ComRun_MLv2.4_dly.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "s7n20Yz89dwe"
   },
   "source": [
    "# Importing and Data Preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "executionInfo": {
     "elapsed": 14,
     "status": "ok",
     "timestamp": 1662732225454,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "_01TAJ-GjI0y"
   },
   "outputs": [],
   "source": [
    "# !pip install h2o # we need to run this at every instance if we are using google colab, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "executionInfo": {
     "elapsed": 13,
     "status": "ok",
     "timestamp": 1662732225455,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "WgcpKC5LKx0n"
   },
   "outputs": [],
   "source": [
    "# import h2o\n",
    "# from h2o.estimators.deeplearning import H2OAutoEncoderEstimator, H2ODeepLearningEstimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "executionInfo": {
     "elapsed": 298,
     "status": "ok",
     "timestamp": 1662732667396,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "paYvI9iy5izh",
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# importing relevant libraries\n",
    "\n",
    "import time\n",
    "\n",
    "import matplotlib.pyplot as plt # plotting graphs\n",
    "import seaborn as sns\n",
    "\n",
    "from pylab import rcParams\n",
    "\n",
    "import sys\n",
    "\n",
    "import numpy as np # linear algebra\n",
    "np.set_printoptions(threshold=sys.maxsize) # view max elements of numpy array\n",
    "\n",
    "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n",
    "# pd.set_option('display.max_columns', None) #pandas shows maximum columns\n",
    "# pd.set_option('display.max_rows', 146) #pandas shows 73 rows\n",
    "\n",
    "import os # for file directories\n",
    "\n",
    "from sklearn.metrics import (confusion_matrix, precision_recall_curve, auc,\n",
    "                             roc_curve, recall_score, classification_report, f1_score,\n",
    "                             precision_recall_fscore_support, make_scorer)\n",
    "from sklearn.ensemble import IsolationForest\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn import model_selection\n",
    "\n",
    "from sklearn.pipeline import Pipeline, make_pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from sklearn.svm import OneClassSVM\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1662732226714,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "pyC1C5CMiiQB",
    "outputId": "9915d28f-0188-4661-9cc6-77c4bfdfefc7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Volume in drive C has no label.\n",
      " Volume Serial Number is 7EC7-B585\n",
      "\n",
      " Directory of C:\\Users\\USER\\Downloads\\MercJbs\n",
      "\n",
      "\n",
      " Directory of C:\\Users\\USER\\Downloads\\MercJbs\n",
      "\n",
      "\n",
      " Directory of C:\\Users\\USER\\Downloads\\MercJbs\n",
      "\n",
      "\n",
      " Directory of C:\\Users\\USER\\Downloads\\MercJbs\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "File Not Found\n"
     ]
    }
   ],
   "source": [
    "# ls # shows our directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662732226715,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "7crl5r-O5qRi"
   },
   "outputs": [],
   "source": [
    "# os.chdir('/content/drive/MyDrive/Colab Files/MercJbs') # change directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 338,
     "status": "ok",
     "timestamp": 1662732227048,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "mOJ6ZDRL-zVD",
    "outputId": "ed2f1ed1-9dc1-4e6c-8de5-0a5faad693d2"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Invalid switch - \"folders\".\n"
     ]
    }
   ],
   "source": [
    "# ls # show current files/folders in directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pwd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "executionInfo": {
     "elapsed": 860,
     "status": "ok",
     "timestamp": 1662732227906,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "QGgz2H94ZFT8"
   },
   "outputs": [],
   "source": [
    "# import dataset\n",
    "df_day = pd.read_csv(r'data/day_approach_maskedID_timeseries_v2.0.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 179
    },
    "executionInfo": {
     "elapsed": 338,
     "status": "ok",
     "timestamp": 1662732228240,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "YNCTAcN7wVKO",
    "outputId": "4ccd4988-ea93-455b-e674-59001600bc25"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>athlete_id</th>\n",
       "      <th>injury</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  nr_sessions  total_km  km_z3-4  km_z5-t1-t2  km_sprinting  \\\n",
       "0           0            1       5.8      0.0          0.6           1.2   \n",
       "\n",
       "   strength_training  hours_alternative  perceived_exertion  \\\n",
       "0                  0                0.0                0.11   \n",
       "\n",
       "   perceived_trainingsuccess  perceived_recovery  nr_sessions_1  total_km_1  \\\n",
       "0                        0.0                0.18              0         0.0   \n",
       "\n",
       "   km_z3-4_1  km_z5-t1-t2_1  km_sprinting_1  strength_training_1  \\\n",
       "0        0.0            0.0             0.0                    0   \n",
       "\n",
       "   hours_alternative_1  perceived_exertion_1  perceived_trainingsuccess_1  \\\n",
       "0                  0.0                 -0.01                        -0.01   \n",
       "\n",
       "   perceived_recovery_1  nr_sessions_2  total_km_2  km_z3-4_2  km_z5-t1-t2_2  \\\n",
       "0                 -0.01              1         0.0        0.0            0.0   \n",
       "\n",
       "   km_sprinting_2  strength_training_2  hours_alternative_2  \\\n",
       "0             0.0                    1                  0.0   \n",
       "\n",
       "   perceived_exertion_2  perceived_trainingsuccess_2  perceived_recovery_2  \\\n",
       "0                   0.1                          0.0                  0.17   \n",
       "\n",
       "   nr_sessions_3  total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3  \\\n",
       "0              0         0.0        0.0            0.0             0.0   \n",
       "\n",
       "   strength_training_3  hours_alternative_3  perceived_exertion_3  \\\n",
       "0                    0                  0.0                 -0.01   \n",
       "\n",
       "   perceived_trainingsuccess_3  perceived_recovery_3  nr_sessions_4  \\\n",
       "0                        -0.01                 -0.01              1   \n",
       "\n",
       "   total_km_4  km_z3-4_4  km_z5-t1-t2_4  km_sprinting_4  strength_training_4  \\\n",
       "0         0.0        0.0            0.0             0.0                    0   \n",
       "\n",
       "   hours_alternative_4  perceived_exertion_4  perceived_trainingsuccess_4  \\\n",
       "0                 1.08                  0.08                          0.0   \n",
       "\n",
       "   perceived_recovery_4  nr_sessions_5  total_km_5  km_z3-4_5  km_z5-t1-t2_5  \\\n",
       "0                  0.18              1        16.4       10.0            0.0   \n",
       "\n",
       "   km_sprinting_5  strength_training_5  hours_alternative_5  \\\n",
       "0             0.0                    1                  0.0   \n",
       "\n",
       "   perceived_exertion_5  perceived_trainingsuccess_5  perceived_recovery_5  \\\n",
       "0                  0.11                          0.0                  0.17   \n",
       "\n",
       "   nr_sessions_6  total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6  \\\n",
       "0              1         0.0        0.0            0.0             0.0   \n",
       "\n",
       "   strength_training_6  hours_alternative_6  perceived_exertion_6  \\\n",
       "0                    0                  1.0                   0.1   \n",
       "\n",
       "   perceived_trainingsuccess_6  perceived_recovery_6  athlete_id  injury  date  \n",
       "0                          0.0                  0.15           0       0     0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_day.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1662732228241,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "MH3yW-pcDx3C"
   },
   "outputs": [],
   "source": [
    "df_day.drop(['Unnamed: 0'], axis = 1, inplace = True)\n",
    "df_day['date'] = df_day['date'].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1662732228242,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Wowekz941-d6"
   },
   "outputs": [],
   "source": [
    "# df_day['injury'] = df_day['injury'].apply(lambda x: 'yes' if x == 1 else 'no')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 393,
     "status": "ok",
     "timestamp": 1662732228627,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "XB12N9N82dHt",
    "outputId": "1b44ff1a-ba23-4255-b725-d70dc5e0d8b5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0, 1]\n",
       "Categories (2, int64): [0, 1]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_day[df_day.nunique()[df_day.nunique() < 30].index] = df_day[df_day.nunique()[df_day.nunique() < 30].index].astype('category')\n",
    "df_day['injury'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 305
    },
    "executionInfo": {
     "elapsed": 17,
     "status": "ok",
     "timestamp": 1662732228628,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "VJlFVtqGDhb-",
    "outputId": "afcc131a-9e63-42bd-c90c-1198f076d8d3"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>athlete_id</th>\n",
       "      <th>injury</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5.8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "      <td>1</td>\n",
       "      <td>5.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "      <td>1</td>\n",
       "      <td>5.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "      <td>1</td>\n",
       "      <td>5.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.15</td>\n",
       "      <td>1</td>\n",
       "      <td>5.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>17.6</td>\n",
       "      <td>7.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  nr_sessions  total_km  km_z3-4  km_z5-t1-t2  km_sprinting strength_training  \\\n",
       "0           1       5.8      0.0          0.6           1.2                 0   \n",
       "1           0       0.0      0.0          0.0           0.0                 0   \n",
       "2           1       0.0      0.0          0.0           0.0                 1   \n",
       "3           0       0.0      0.0          0.0           0.0                 0   \n",
       "4           1       0.0      0.0          0.0           0.0                 0   \n",
       "\n",
       "   hours_alternative  perceived_exertion  perceived_trainingsuccess  \\\n",
       "0               0.00                0.11                       0.00   \n",
       "1               0.00               -0.01                      -0.01   \n",
       "2               0.00                0.10                       0.00   \n",
       "3               0.00               -0.01                      -0.01   \n",
       "4               1.08                0.08                       0.00   \n",
       "\n",
       "   perceived_recovery nr_sessions_1  total_km_1  km_z3-4_1  km_z5-t1-t2_1  \\\n",
       "0                0.18             0         0.0        0.0            0.0   \n",
       "1               -0.01             1         0.0        0.0            0.0   \n",
       "2                0.17             0         0.0        0.0            0.0   \n",
       "3               -0.01             1         0.0        0.0            0.0   \n",
       "4                0.18             1        16.4       10.0            0.0   \n",
       "\n",
       "   km_sprinting_1 strength_training_1  hours_alternative_1  \\\n",
       "0             0.0                   0                 0.00   \n",
       "1             0.0                   1                 0.00   \n",
       "2             0.0                   0                 0.00   \n",
       "3             0.0                   0                 1.08   \n",
       "4             0.0                   1                 0.00   \n",
       "\n",
       "   perceived_exertion_1  perceived_trainingsuccess_1  perceived_recovery_1  \\\n",
       "0                 -0.01                        -0.01                 -0.01   \n",
       "1                  0.10                         0.00                  0.17   \n",
       "2                 -0.01                        -0.01                 -0.01   \n",
       "3                  0.08                         0.00                  0.18   \n",
       "4                  0.11                         0.00                  0.17   \n",
       "\n",
       "  nr_sessions_2  total_km_2  km_z3-4_2  km_z5-t1-t2_2  km_sprinting_2  \\\n",
       "0             1         0.0        0.0            0.0             0.0   \n",
       "1             0         0.0        0.0            0.0             0.0   \n",
       "2             1         0.0        0.0            0.0             0.0   \n",
       "3             1        16.4       10.0            0.0             0.0   \n",
       "4             1         0.0        0.0            0.0             0.0   \n",
       "\n",
       "  strength_training_2  hours_alternative_2  perceived_exertion_2  \\\n",
       "0                   1                 0.00                  0.10   \n",
       "1                   0                 0.00                 -0.01   \n",
       "2                   0                 1.08                  0.08   \n",
       "3                   1                 0.00                  0.11   \n",
       "4                   0                 1.00                  0.10   \n",
       "\n",
       "   perceived_trainingsuccess_2  perceived_recovery_2 nr_sessions_3  \\\n",
       "0                         0.00                  0.17             0   \n",
       "1                        -0.01                 -0.01             1   \n",
       "2                         0.00                  0.18             1   \n",
       "3                         0.00                  0.17             1   \n",
       "4                         0.00                  0.15             1   \n",
       "\n",
       "   total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3 strength_training_3  \\\n",
       "0         0.0        0.0            0.0             0.0                   0   \n",
       "1         0.0        0.0            0.0             0.0                   0   \n",
       "2        16.4       10.0            0.0             0.0                   1   \n",
       "3         0.0        0.0            0.0             0.0                   0   \n",
       "4         5.2        0.0            0.5             1.2                   0   \n",
       "\n",
       "   hours_alternative_3  perceived_exertion_3  perceived_trainingsuccess_3  \\\n",
       "0                 0.00                 -0.01                        -0.01   \n",
       "1                 1.08                  0.08                         0.00   \n",
       "2                 0.00                  0.11                         0.00   \n",
       "3                 1.00                  0.10                         0.00   \n",
       "4                 0.00                  0.10                         0.00   \n",
       "\n",
       "   perceived_recovery_3 nr_sessions_4  total_km_4  km_z3-4_4  km_z5-t1-t2_4  \\\n",
       "0                 -0.01             1         0.0        0.0            0.0   \n",
       "1                  0.18             1        16.4       10.0            0.0   \n",
       "2                  0.17             1         0.0        0.0            0.0   \n",
       "3                  0.15             1         5.2        0.0            0.5   \n",
       "4                  0.17             0         0.0        0.0            0.0   \n",
       "\n",
       "   km_sprinting_4 strength_training_4  hours_alternative_4  \\\n",
       "0             0.0                   0                 1.08   \n",
       "1             0.0                   1                 0.00   \n",
       "2             0.0                   0                 1.00   \n",
       "3             1.2                   0                 0.00   \n",
       "4             0.0                   0                 0.00   \n",
       "\n",
       "   perceived_exertion_4  perceived_trainingsuccess_4  perceived_recovery_4  \\\n",
       "0                  0.08                         0.00                  0.18   \n",
       "1                  0.11                         0.00                  0.17   \n",
       "2                  0.10                         0.00                  0.15   \n",
       "3                  0.10                         0.00                  0.17   \n",
       "4                 -0.01                        -0.01                 -0.01   \n",
       "\n",
       "  nr_sessions_5  total_km_5  km_z3-4_5  km_z5-t1-t2_5  km_sprinting_5  \\\n",
       "0             1        16.4       10.0            0.0             0.0   \n",
       "1             1         0.0        0.0            0.0             0.0   \n",
       "2             1         5.2        0.0            0.5             1.2   \n",
       "3             0         0.0        0.0            0.0             0.0   \n",
       "4             1         0.0        0.0            0.0             0.0   \n",
       "\n",
       "  strength_training_5  hours_alternative_5  perceived_exertion_5  \\\n",
       "0                   1                  0.0                  0.11   \n",
       "1                   0                  1.0                  0.10   \n",
       "2                   0                  0.0                  0.10   \n",
       "3                   0                  0.0                 -0.01   \n",
       "4                   1                  0.0                  0.10   \n",
       "\n",
       "   perceived_trainingsuccess_5  perceived_recovery_5 nr_sessions_6  \\\n",
       "0                         0.00                  0.17             1   \n",
       "1                         0.00                  0.15             1   \n",
       "2                         0.00                  0.17             0   \n",
       "3                        -0.01                 -0.01             1   \n",
       "4                         0.00                  0.17             1   \n",
       "\n",
       "   total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6 strength_training_6  \\\n",
       "0         0.0        0.0            0.0             0.0                   0   \n",
       "1         5.2        0.0            0.5             1.2                   0   \n",
       "2         0.0        0.0            0.0             0.0                   0   \n",
       "3         0.0        0.0            0.0             0.0                   1   \n",
       "4        17.6        7.2            0.0             0.0                   0   \n",
       "\n",
       "   hours_alternative_6  perceived_exertion_6  perceived_trainingsuccess_6  \\\n",
       "0                  1.0                  0.10                         0.00   \n",
       "1                  0.0                  0.10                         0.00   \n",
       "2                  0.0                 -0.01                        -0.01   \n",
       "3                  0.0                  0.10                         0.00   \n",
       "4                  0.0                  0.11                         0.00   \n",
       "\n",
       "   perceived_recovery_6  athlete_id injury date  \n",
       "0                  0.15           0      0    0  \n",
       "1                  0.17           0      0    1  \n",
       "2                 -0.01           0      0    2  \n",
       "3                  0.17           0      0    3  \n",
       "4                  0.17           0      0    4  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_day.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 16,
     "status": "ok",
     "timestamp": 1662732228629,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "nmUf6Fhz69V4",
    "outputId": "32fed26a-479e-46d4-af18-4896e7e9446a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 42766 entries, 0 to 42765\n",
      "Data columns (total 73 columns):\n",
      " #   Column                       Non-Null Count  Dtype   \n",
      "---  ------                       --------------  -----   \n",
      " 0   nr_sessions                  42766 non-null  category\n",
      " 1   total_km                     42766 non-null  float64 \n",
      " 2   km_z3-4                      42766 non-null  float64 \n",
      " 3   km_z5-t1-t2                  42766 non-null  float64 \n",
      " 4   km_sprinting                 42766 non-null  float64 \n",
      " 5   strength_training            42766 non-null  category\n",
      " 6   hours_alternative            42766 non-null  float64 \n",
      " 7   perceived_exertion           42766 non-null  float64 \n",
      " 8   perceived_trainingsuccess    42766 non-null  float64 \n",
      " 9   perceived_recovery           42766 non-null  float64 \n",
      " 10  nr_sessions_1                42766 non-null  category\n",
      " 11  total_km_1                   42766 non-null  float64 \n",
      " 12  km_z3-4_1                    42766 non-null  float64 \n",
      " 13  km_z5-t1-t2_1                42766 non-null  float64 \n",
      " 14  km_sprinting_1               42766 non-null  float64 \n",
      " 15  strength_training_1          42766 non-null  category\n",
      " 16  hours_alternative_1          42766 non-null  float64 \n",
      " 17  perceived_exertion_1         42766 non-null  float64 \n",
      " 18  perceived_trainingsuccess_1  42766 non-null  float64 \n",
      " 19  perceived_recovery_1         42766 non-null  float64 \n",
      " 20  nr_sessions_2                42766 non-null  category\n",
      " 21  total_km_2                   42766 non-null  float64 \n",
      " 22  km_z3-4_2                    42766 non-null  float64 \n",
      " 23  km_z5-t1-t2_2                42766 non-null  float64 \n",
      " 24  km_sprinting_2               42766 non-null  float64 \n",
      " 25  strength_training_2          42766 non-null  category\n",
      " 26  hours_alternative_2          42766 non-null  float64 \n",
      " 27  perceived_exertion_2         42766 non-null  float64 \n",
      " 28  perceived_trainingsuccess_2  42766 non-null  float64 \n",
      " 29  perceived_recovery_2         42766 non-null  float64 \n",
      " 30  nr_sessions_3                42766 non-null  category\n",
      " 31  total_km_3                   42766 non-null  float64 \n",
      " 32  km_z3-4_3                    42766 non-null  float64 \n",
      " 33  km_z5-t1-t2_3                42766 non-null  float64 \n",
      " 34  km_sprinting_3               42766 non-null  float64 \n",
      " 35  strength_training_3          42766 non-null  category\n",
      " 36  hours_alternative_3          42766 non-null  float64 \n",
      " 37  perceived_exertion_3         42766 non-null  float64 \n",
      " 38  perceived_trainingsuccess_3  42766 non-null  float64 \n",
      " 39  perceived_recovery_3         42766 non-null  float64 \n",
      " 40  nr_sessions_4                42766 non-null  category\n",
      " 41  total_km_4                   42766 non-null  float64 \n",
      " 42  km_z3-4_4                    42766 non-null  float64 \n",
      " 43  km_z5-t1-t2_4                42766 non-null  float64 \n",
      " 44  km_sprinting_4               42766 non-null  float64 \n",
      " 45  strength_training_4          42766 non-null  category\n",
      " 46  hours_alternative_4          42766 non-null  float64 \n",
      " 47  perceived_exertion_4         42766 non-null  float64 \n",
      " 48  perceived_trainingsuccess_4  42766 non-null  float64 \n",
      " 49  perceived_recovery_4         42766 non-null  float64 \n",
      " 50  nr_sessions_5                42766 non-null  category\n",
      " 51  total_km_5                   42766 non-null  float64 \n",
      " 52  km_z3-4_5                    42766 non-null  float64 \n",
      " 53  km_z5-t1-t2_5                42766 non-null  float64 \n",
      " 54  km_sprinting_5               42766 non-null  float64 \n",
      " 55  strength_training_5          42766 non-null  category\n",
      " 56  hours_alternative_5          42766 non-null  float64 \n",
      " 57  perceived_exertion_5         42766 non-null  float64 \n",
      " 58  perceived_trainingsuccess_5  42766 non-null  float64 \n",
      " 59  perceived_recovery_5         42766 non-null  float64 \n",
      " 60  nr_sessions_6                42766 non-null  category\n",
      " 61  total_km_6                   42766 non-null  float64 \n",
      " 62  km_z3-4_6                    42766 non-null  float64 \n",
      " 63  km_z5-t1-t2_6                42766 non-null  float64 \n",
      " 64  km_sprinting_6               42766 non-null  float64 \n",
      " 65  strength_training_6          42766 non-null  category\n",
      " 66  hours_alternative_6          42766 non-null  float64 \n",
      " 67  perceived_exertion_6         42766 non-null  float64 \n",
      " 68  perceived_trainingsuccess_6  42766 non-null  float64 \n",
      " 69  perceived_recovery_6         42766 non-null  float64 \n",
      " 70  athlete_id                   42766 non-null  int64   \n",
      " 71  injury                       42766 non-null  category\n",
      " 72  date                         42766 non-null  category\n",
      "dtypes: category(16), float64(56), int64(1)\n",
      "memory usage: 19.4 MB\n"
     ]
    }
   ],
   "source": [
    "df_day.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "executionInfo": {
     "elapsed": 1845,
     "status": "ok",
     "timestamp": 1662705831978,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "-NrcE8oR3DWM"
   },
   "outputs": [],
   "source": [
    "df_day.to_csv('data/day_approach_maskedID_timeseries_v3.0.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CPiXjFJ1YEGP"
   },
   "source": [
    "# Introduction\n",
    "\n",
    "In this notebook, we run some advance classification ML algorithms on our datasets. As a result of our dataset having many rows/entries (42766) and columns (73), we use a dimensionality reduction technique (PCA) to see the variability on the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mrMcJotaSJyE"
   },
   "source": [
    "<!-- ## PCA -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ovll_ZxqQ0TF"
   },
   "source": [
    "<!-- from sklearn.model_selection import train_test_split\n",
    "from sklearn.decomposition import PCA -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XYly1zejVFcA"
   },
   "source": [
    "<!-- Xd = df_day.drop(columns = ['injury'])\n",
    "yd = df_day['injury'] -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "006qxTRhSTm3"
   },
   "source": [
    "<!-- pca = make_pipeline(StandardScaler(), PCA(.95))\n",
    "# pca = PCA()\n",
    "pca.fit_transform(Xd)\n",
    "pca -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "3ZW-PDn-udze"
   },
   "source": [
    "<!-- **Explained variance** is the amount of variance explained by each of the selected components.  -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "fe7WxdrpVbc4"
   },
   "source": [
    "<!-- # pca[1].n_components_ -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_qDdDcPEUn1k"
   },
   "source": [
    "<!-- explained = pca[1].explained_variance_ratio_\n",
    "explained -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LiLqc2RrnTs2"
   },
   "source": [
    "<!-- cumulative_explained = explained.cumsum()\n",
    "\n",
    "cumulative_explained -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "thW8N6EhoeE9"
   },
   "source": [
    "<!-- pca[1].components_.shape -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LTAq_wQsowSW"
   },
   "source": [
    "<!-- plt.figure(figsize=(15,7.5))\n",
    "plt.plot(range(0, 53), explained, label='Explained Variance')\n",
    "plt.plot(range(0, 53), cumulative_explained, label='Cumulative Explained Variance')\n",
    "plt.xlabel('Principal Components')\n",
    "plt.vlines(x=42, ymin= -0.1, ymax=1.1, colors= 'k', label = \"42nd Component\")\n",
    "plt.hlines(y=0.90, xmin=-0.5, xmax=72, colors= 'g', linestyles='--', label = \"90% Explained\")\n",
    "plt.title('Explained Variance Ratio')\n",
    "plt.legend()\n",
    "plt.show() -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Kof3dEI9YB99"
   },
   "source": [
    "# Model 1 : Autoencoder\n",
    "**What is Autoencoder?**\n",
    "\n",
    "Autoencoder is an unsupervised artificial neural network that learns how to efficiently compress and encode data then learns how to reconstruct the data back from the reduced encoded representation to a representation that is as close to the original input as possible. Autoencoder, by design, reduces data dimensions by learning how to ignore the noise in the data. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5zuZXUOEpg-P"
   },
   "source": [
    "**Why Autoencoder works for Anomaly Detection?**\n",
    "\n",
    "Autoencoders are great in learning what is normal. I can identify the normal region ina distribution then anything that is considered an outlier will return a high reconstruction error.\n",
    "\n",
    "Let's start by installing some pre-requisite libraries, h20. \n",
    "\n",
    "**H2O** is an open-source library that can be used to do a number of machine learning tasks. It has a function for the autoencoders. Let's have a look"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wRqCXGxmZhAr"
   },
   "source": [
    "## Starting and initializing H2O server."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 491
    },
    "executionInfo": {
     "elapsed": 12449,
     "status": "ok",
     "timestamp": 1661545719192,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "wQFBKEG0ZjQ3",
    "outputId": "b9ee3732-e2bb-473d-9bfe-df0bac427685"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321 ..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "  Java Version: openjdk version \"11.0.16\" 2022-07-19; OpenJDK Runtime Environment (build 11.0.16+8-post-Ubuntu-0ubuntu118.04); OpenJDK 64-Bit Server VM (build 11.0.16+8-post-Ubuntu-0ubuntu118.04, mixed mode, sharing)\n",
      "  Starting server from /usr/local/lib/python3.7/dist-packages/h2o/backend/bin/h2o.jar\n",
      "  Ice root: /tmp/tmplfwolugk\n",
      "  JVM stdout: /tmp/tmplfwolugk/h2o_unknownUser_started_from_python.out\n",
      "  JVM stderr: /tmp/tmplfwolugk/h2o_unknownUser_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321 ... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O_cluster_uptime:</td>\n",
       "<td>05 secs</td></tr>\n",
       "<tr><td>H2O_cluster_timezone:</td>\n",
       "<td>Etc/UTC</td></tr>\n",
       "<tr><td>H2O_data_parsing_timezone:</td>\n",
       "<td>UTC</td></tr>\n",
       "<tr><td>H2O_cluster_version:</td>\n",
       "<td>3.36.1.4</td></tr>\n",
       "<tr><td>H2O_cluster_version_age:</td>\n",
       "<td>23 days </td></tr>\n",
       "<tr><td>H2O_cluster_name:</td>\n",
       "<td>H2O_from_python_unknownUser_vhcj8a</td></tr>\n",
       "<tr><td>H2O_cluster_total_nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O_cluster_free_memory:</td>\n",
       "<td>3.172 Gb</td></tr>\n",
       "<tr><td>H2O_cluster_total_cores:</td>\n",
       "<td>2</td></tr>\n",
       "<tr><td>H2O_cluster_allowed_cores:</td>\n",
       "<td>2</td></tr>\n",
       "<tr><td>H2O_cluster_status:</td>\n",
       "<td>locked, healthy</td></tr>\n",
       "<tr><td>H2O_connection_url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O_connection_proxy:</td>\n",
       "<td>{\"http\": null, \"https\": null}</td></tr>\n",
       "<tr><td>H2O_internal_security:</td>\n",
       "<td>False</td></tr>\n",
       "<tr><td>Python_version:</td>\n",
       "<td>3.7.13 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ----------------------------------\n",
       "H2O_cluster_uptime:         05 secs\n",
       "H2O_cluster_timezone:       Etc/UTC\n",
       "H2O_data_parsing_timezone:  UTC\n",
       "H2O_cluster_version:        3.36.1.4\n",
       "H2O_cluster_version_age:    23 days\n",
       "H2O_cluster_name:           H2O_from_python_unknownUser_vhcj8a\n",
       "H2O_cluster_total_nodes:    1\n",
       "H2O_cluster_free_memory:    3.172 Gb\n",
       "H2O_cluster_total_cores:    2\n",
       "H2O_cluster_allowed_cores:  2\n",
       "H2O_cluster_status:         locked, healthy\n",
       "H2O_connection_url:         http://127.0.0.1:54321\n",
       "H2O_connection_proxy:       {\"http\": null, \"https\": null}\n",
       "H2O_internal_security:      False\n",
       "Python_version:             3.7.13 final\n",
       "--------------------------  ----------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "h2o.init() # initializing h2o server\n",
    "# h2o.remove_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wI2O7uOQZsZU"
   },
   "source": [
    "We now load the dataset into a H2O dataframe and splitting the DS."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 6950,
     "status": "ok",
     "timestamp": 1661545778080,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "KOf-6XZracj9",
    "outputId": "07f7059d-cbf1-419c-fff1-f52daa0c0f12"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n"
     ]
    }
   ],
   "source": [
    "hdf_day = h2o.import_file(r\"data/day_approach_maskedID_timeseries_v3.0.csv\") #import the file as a h2o file\n",
    "hdf_day[\"date\"] = hdf_day[\"date\"].asfactor() # coonvert date to factor i.e categorical\n",
    "hdf_day[\"injury\"] = hdf_day[\"injury\"].asfactor()\n",
    "features_day  = hdf_day.drop(['athlete_id', 'C1'], axis=1) # we do not need the athlete id again alonside the 'C1' column which is an automatic addition by h2o import\n",
    "train_day , test_day = features_day.split_frame([0.8]) # 80% of data for training and 20% for testing "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "e3qmuar_rKP-"
   },
   "source": [
    "We used a 80-20 split between the training and testing because of the imbalance in the dataset. This spit enables us to get enough of the undersampled class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 137
    },
    "executionInfo": {
     "elapsed": 1539,
     "status": "ok",
     "timestamp": 1661545805988,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "KbNDIY3GGQ3x",
    "outputId": "1d9f0ed5-ad70-43c9-997e-6fdee67a3839"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th style=\"text-align: right;\">  nr__sessions</th><th style=\"text-align: right;\">  total_km</th><th style=\"text-align: right;\">  km_z3-4</th><th style=\"text-align: right;\">  km_z5-t1-t2</th><th style=\"text-align: right;\">  km_sprinting</th><th style=\"text-align: right;\">  strength_training</th><th style=\"text-align: right;\">  hours_alternative</th><th style=\"text-align: right;\">  perceived_exertion</th><th style=\"text-align: right;\">  perceived_trainingsuccess</th><th style=\"text-align: right;\">  perceived_recovery</th><th style=\"text-align: right;\">  nr__sessions_1</th><th style=\"text-align: right;\">  total_km_1</th><th style=\"text-align: right;\">  km_z3-4_1</th><th style=\"text-align: right;\">  km_z5-t1-t2_1</th><th style=\"text-align: right;\">  km_sprinting_1</th><th style=\"text-align: right;\">  strength_training_1</th><th style=\"text-align: right;\">  hours_alternative_1</th><th style=\"text-align: right;\">  perceived_exertion_1</th><th style=\"text-align: right;\">  perceived_trainingsuccess_1</th><th style=\"text-align: right;\">  perceived_recovery_1</th><th style=\"text-align: right;\">  nr__sessions_2</th><th style=\"text-align: right;\">  total_km_2</th><th style=\"text-align: right;\">  km_z3-4_2</th><th style=\"text-align: right;\">  km_z5-t1-t2_2</th><th style=\"text-align: right;\">  km_sprinting_2</th><th style=\"text-align: right;\">  strength_training_2</th><th style=\"text-align: right;\">  hours_alternative_2</th><th style=\"text-align: right;\">  perceived_exertion_2</th><th style=\"text-align: right;\">  perceived_trainingsuccess_2</th><th style=\"text-align: right;\">  perceived_recovery_2</th><th style=\"text-align: right;\">  nr__sessions_3</th><th style=\"text-align: right;\">  total_km_3</th><th style=\"text-align: right;\">  km_z3-4_3</th><th style=\"text-align: right;\">  km_z5-t1-t2_3</th><th style=\"text-align: right;\">  km_sprinting_3</th><th style=\"text-align: right;\">  strength_training_3</th><th style=\"text-align: right;\">  hours_alternative_3</th><th style=\"text-align: right;\">  perceived_exertion_3</th><th style=\"text-align: right;\">  perceived_trainingsuccess_3</th><th style=\"text-align: right;\">  perceived_recovery_3</th><th style=\"text-align: right;\">  nr__sessions_4</th><th style=\"text-align: right;\">  total_km_4</th><th style=\"text-align: right;\">  km_z3-4_4</th><th style=\"text-align: right;\">  km_z5-t1-t2_4</th><th style=\"text-align: right;\">  km_sprinting_4</th><th style=\"text-align: right;\">  strength_training_4</th><th style=\"text-align: right;\">  hours_alternative_4</th><th style=\"text-align: right;\">  perceived_exertion_4</th><th style=\"text-align: right;\">  perceived_trainingsuccess_4</th><th style=\"text-align: right;\">  perceived_recovery_4</th><th style=\"text-align: right;\">  nr__sessions_5</th><th style=\"text-align: right;\">  total_km_5</th><th style=\"text-align: right;\">  km_z3-4_5</th><th style=\"text-align: right;\">  km_z5-t1-t2_5</th><th style=\"text-align: right;\">  km_sprinting_5</th><th style=\"text-align: right;\">  strength_training_5</th><th style=\"text-align: right;\">  hours_alternative_5</th><th style=\"text-align: right;\">  perceived_exertion_5</th><th style=\"text-align: right;\">  perceived_trainingsuccess_5</th><th style=\"text-align: right;\">  perceived_recovery_5</th><th style=\"text-align: right;\">  nr__sessions_6</th><th style=\"text-align: right;\">  total_km_6</th><th style=\"text-align: right;\">  km_z3-4_6</th><th style=\"text-align: right;\">  km_z5-t1-t2_6</th><th style=\"text-align: right;\">  km_sprinting_6</th><th style=\"text-align: right;\">  strength_training_6</th><th style=\"text-align: right;\">  hours_alternative_6</th><th style=\"text-align: right;\">  perceived_exertion_6</th><th style=\"text-align: right;\">  perceived_trainingsuccess_6</th><th style=\"text-align: right;\">  perceived_recovery_6</th><th style=\"text-align: right;\">  injury</th><th style=\"text-align: right;\">  date</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td style=\"text-align: right;\">             1</td><td style=\"text-align: right;\">       5.8</td><td style=\"text-align: right;\">        0</td><td style=\"text-align: right;\">          0.6</td><td style=\"text-align: right;\">           1.2</td><td style=\"text-align: right;\">                  0</td><td style=\"text-align: right;\">                  0</td><td style=\"text-align: right;\">                0.11</td><td style=\"text-align: right;\">                       0   </td><td style=\"text-align: right;\">                0.18</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">                        -0.01</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                 0   </td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 0   </td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">                        -0.01</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 1.08</td><td style=\"text-align: right;\">                  0.08</td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.18</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">        16.4</td><td style=\"text-align: right;\">         10</td><td style=\"text-align: right;\">            0  </td><td style=\"text-align: right;\">             0  </td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                  0.11</td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">            0  </td><td style=\"text-align: right;\">             0  </td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.15</td><td style=\"text-align: right;\">       0</td><td style=\"text-align: right;\">     0</td></tr>\n",
       "<tr><td style=\"text-align: right;\">             0</td><td style=\"text-align: right;\">       0  </td><td style=\"text-align: right;\">        0</td><td style=\"text-align: right;\">          0  </td><td style=\"text-align: right;\">           0  </td><td style=\"text-align: right;\">                  0</td><td style=\"text-align: right;\">                  0</td><td style=\"text-align: right;\">               -0.01</td><td style=\"text-align: right;\">                      -0.01</td><td style=\"text-align: right;\">               -0.01</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 0   </td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">                        -0.01</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 1.08</td><td style=\"text-align: right;\">                  0.08</td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.18</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">        16.4</td><td style=\"text-align: right;\">         10</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                 0   </td><td style=\"text-align: right;\">                  0.11</td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">            0  </td><td style=\"text-align: right;\">             0  </td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.15</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         5.2</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">            0.5</td><td style=\"text-align: right;\">             1.2</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">       0</td><td style=\"text-align: right;\">     1</td></tr>\n",
       "<tr><td style=\"text-align: right;\">             1</td><td style=\"text-align: right;\">       0  </td><td style=\"text-align: right;\">        0</td><td style=\"text-align: right;\">          0  </td><td style=\"text-align: right;\">           0  </td><td style=\"text-align: right;\">                  1</td><td style=\"text-align: right;\">                  0</td><td style=\"text-align: right;\">                0.1 </td><td style=\"text-align: right;\">                       0   </td><td style=\"text-align: right;\">                0.17</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">                        -0.01</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">           0</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 1.08</td><td style=\"text-align: right;\">                  0.08</td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.18</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">        16.4</td><td style=\"text-align: right;\">         10</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    1</td><td style=\"text-align: right;\">                 0   </td><td style=\"text-align: right;\">                  0.11</td><td style=\"text-align: right;\">                         0   </td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">              0</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 1   </td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.15</td><td style=\"text-align: right;\">               1</td><td style=\"text-align: right;\">         5.2</td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">            0.5</td><td style=\"text-align: right;\">             1.2</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                  0.1 </td><td style=\"text-align: right;\">                            0</td><td style=\"text-align: right;\">                  0.17</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">         0  </td><td style=\"text-align: right;\">          0</td><td style=\"text-align: right;\">            0  </td><td style=\"text-align: right;\">             0  </td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                    0</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">                        -0.01</td><td style=\"text-align: right;\">                 -0.01</td><td style=\"text-align: right;\">       0</td><td style=\"text-align: right;\">     2</td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_day.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 466,
     "status": "ok",
     "timestamp": 1661545813800,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "QSSl9ZHW-uwg",
    "outputId": "4c22aa90-b431-4ed4-dac5-b57c3cbd59e3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'nr__sessions': 'int',\n",
       " 'total_km': 'real',\n",
       " 'km_z3-4': 'real',\n",
       " 'km_z5-t1-t2': 'real',\n",
       " 'km_sprinting': 'real',\n",
       " 'strength_training': 'int',\n",
       " 'hours_alternative': 'real',\n",
       " 'perceived_exertion': 'real',\n",
       " 'perceived_trainingsuccess': 'real',\n",
       " 'perceived_recovery': 'real',\n",
       " 'nr__sessions_1': 'int',\n",
       " 'total_km_1': 'real',\n",
       " 'km_z3-4_1': 'real',\n",
       " 'km_z5-t1-t2_1': 'real',\n",
       " 'km_sprinting_1': 'real',\n",
       " 'strength_training_1': 'int',\n",
       " 'hours_alternative_1': 'real',\n",
       " 'perceived_exertion_1': 'real',\n",
       " 'perceived_trainingsuccess_1': 'real',\n",
       " 'perceived_recovery_1': 'real',\n",
       " 'nr__sessions_2': 'int',\n",
       " 'total_km_2': 'real',\n",
       " 'km_z3-4_2': 'real',\n",
       " 'km_z5-t1-t2_2': 'real',\n",
       " 'km_sprinting_2': 'real',\n",
       " 'strength_training_2': 'int',\n",
       " 'hours_alternative_2': 'real',\n",
       " 'perceived_exertion_2': 'real',\n",
       " 'perceived_trainingsuccess_2': 'real',\n",
       " 'perceived_recovery_2': 'real',\n",
       " 'nr__sessions_3': 'int',\n",
       " 'total_km_3': 'real',\n",
       " 'km_z3-4_3': 'real',\n",
       " 'km_z5-t1-t2_3': 'real',\n",
       " 'km_sprinting_3': 'real',\n",
       " 'strength_training_3': 'int',\n",
       " 'hours_alternative_3': 'real',\n",
       " 'perceived_exertion_3': 'real',\n",
       " 'perceived_trainingsuccess_3': 'real',\n",
       " 'perceived_recovery_3': 'real',\n",
       " 'nr__sessions_4': 'int',\n",
       " 'total_km_4': 'real',\n",
       " 'km_z3-4_4': 'real',\n",
       " 'km_z5-t1-t2_4': 'real',\n",
       " 'km_sprinting_4': 'real',\n",
       " 'strength_training_4': 'int',\n",
       " 'hours_alternative_4': 'real',\n",
       " 'perceived_exertion_4': 'real',\n",
       " 'perceived_trainingsuccess_4': 'real',\n",
       " 'perceived_recovery_4': 'real',\n",
       " 'nr__sessions_5': 'int',\n",
       " 'total_km_5': 'real',\n",
       " 'km_z3-4_5': 'real',\n",
       " 'km_z5-t1-t2_5': 'real',\n",
       " 'km_sprinting_5': 'real',\n",
       " 'strength_training_5': 'int',\n",
       " 'hours_alternative_5': 'real',\n",
       " 'perceived_exertion_5': 'real',\n",
       " 'perceived_trainingsuccess_5': 'real',\n",
       " 'perceived_recovery_5': 'real',\n",
       " 'nr__sessions_6': 'int',\n",
       " 'total_km_6': 'real',\n",
       " 'km_z3-4_6': 'real',\n",
       " 'km_z5-t1-t2_6': 'real',\n",
       " 'km_sprinting_6': 'real',\n",
       " 'strength_training_6': 'int',\n",
       " 'hours_alternative_6': 'real',\n",
       " 'perceived_exertion_6': 'real',\n",
       " 'perceived_trainingsuccess_6': 'real',\n",
       " 'perceived_recovery_6': 'real',\n",
       " 'injury': 'enum',\n",
       " 'date': 'enum'}"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features_day.types"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "17tYnFWfcV9M"
   },
   "source": [
    "We now preprocess the dataset by seperating training and test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ByKA-DPJdcHG"
   },
   "outputs": [],
   "source": [
    "# converting to pandas dataframe\n",
    "train_day_df = train_day.as_data_frame()\n",
    "test_day_df = test_day.as_data_frame()\n",
    "\n",
    "# noinjury_df_tr = train_day_df[train_day_df['injury'] == 0]\n",
    "# injury_df_tr = train_day_df[train_day_df['injury'] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "aaIfZzwecTMK"
   },
   "outputs": [],
   "source": [
    "# drop the Class variable\n",
    "\n",
    "Y_test_df = test_day_df['injury'] # true labels of the testing set\n",
    "Y_test_df_ = pd.DataFrame(Y_test_df) # convert to pandas dataframe from pandas series\n",
    "test_df = test_day_df.drop(['injury'], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 4376,
     "status": "ok",
     "timestamp": 1661545963462,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "DGbyLdAVfP44",
    "outputId": "24720f8e-f68f-4ea1-c468-ca32e7001253"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n",
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n",
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n"
     ]
    }
   ],
   "source": [
    "# converting to h2o frame\n",
    "Y_test_h2o = h2o.H2OFrame(Y_test_df_) \n",
    "train_h2o = h2o.H2OFrame(train_day_df) \n",
    "test_h2o = h2o.H2OFrame(test_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ddpFQeSPgvDW"
   },
   "outputs": [],
   "source": [
    "cols = train_h2o.columns\n",
    "cols.remove('injury')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 15,
     "status": "ok",
     "timestamp": 1661192382378,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Z78C7aMEClx0",
    "outputId": "876eb936-d137-492d-a7a8-88cd3706c556"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['nr__sessions',\n",
       " 'total_km',\n",
       " 'km_z3-4',\n",
       " 'km_z5-t1-t2',\n",
       " 'km_sprinting',\n",
       " 'strength_training',\n",
       " 'hours_alternative',\n",
       " 'perceived_exertion',\n",
       " 'perceived_trainingsuccess',\n",
       " 'perceived_recovery',\n",
       " 'nr__sessions_1',\n",
       " 'total_km_1',\n",
       " 'km_z3-4_1',\n",
       " 'km_z5-t1-t2_1',\n",
       " 'km_sprinting_1',\n",
       " 'strength_training_1',\n",
       " 'hours_alternative_1',\n",
       " 'perceived_exertion_1',\n",
       " 'perceived_trainingsuccess_1',\n",
       " 'perceived_recovery_1',\n",
       " 'nr__sessions_2',\n",
       " 'total_km_2',\n",
       " 'km_z3-4_2',\n",
       " 'km_z5-t1-t2_2',\n",
       " 'km_sprinting_2',\n",
       " 'strength_training_2',\n",
       " 'hours_alternative_2',\n",
       " 'perceived_exertion_2',\n",
       " 'perceived_trainingsuccess_2',\n",
       " 'perceived_recovery_2',\n",
       " 'nr__sessions_3',\n",
       " 'total_km_3',\n",
       " 'km_z3-4_3',\n",
       " 'km_z5-t1-t2_3',\n",
       " 'km_sprinting_3',\n",
       " 'strength_training_3',\n",
       " 'hours_alternative_3',\n",
       " 'perceived_exertion_3',\n",
       " 'perceived_trainingsuccess_3',\n",
       " 'perceived_recovery_3',\n",
       " 'nr__sessions_4',\n",
       " 'total_km_4',\n",
       " 'km_z3-4_4',\n",
       " 'km_z5-t1-t2_4',\n",
       " 'km_sprinting_4',\n",
       " 'strength_training_4',\n",
       " 'hours_alternative_4',\n",
       " 'perceived_exertion_4',\n",
       " 'perceived_trainingsuccess_4',\n",
       " 'perceived_recovery_4',\n",
       " 'nr__sessions_5',\n",
       " 'total_km_5',\n",
       " 'km_z3-4_5',\n",
       " 'km_z5-t1-t2_5',\n",
       " 'km_sprinting_5',\n",
       " 'strength_training_5',\n",
       " 'hours_alternative_5',\n",
       " 'perceived_exertion_5',\n",
       " 'perceived_trainingsuccess_5',\n",
       " 'perceived_recovery_5',\n",
       " 'nr__sessions_6',\n",
       " 'total_km_6',\n",
       " 'km_z3-4_6',\n",
       " 'km_z5-t1-t2_6',\n",
       " 'km_sprinting_6',\n",
       " 'strength_training_6',\n",
       " 'hours_alternative_6',\n",
       " 'perceived_exertion_6',\n",
       " 'perceived_trainingsuccess_6',\n",
       " 'perceived_recovery_6',\n",
       " 'date']"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-uyGkZF6cH4U"
   },
   "source": [
    "We now create a local estimator and start the training process"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vIRuF-xaqPta"
   },
   "source": [
    "In the deep learning model below, we included the h2o library for anomaly detection i.e autoencoder."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "EZvOc8B2i-bH"
   },
   "outputs": [],
   "source": [
    "# our model with its hyper parameters\n",
    "\n",
    "deepl_model = H2ODeepLearningEstimator(activation = \"Tanh\",\n",
    "                               hidden = [45,22,22,22,22,22,22,45],\n",
    "                               epochs = 100,\n",
    "                               standardize = True,\n",
    "                                stopping_metric = 'MSE',\n",
    "                                loss = 'automatic',\n",
    "                                train_samples_per_iteration = 32,\n",
    "                                shuffle_training_data = True,     \n",
    "                               autoencoder = True,\n",
    "                               l1 = 10e-5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2972793,
     "status": "ok",
     "timestamp": 1661549483129,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ysiog36JjAY3",
    "outputId": "e91cd1c6-966f-4036-e71c-800eba6cda44"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "deeplearning Model Build progress: |█████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |██████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |█████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |███████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |███████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |██████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |████████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |██████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |█████████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |█████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |████████████████████████████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |██████████████████████████████████| (done) 100%\n",
      "deeplearning Model Build progress: |█████████████████████████████████████████████████████████████| (done) 100%\n",
      "7min 15s ± 1min 8s per loop (mean ± std. dev. of 7 runs, 1 loop each)\n",
      "█████████████████████████████| (done) 100%\n",
      "7min 15s ± 1min 8s per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "%timeit deepl_model.train(x = cols , y = 'injury', training_frame = train_h2o) # this takes a while at least an hour "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 473,
     "status": "ok",
     "timestamp": 1661549953220,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "5Ken-3NPo4Dq",
    "outputId": "64826733-332a-43f4-e562-b57020411837"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "ModelMetricsAutoEncoder: deeplearning\n",
      "** Reported on train data. **\n",
      "\n",
      "MSE: 0.0039337076845405595\n",
      "RMSE: 0.06271927681774209\n"
     ]
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Evaluate performance:\n",
    "\n",
    "perf = deepl_model.model_performance()\n",
    "perf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ic2BV2d-qlE2"
   },
   "source": [
    "The MSE & RMSE helps us to understand how far off our prediction is on the training data as opposed to the actual results. The small number above which almost equals zero tells us that our model performs well on the training set.\n",
    "\n",
    "We also note that $RMSE = \\sqrt{MSE}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cxuafel6kRAt"
   },
   "source": [
    "Show the model feature importance in a dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "executionInfo": {
     "elapsed": 464,
     "status": "ok",
     "timestamp": 1661549956349,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "oA54wpE9jJDG",
    "outputId": "9abef579-3d10-475e-e3a6-2237898c04ba"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-0fd18775-9206-4557-b16e-d5f94f5afde5\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>variable</th>\n",
       "      <th>relative_importance</th>\n",
       "      <th>scaled_importance</th>\n",
       "      <th>percentage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>nr__sessions_1</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.052037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>nr__sessions_3</td>\n",
       "      <td>0.837007</td>\n",
       "      <td>0.837007</td>\n",
       "      <td>0.043555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>nr__sessions_2</td>\n",
       "      <td>0.832789</td>\n",
       "      <td>0.832789</td>\n",
       "      <td>0.043336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>nr__sessions_6</td>\n",
       "      <td>0.812425</td>\n",
       "      <td>0.812425</td>\n",
       "      <td>0.042276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nr__sessions_5</td>\n",
       "      <td>0.792033</td>\n",
       "      <td>0.792033</td>\n",
       "      <td>0.041215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>nr__sessions_4</td>\n",
       "      <td>0.774218</td>\n",
       "      <td>0.774218</td>\n",
       "      <td>0.040288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>strength_training</td>\n",
       "      <td>0.734084</td>\n",
       "      <td>0.734084</td>\n",
       "      <td>0.038200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>strength_training_1</td>\n",
       "      <td>0.694053</td>\n",
       "      <td>0.694053</td>\n",
       "      <td>0.036116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>nr__sessions</td>\n",
       "      <td>0.674160</td>\n",
       "      <td>0.674160</td>\n",
       "      <td>0.035081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>perceived_exertion_2</td>\n",
       "      <td>0.607244</td>\n",
       "      <td>0.607244</td>\n",
       "      <td>0.031599</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>strength_training_4</td>\n",
       "      <td>0.597771</td>\n",
       "      <td>0.597771</td>\n",
       "      <td>0.031106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>date</td>\n",
       "      <td>0.570068</td>\n",
       "      <td>0.570068</td>\n",
       "      <td>0.029665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>perceived_exertion_3</td>\n",
       "      <td>0.549938</td>\n",
       "      <td>0.549938</td>\n",
       "      <td>0.028617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>perceived_trainingsuccess_5</td>\n",
       "      <td>0.503045</td>\n",
       "      <td>0.503045</td>\n",
       "      <td>0.026177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>perceived_exertion_5</td>\n",
       "      <td>0.459329</td>\n",
       "      <td>0.459329</td>\n",
       "      <td>0.023902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>perceived_exertion_1</td>\n",
       "      <td>0.416877</td>\n",
       "      <td>0.416877</td>\n",
       "      <td>0.021693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>perceived_exertion_6</td>\n",
       "      <td>0.402688</td>\n",
       "      <td>0.402688</td>\n",
       "      <td>0.020955</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>strength_training_3</td>\n",
       "      <td>0.385178</td>\n",
       "      <td>0.385178</td>\n",
       "      <td>0.020044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>perceived_trainingsuccess_3</td>\n",
       "      <td>0.363605</td>\n",
       "      <td>0.363605</td>\n",
       "      <td>0.018921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>perceived_trainingsuccess_2</td>\n",
       "      <td>0.338509</td>\n",
       "      <td>0.338509</td>\n",
       "      <td>0.017615</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>perceived_exertion</td>\n",
       "      <td>0.335530</td>\n",
       "      <td>0.335530</td>\n",
       "      <td>0.017460</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>strength_training_6</td>\n",
       "      <td>0.320976</td>\n",
       "      <td>0.320976</td>\n",
       "      <td>0.016703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>perceived_trainingsuccess_6</td>\n",
       "      <td>0.302537</td>\n",
       "      <td>0.302537</td>\n",
       "      <td>0.015743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>perceived_recovery_5</td>\n",
       "      <td>0.293750</td>\n",
       "      <td>0.293750</td>\n",
       "      <td>0.015286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>perceived_trainingsuccess</td>\n",
       "      <td>0.290505</td>\n",
       "      <td>0.290505</td>\n",
       "      <td>0.015117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>perceived_recovery_3</td>\n",
       "      <td>0.285945</td>\n",
       "      <td>0.285945</td>\n",
       "      <td>0.014880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>perceived_recovery_4</td>\n",
       "      <td>0.284656</td>\n",
       "      <td>0.284656</td>\n",
       "      <td>0.014813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>perceived_trainingsuccess_1</td>\n",
       "      <td>0.280605</td>\n",
       "      <td>0.280605</td>\n",
       "      <td>0.014602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>perceived_recovery_6</td>\n",
       "      <td>0.265760</td>\n",
       "      <td>0.265760</td>\n",
       "      <td>0.013829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>perceived_recovery</td>\n",
       "      <td>0.263282</td>\n",
       "      <td>0.263282</td>\n",
       "      <td>0.013700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>strength_training_2</td>\n",
       "      <td>0.252451</td>\n",
       "      <td>0.252451</td>\n",
       "      <td>0.013137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>perceived_recovery_2</td>\n",
       "      <td>0.233755</td>\n",
       "      <td>0.233755</td>\n",
       "      <td>0.012164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>perceived_recovery_1</td>\n",
       "      <td>0.232987</td>\n",
       "      <td>0.232987</td>\n",
       "      <td>0.012124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>strength_training_5</td>\n",
       "      <td>0.231050</td>\n",
       "      <td>0.231050</td>\n",
       "      <td>0.012023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>perceived_trainingsuccess_4</td>\n",
       "      <td>0.219581</td>\n",
       "      <td>0.219581</td>\n",
       "      <td>0.011426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>perceived_exertion_4</td>\n",
       "      <td>0.218597</td>\n",
       "      <td>0.218597</td>\n",
       "      <td>0.011375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>total_km_3</td>\n",
       "      <td>0.187769</td>\n",
       "      <td>0.187769</td>\n",
       "      <td>0.009771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>total_km_2</td>\n",
       "      <td>0.184078</td>\n",
       "      <td>0.184078</td>\n",
       "      <td>0.009579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>total_km_5</td>\n",
       "      <td>0.166356</td>\n",
       "      <td>0.166356</td>\n",
       "      <td>0.008657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>total_km_4</td>\n",
       "      <td>0.154513</td>\n",
       "      <td>0.154513</td>\n",
       "      <td>0.008040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>total_km_1</td>\n",
       "      <td>0.150580</td>\n",
       "      <td>0.150580</td>\n",
       "      <td>0.007836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>total_km_6</td>\n",
       "      <td>0.150452</td>\n",
       "      <td>0.150452</td>\n",
       "      <td>0.007829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>total_km</td>\n",
       "      <td>0.140971</td>\n",
       "      <td>0.140971</td>\n",
       "      <td>0.007336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>km_z5-t1-t2_1</td>\n",
       "      <td>0.068250</td>\n",
       "      <td>0.068250</td>\n",
       "      <td>0.003552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>hours_alternative_3</td>\n",
       "      <td>0.067490</td>\n",
       "      <td>0.067490</td>\n",
       "      <td>0.003512</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>hours_alternative</td>\n",
       "      <td>0.067452</td>\n",
       "      <td>0.067452</td>\n",
       "      <td>0.003510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>km_z3-4_4</td>\n",
       "      <td>0.063417</td>\n",
       "      <td>0.063417</td>\n",
       "      <td>0.003300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>km_z3-4_3</td>\n",
       "      <td>0.061998</td>\n",
       "      <td>0.061998</td>\n",
       "      <td>0.003226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>hours_alternative_5</td>\n",
       "      <td>0.061292</td>\n",
       "      <td>0.061292</td>\n",
       "      <td>0.003189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>km_z3-4_2</td>\n",
       "      <td>0.061083</td>\n",
       "      <td>0.061083</td>\n",
       "      <td>0.003179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>km_z3-4_6</td>\n",
       "      <td>0.058273</td>\n",
       "      <td>0.058273</td>\n",
       "      <td>0.003032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>hours_alternative_2</td>\n",
       "      <td>0.057940</td>\n",
       "      <td>0.057940</td>\n",
       "      <td>0.003015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>hours_alternative_1</td>\n",
       "      <td>0.057845</td>\n",
       "      <td>0.057845</td>\n",
       "      <td>0.003010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>km_z3-4_1</td>\n",
       "      <td>0.056962</td>\n",
       "      <td>0.056962</td>\n",
       "      <td>0.002964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>hours_alternative_4</td>\n",
       "      <td>0.055539</td>\n",
       "      <td>0.055539</td>\n",
       "      <td>0.002890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>km_z5-t1-t2_5</td>\n",
       "      <td>0.053585</td>\n",
       "      <td>0.053585</td>\n",
       "      <td>0.002788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>km_z3-4</td>\n",
       "      <td>0.051384</td>\n",
       "      <td>0.051384</td>\n",
       "      <td>0.002674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>km_z5-t1-t2</td>\n",
       "      <td>0.050723</td>\n",
       "      <td>0.050723</td>\n",
       "      <td>0.002639</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>km_z3-4_5</td>\n",
       "      <td>0.050161</td>\n",
       "      <td>0.050161</td>\n",
       "      <td>0.002610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>injury</td>\n",
       "      <td>0.049861</td>\n",
       "      <td>0.049861</td>\n",
       "      <td>0.002595</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>km_z5-t1-t2_4</td>\n",
       "      <td>0.047651</td>\n",
       "      <td>0.047651</td>\n",
       "      <td>0.002480</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>km_z5-t1-t2_2</td>\n",
       "      <td>0.044871</td>\n",
       "      <td>0.044871</td>\n",
       "      <td>0.002335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>km_z5-t1-t2_6</td>\n",
       "      <td>0.043690</td>\n",
       "      <td>0.043690</td>\n",
       "      <td>0.002274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>km_z5-t1-t2_3</td>\n",
       "      <td>0.041144</td>\n",
       "      <td>0.041144</td>\n",
       "      <td>0.002141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>hours_alternative_6</td>\n",
       "      <td>0.036241</td>\n",
       "      <td>0.036241</td>\n",
       "      <td>0.001886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>km_sprinting_1</td>\n",
       "      <td>0.035335</td>\n",
       "      <td>0.035335</td>\n",
       "      <td>0.001839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>km_sprinting_6</td>\n",
       "      <td>0.033413</td>\n",
       "      <td>0.033413</td>\n",
       "      <td>0.001739</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>km_sprinting_4</td>\n",
       "      <td>0.033172</td>\n",
       "      <td>0.033172</td>\n",
       "      <td>0.001726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>km_sprinting</td>\n",
       "      <td>0.031600</td>\n",
       "      <td>0.031600</td>\n",
       "      <td>0.001644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>km_sprinting_5</td>\n",
       "      <td>0.029190</td>\n",
       "      <td>0.029190</td>\n",
       "      <td>0.001519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>km_sprinting_2</td>\n",
       "      <td>0.027998</td>\n",
       "      <td>0.027998</td>\n",
       "      <td>0.001457</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>km_sprinting_3</td>\n",
       "      <td>0.027810</td>\n",
       "      <td>0.027810</td>\n",
       "      <td>0.001447</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0fd18775-9206-4557-b16e-d5f94f5afde5')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-0fd18775-9206-4557-b16e-d5f94f5afde5 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-0fd18775-9206-4557-b16e-d5f94f5afde5');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                       variable  relative_importance  scaled_importance  \\\n",
       "0                nr__sessions_1             1.000000           1.000000   \n",
       "1                nr__sessions_3             0.837007           0.837007   \n",
       "2                nr__sessions_2             0.832789           0.832789   \n",
       "3                nr__sessions_6             0.812425           0.812425   \n",
       "4                nr__sessions_5             0.792033           0.792033   \n",
       "5                nr__sessions_4             0.774218           0.774218   \n",
       "6             strength_training             0.734084           0.734084   \n",
       "7           strength_training_1             0.694053           0.694053   \n",
       "8                  nr__sessions             0.674160           0.674160   \n",
       "9          perceived_exertion_2             0.607244           0.607244   \n",
       "10          strength_training_4             0.597771           0.597771   \n",
       "11                         date             0.570068           0.570068   \n",
       "12         perceived_exertion_3             0.549938           0.549938   \n",
       "13  perceived_trainingsuccess_5             0.503045           0.503045   \n",
       "14         perceived_exertion_5             0.459329           0.459329   \n",
       "15         perceived_exertion_1             0.416877           0.416877   \n",
       "16         perceived_exertion_6             0.402688           0.402688   \n",
       "17          strength_training_3             0.385178           0.385178   \n",
       "18  perceived_trainingsuccess_3             0.363605           0.363605   \n",
       "19  perceived_trainingsuccess_2             0.338509           0.338509   \n",
       "20           perceived_exertion             0.335530           0.335530   \n",
       "21          strength_training_6             0.320976           0.320976   \n",
       "22  perceived_trainingsuccess_6             0.302537           0.302537   \n",
       "23         perceived_recovery_5             0.293750           0.293750   \n",
       "24    perceived_trainingsuccess             0.290505           0.290505   \n",
       "25         perceived_recovery_3             0.285945           0.285945   \n",
       "26         perceived_recovery_4             0.284656           0.284656   \n",
       "27  perceived_trainingsuccess_1             0.280605           0.280605   \n",
       "28         perceived_recovery_6             0.265760           0.265760   \n",
       "29           perceived_recovery             0.263282           0.263282   \n",
       "30          strength_training_2             0.252451           0.252451   \n",
       "31         perceived_recovery_2             0.233755           0.233755   \n",
       "32         perceived_recovery_1             0.232987           0.232987   \n",
       "33          strength_training_5             0.231050           0.231050   \n",
       "34  perceived_trainingsuccess_4             0.219581           0.219581   \n",
       "35         perceived_exertion_4             0.218597           0.218597   \n",
       "36                   total_km_3             0.187769           0.187769   \n",
       "37                   total_km_2             0.184078           0.184078   \n",
       "38                   total_km_5             0.166356           0.166356   \n",
       "39                   total_km_4             0.154513           0.154513   \n",
       "40                   total_km_1             0.150580           0.150580   \n",
       "41                   total_km_6             0.150452           0.150452   \n",
       "42                     total_km             0.140971           0.140971   \n",
       "43                km_z5-t1-t2_1             0.068250           0.068250   \n",
       "44          hours_alternative_3             0.067490           0.067490   \n",
       "45            hours_alternative             0.067452           0.067452   \n",
       "46                    km_z3-4_4             0.063417           0.063417   \n",
       "47                    km_z3-4_3             0.061998           0.061998   \n",
       "48          hours_alternative_5             0.061292           0.061292   \n",
       "49                    km_z3-4_2             0.061083           0.061083   \n",
       "50                    km_z3-4_6             0.058273           0.058273   \n",
       "51          hours_alternative_2             0.057940           0.057940   \n",
       "52          hours_alternative_1             0.057845           0.057845   \n",
       "53                    km_z3-4_1             0.056962           0.056962   \n",
       "54          hours_alternative_4             0.055539           0.055539   \n",
       "55                km_z5-t1-t2_5             0.053585           0.053585   \n",
       "56                      km_z3-4             0.051384           0.051384   \n",
       "57                  km_z5-t1-t2             0.050723           0.050723   \n",
       "58                    km_z3-4_5             0.050161           0.050161   \n",
       "59                       injury             0.049861           0.049861   \n",
       "60                km_z5-t1-t2_4             0.047651           0.047651   \n",
       "61                km_z5-t1-t2_2             0.044871           0.044871   \n",
       "62                km_z5-t1-t2_6             0.043690           0.043690   \n",
       "63                km_z5-t1-t2_3             0.041144           0.041144   \n",
       "64          hours_alternative_6             0.036241           0.036241   \n",
       "65               km_sprinting_1             0.035335           0.035335   \n",
       "66               km_sprinting_6             0.033413           0.033413   \n",
       "67               km_sprinting_4             0.033172           0.033172   \n",
       "68                 km_sprinting             0.031600           0.031600   \n",
       "69               km_sprinting_5             0.029190           0.029190   \n",
       "70               km_sprinting_2             0.027998           0.027998   \n",
       "71               km_sprinting_3             0.027810           0.027810   \n",
       "\n",
       "    percentage  \n",
       "0     0.052037  \n",
       "1     0.043555  \n",
       "2     0.043336  \n",
       "3     0.042276  \n",
       "4     0.041215  \n",
       "5     0.040288  \n",
       "6     0.038200  \n",
       "7     0.036116  \n",
       "8     0.035081  \n",
       "9     0.031599  \n",
       "10    0.031106  \n",
       "11    0.029665  \n",
       "12    0.028617  \n",
       "13    0.026177  \n",
       "14    0.023902  \n",
       "15    0.021693  \n",
       "16    0.020955  \n",
       "17    0.020044  \n",
       "18    0.018921  \n",
       "19    0.017615  \n",
       "20    0.017460  \n",
       "21    0.016703  \n",
       "22    0.015743  \n",
       "23    0.015286  \n",
       "24    0.015117  \n",
       "25    0.014880  \n",
       "26    0.014813  \n",
       "27    0.014602  \n",
       "28    0.013829  \n",
       "29    0.013700  \n",
       "30    0.013137  \n",
       "31    0.012164  \n",
       "32    0.012124  \n",
       "33    0.012023  \n",
       "34    0.011426  \n",
       "35    0.011375  \n",
       "36    0.009771  \n",
       "37    0.009579  \n",
       "38    0.008657  \n",
       "39    0.008040  \n",
       "40    0.007836  \n",
       "41    0.007829  \n",
       "42    0.007336  \n",
       "43    0.003552  \n",
       "44    0.003512  \n",
       "45    0.003510  \n",
       "46    0.003300  \n",
       "47    0.003226  \n",
       "48    0.003189  \n",
       "49    0.003179  \n",
       "50    0.003032  \n",
       "51    0.003015  \n",
       "52    0.003010  \n",
       "53    0.002964  \n",
       "54    0.002890  \n",
       "55    0.002788  \n",
       "56    0.002674  \n",
       "57    0.002639  \n",
       "58    0.002610  \n",
       "59    0.002595  \n",
       "60    0.002480  \n",
       "61    0.002335  \n",
       "62    0.002274  \n",
       "63    0.002141  \n",
       "64    0.001886  \n",
       "65    0.001839  \n",
       "66    0.001739  \n",
       "67    0.001726  \n",
       "68    0.001644  \n",
       "69    0.001519  \n",
       "70    0.001457  \n",
       "71    0.001447  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deepl_model._model_json['output']['variable_importances'].as_data_frame()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0ctJxD89kgAw"
   },
   "source": [
    "Visualization of the feature importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 513
    },
    "executionInfo": {
     "elapsed": 851,
     "status": "ok",
     "timestamp": 1661549965239,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "73vZUHmmkfhI",
    "outputId": "1f6f7393-0d14-49a9-b3f9-29cda6aba5a5"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting the variable importance\n",
    "rcParams['figure.figsize'] = 14, 8\n",
    "#plt.rcdefaults()\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "variables = deepl_model._model_json['output']['variable_importances']['variable']\n",
    "var = variables[0:35]\n",
    "y_pos = np.arange(len(var))\n",
    "\n",
    "scaled_importance = deepl_model._model_json['output']['variable_importances']['scaled_importance']\n",
    "sc = scaled_importance[0:35]\n",
    "\n",
    "ax.barh(y_pos, sc, align='center', color='green', ecolor='black')\n",
    "ax.set_yticks(y_pos)\n",
    "ax.set_yticklabels(variables)\n",
    "ax.invert_yaxis()\n",
    "ax.set_xlabel('Scaled Importance')\n",
    "ax.set_title('Variable Importance of Top 35 Features ')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ir90Ed2_krem"
   },
   "source": [
    "Visualize the model loss."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 531
    },
    "executionInfo": {
     "elapsed": 558,
     "status": "ok",
     "timestamp": 1661549977060,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "G7ZoE668kfdh",
    "outputId": "186e2a3e-d233-488b-a088-02f8ca814870"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'epoch')"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting the loss\n",
    "\n",
    "scoring_history = deepl_model.score_history()\n",
    "\n",
    "rcParams['figure.figsize'] = 14, 8\n",
    "plt.plot(scoring_history['training_mse'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2SFWIkEUkyrQ"
   },
   "source": [
    "Visualize the scatter plot across the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 497
    },
    "executionInfo": {
     "elapsed": 1284,
     "status": "ok",
     "timestamp": 1661550047053,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "BQFsmBYqkedk",
    "outputId": "45a6bc5f-7683-40d8-f615-3d58497b7a7b"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_rec_error = deepl_model.anomaly(test_h2o) \n",
    "\n",
    "# deepl is a H2O function which calculates the error for the dataset\n",
    "\n",
    "# converting to pandas dataframe\n",
    "test_rec_error_df = test_rec_error.as_data_frame()\n",
    "\n",
    "# plotting the testing dataset against the error\n",
    "test_rec_error_df['id'] = test_rec_error_df.index\n",
    "rcParams['figure.figsize'] = 14, 8\n",
    "test_rec_error_df.plot(kind=\"scatter\", x='id', y=\"Reconstruction.MSE\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 162
    },
    "executionInfo": {
     "elapsed": 923,
     "status": "ok",
     "timestamp": 1661550105410,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "bh5vqpwrk-6f",
    "outputId": "025423f6-730d-45cb-f405-43343b980dee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "deeplearning prediction progress: |██████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-b6ac792b-e3f7-413c-b6c9-4b717fa44933\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>reconstruction_error</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true_class</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>8602</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>110</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b6ac792b-e3f7-413c-b6c9-4b717fa44933')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-b6ac792b-e3f7-413c-b6c9-4b717fa44933 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-b6ac792b-e3f7-413c-b6c9-4b717fa44933');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "            reconstruction_error\n",
       "true_class                      \n",
       "0                           8602\n",
       "1                            110"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# predicting the class for the testing dataset\n",
    "predictions = deepl_model.predict(test_h2o)\n",
    "error_df = pd.DataFrame({'reconstruction_error': test_rec_error_df['Reconstruction.MSE'],\n",
    "                        'true_class': Y_test_df})\n",
    "error_df.groupby('true_class').count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1w7rTflOmqaR"
   },
   "source": [
    "Calculate the threshold score/ cutoff score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1053,
     "status": "ok",
     "timestamp": 1661550124839,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Eh8uvN88lJCG",
    "outputId": "7e0c71d8-88a3-4e6c-cf73-14ed3e3d65db"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.016328085790992836"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_mean = error_df['reconstruction_error'].mean()\n",
    "score_std = error_df['reconstruction_error'].std()\n",
    "score_cutoff = score_mean + 4*score_std\n",
    "score_cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 923,
     "status": "ok",
     "timestamp": 1661550129468,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "QXGWE7GEmuZV",
    "outputId": "52ab71a1-a420-40b3-df20-85a100072836"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "reconstruction_error    110\n",
       "true_class              110\n",
       "dtype: int64"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_df[(error_df['true_class'] == 1) & error_df['reconstruction_error'] > 0.0095].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "R3ostMdJm446"
   },
   "source": [
    "Print out the confusion matrix and adjust and tune the confusion matrix from the threshold."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ysr4PhagVgjL"
   },
   "outputs": [],
   "source": [
    "# len(y_pred), len(error_df.true_class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 49
    },
    "executionInfo": {
     "elapsed": 420,
     "status": "ok",
     "timestamp": 1661550139734,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Qv9BWEtgXSA4",
    "outputId": "212b4761-b96d-4326-f3f7-bea7151125ab"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-ac891abf-1ec9-4ff8-865e-337d6443c614\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>reconstruction_error</th>\n",
       "      <th>true_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ac891abf-1ec9-4ff8-865e-337d6443c614')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-ac891abf-1ec9-4ff8-865e-337d6443c614 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-ac891abf-1ec9-4ff8-865e-337d6443c614');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [reconstruction_error, true_class]\n",
       "Index: []"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "error_df[error_df['true_class'] == 'NaN' ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 730
    },
    "executionInfo": {
     "elapsed": 957,
     "status": "ok",
     "timestamp": 1661550570050,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "9zhQqaJ5mw6h",
    "outputId": "33383ec3-8f68-49da-f984-4625be177fd9"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = 0.009 # we can equally choose to use different values for our threshold [0.009]\n",
    "LABELS = ['Injury', 'No Injury']\n",
    "y_pred = [1 if e > threshold else 0 for e in error_df.reconstruction_error.values]\n",
    "conf_matrix = confusion_matrix(error_df.true_class, y_pred)\n",
    "plt.figure(figsize=(12, 12))\n",
    "sns.heatmap(conf_matrix, xticklabels=LABELS, yticklabels=LABELS, annot=True, fmt=\"d\");\n",
    "plt.title(\"Confusion matrix\")\n",
    "plt.ylabel('True class')\n",
    "plt.xlabel('Predicted class')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "E7WigCeargv-"
   },
   "source": [
    "The confusion matrix above helps us to understand our accuracy on the test data. \n",
    "\n",
    "We note that:\n",
    "- 8099: The model predicts this case to be actually injury where the true case is injury too.\n",
    "- 10: The model predicts no-injury with the true case being no injury.\n",
    "- 100: The model predicts injury where the actual class in no-injury\n",
    "- 503: The model predicts no-injury where the actual class is injury.\n",
    "\n",
    "The above explanation is given in the classification report below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 604,
     "status": "ok",
     "timestamp": 1661550572900,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "1r58H_V9HjyY",
    "outputId": "de25a40b-0a02-41d3-f297-c44167817cd1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.99      0.94      0.96      8602\n",
      "           1       0.02      0.09      0.03       110\n",
      "\n",
      "    accuracy                           0.93      8712\n",
      "   macro avg       0.50      0.52      0.50      8712\n",
      "weighted avg       0.98      0.93      0.95      8712\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(error_df.true_class, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MNXhdwY0ZtSb"
   },
   "source": [
    "# Model 2: Isolation Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "executionInfo": {
     "elapsed": 292,
     "status": "ok",
     "timestamp": 1662705368957,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ZPYa1f_8B7hk"
   },
   "outputs": [],
   "source": [
    "def drop_a_column(df, cols):\n",
    "    \"\"\"\n",
    "    This functiion drops a column in a dataframe and returns the new df without\n",
    "    the dropped column.\n",
    "\n",
    "    cols is a list\n",
    "    df is a dataframe\n",
    "    \"\"\"\n",
    "\n",
    "    df_dr = df.drop(cols, axis = 1)\n",
    "\n",
    "    return df_dr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1662705370145,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "I4VBDJdW5-SS"
   },
   "outputs": [],
   "source": [
    "# inliers = df_day[df_day.injury == 'no']\n",
    "inliers = df_day[df_day.injury == 0]\n",
    "inliers_y = inliers.injury\n",
    "inliers = inliers.drop(['injury', 'athlete_id'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662705371351,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "o7cvpMDt6nPF"
   },
   "outputs": [],
   "source": [
    "# outliers = df_day[df_day.injury == 'yes']\n",
    "outliers = df_day[df_day.injury == 1]\n",
    "outliers_y = outliers.injury\n",
    "outliers = outliers.drop(['injury', 'athlete_id'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 211
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1662705372845,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "bnIQ2LkvJe38",
    "outputId": "0aff445a-e326-4ee5-eb42-9e5e22bfbdd5"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5.8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.6</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.17</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.08</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.18</td>\n",
       "      <td>1</td>\n",
       "      <td>16.4</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.15</td>\n",
       "      <td>1</td>\n",
       "      <td>5.2</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.5</td>\n",
       "      <td>1.2</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  nr_sessions  total_km  km_z3-4  km_z5-t1-t2  km_sprinting strength_training  \\\n",
       "0           1       5.8      0.0          0.6           1.2                 0   \n",
       "1           0       0.0      0.0          0.0           0.0                 0   \n",
       "\n",
       "   hours_alternative  perceived_exertion  perceived_trainingsuccess  \\\n",
       "0                0.0                0.11                       0.00   \n",
       "1                0.0               -0.01                      -0.01   \n",
       "\n",
       "   perceived_recovery nr_sessions_1  total_km_1  km_z3-4_1  km_z5-t1-t2_1  \\\n",
       "0                0.18             0         0.0        0.0            0.0   \n",
       "1               -0.01             1         0.0        0.0            0.0   \n",
       "\n",
       "   km_sprinting_1 strength_training_1  hours_alternative_1  \\\n",
       "0             0.0                   0                  0.0   \n",
       "1             0.0                   1                  0.0   \n",
       "\n",
       "   perceived_exertion_1  perceived_trainingsuccess_1  perceived_recovery_1  \\\n",
       "0                 -0.01                        -0.01                 -0.01   \n",
       "1                  0.10                         0.00                  0.17   \n",
       "\n",
       "  nr_sessions_2  total_km_2  km_z3-4_2  km_z5-t1-t2_2  km_sprinting_2  \\\n",
       "0             1         0.0        0.0            0.0             0.0   \n",
       "1             0         0.0        0.0            0.0             0.0   \n",
       "\n",
       "  strength_training_2  hours_alternative_2  perceived_exertion_2  \\\n",
       "0                   1                  0.0                  0.10   \n",
       "1                   0                  0.0                 -0.01   \n",
       "\n",
       "   perceived_trainingsuccess_2  perceived_recovery_2 nr_sessions_3  \\\n",
       "0                         0.00                  0.17             0   \n",
       "1                        -0.01                 -0.01             1   \n",
       "\n",
       "   total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3 strength_training_3  \\\n",
       "0         0.0        0.0            0.0             0.0                   0   \n",
       "1         0.0        0.0            0.0             0.0                   0   \n",
       "\n",
       "   hours_alternative_3  perceived_exertion_3  perceived_trainingsuccess_3  \\\n",
       "0                 0.00                 -0.01                        -0.01   \n",
       "1                 1.08                  0.08                         0.00   \n",
       "\n",
       "   perceived_recovery_3 nr_sessions_4  total_km_4  km_z3-4_4  km_z5-t1-t2_4  \\\n",
       "0                 -0.01             1         0.0        0.0            0.0   \n",
       "1                  0.18             1        16.4       10.0            0.0   \n",
       "\n",
       "   km_sprinting_4 strength_training_4  hours_alternative_4  \\\n",
       "0             0.0                   0                 1.08   \n",
       "1             0.0                   1                 0.00   \n",
       "\n",
       "   perceived_exertion_4  perceived_trainingsuccess_4  perceived_recovery_4  \\\n",
       "0                  0.08                          0.0                  0.18   \n",
       "1                  0.11                          0.0                  0.17   \n",
       "\n",
       "  nr_sessions_5  total_km_5  km_z3-4_5  km_z5-t1-t2_5  km_sprinting_5  \\\n",
       "0             1        16.4       10.0            0.0             0.0   \n",
       "1             1         0.0        0.0            0.0             0.0   \n",
       "\n",
       "  strength_training_5  hours_alternative_5  perceived_exertion_5  \\\n",
       "0                   1                  0.0                  0.11   \n",
       "1                   0                  1.0                  0.10   \n",
       "\n",
       "   perceived_trainingsuccess_5  perceived_recovery_5 nr_sessions_6  \\\n",
       "0                          0.0                  0.17             1   \n",
       "1                          0.0                  0.15             1   \n",
       "\n",
       "   total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6 strength_training_6  \\\n",
       "0         0.0        0.0            0.0             0.0                   0   \n",
       "1         5.2        0.0            0.5             1.2                   0   \n",
       "\n",
       "   hours_alternative_6  perceived_exertion_6  perceived_trainingsuccess_6  \\\n",
       "0                  1.0                   0.1                          0.0   \n",
       "1                  0.0                   0.1                          0.0   \n",
       "\n",
       "   perceived_recovery_6 date  \n",
       "0                  0.15    0  \n",
       "1                  0.17    1  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 318,
     "status": "ok",
     "timestamp": 1662705378663,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Ga9WMzXt89Y3",
    "outputId": "dbd4712d-ffb2-4611-fe09-1085cc70d206"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((42183, 71), (583, 71))"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.shape, outliers.shape # this is the respective dimension of our no-injury and injury data in our dataset respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {
    "executionInfo": {
     "elapsed": 305,
     "status": "ok",
     "timestamp": 1662705380062,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "56HuFunU66Lg"
   },
   "outputs": [],
   "source": [
    "inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "hzQwUMU57MB6"
   },
   "source": [
    "## Train the isolation Forest Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "executionInfo": {
     "elapsed": 471,
     "status": "ok",
     "timestamp": 1662705382282,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "YkeH4DA37UjD"
   },
   "outputs": [],
   "source": [
    "rng = np.random.RandomState(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1662704175406,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "vNrHZsT3DPmG"
   },
   "outputs": [],
   "source": [
    "# inliers = df_day[df_day.injury == 'no']\n",
    "# inliers = df_day[df_day.injury == 1]\n",
    "# inliers_y = inliers.injury\n",
    "# inliers = inliers.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "executionInfo": {
     "elapsed": 311,
     "status": "ok",
     "timestamp": 1662705384534,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "B7Tq24fuDai1"
   },
   "outputs": [],
   "source": [
    "inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "executionInfo": {
     "elapsed": 334,
     "status": "ok",
     "timestamp": 1662704553965,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "cAAUapyNDb7Y"
   },
   "outputs": [],
   "source": [
    "# inliers_tr_y = inliers_train.injury\n",
    "# inliers_train = drop_a_column(inliers_train, ['injury', 'athlete_id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1662704555307,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "FL7pBN7KKX7z"
   },
   "outputs": [],
   "source": [
    "# inliers_test = drop_a_column(inliers_test, ['injury', 'athlete_id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "executionInfo": {
     "elapsed": 1302,
     "status": "ok",
     "timestamp": 1662705392224,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "DWwb9Dev66B-"
   },
   "outputs": [],
   "source": [
    "model_iso = IsolationForest( max_samples=150, random_state=rng, contamination='auto')\n",
    "# model_iso = IsolationForest( max_samples=150, random_state=rng, contamination=0.01)\n",
    "model_iso.fit(inliers_train)\n",
    "inlier_pred_test = model_iso.predict(inliers_test)\n",
    "outlier_pred = model_iso.predict(outliers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 313,
     "status": "ok",
     "timestamp": 1662705396129,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "_7qKk8wi7eHF",
    "outputId": "5048b19b-dee9-44ba-a50e-d23a6b73a5c4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy in Detecting injury Cases: 0.9615977243095887\n",
      "Accuracy in Detecting no-injury Cases: 0.0823327615780446\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy in Detecting injury Cases:\", list(inlier_pred_test).count(1)/inlier_pred_test.shape[0])\n",
    "print(\"Accuracy in Detecting no-injury Cases:\", list(outlier_pred).count(-1)/outlier_pred.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 331,
     "status": "ok",
     "timestamp": 1662705400078,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "jfwEOG8m7AHu",
    "outputId": "0314db4d-644c-4c64-aeda-9486f3b5d334"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1,  1])"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(outlier_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662705400976,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "eGqbEW5Q6qlA",
    "outputId": "038bbf07-db59-487d-e2bd-e5433157d762"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.152\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "\n",
    "# mark inliers 1, outliers -1\n",
    "\n",
    "outliers_y = outliers_y.replace(1, -1)\n",
    "\n",
    "outliers_y = outliers_y.replace(0, 1)\n",
    "\n",
    "score = f1_score(outliers_y, outlier_pred, pos_label=-1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RFEBUpz2ANF6"
   },
   "source": [
    "We can then compare the predictions from the model to the expected target values and calculate a score. Given that we have crisp class labels, we might use a score like precision, recall, or a combination of both, such as the F-measure (F1-score).\n",
    "\n",
    "In this case, we will use F-measure score, which is the harmonic mean of precision and recall. We can calculate the F-measure using the f1_score() function and specify the label of the minority class as -1 via the “pos_label” argument."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "66hA5N6humEZ"
   },
   "source": [
    "# Model 3: One Class SVM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wVp9xkgfXGfS"
   },
   "source": [
    "One Class Classification (OCC) aims to differentiate samples of one particular class by learning from single class samples during training. It is one of the most commonly used approaches to solve Anamoly Detection (AD), a subfield of machine learning that deals with identifying anomalous cases which the model has never seen before. OCC is also called unary classification, class-modelling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "executionInfo": {
     "elapsed": 307,
     "status": "ok",
     "timestamp": 1662700785733,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Ap3-8zUrZiGm"
   },
   "outputs": [],
   "source": [
    "X = df_day.drop(columns = ['injury', 'athlete_id'])\n",
    "y = df_day['injury']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1662700786809,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "n69G-RJ7ZrtH"
   },
   "outputs": [],
   "source": [
    "# split into train/test sets\n",
    "trainX, testX, trainy, testy = train_test_split(X, y, test_size=0.80, random_state=2, stratify=y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kJ9kaP2hdjUT"
   },
   "source": [
    "<!-- train = trainX.copy()\n",
    "train['injury'] = trainy\n",
    "# train -->"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xP3naJcadIGl"
   },
   "source": [
    "<!-- # standardize the dataset\n",
    "scaler = StandardScaler()\n",
    "\n",
    "# Fit on training set only.\n",
    "scaler.fit(trainX)\n",
    "\n",
    "# Apply transform to both the training set and the test set.\n",
    "trainX = scaler.transform(trainX)\n",
    "testX = scaler.transform(testX) -->\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "executionInfo": {
     "elapsed": 403,
     "status": "ok",
     "timestamp": 1662701473047,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "Kj1I0Rem3loL"
   },
   "outputs": [],
   "source": [
    "model_onec_svm = OneClassSVM() #one class svm model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1662701474659,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "At88wiRHZ2BZ"
   },
   "outputs": [],
   "source": [
    "# kernel = ['linear', 'poly', 'rbf', 'sigmoid', 'precomputed']\n",
    "kernel = ['linear', 'poly', 'rbf', 'sigmoid']\n",
    "degree = np.arange(2,5)\n",
    "gamma = ['scale', 'auto']\n",
    "shrinking = [True, False]\n",
    "\n",
    "tune_parameters = [{\"kernel\": kernel, 'degree':degree, 'gamma':gamma, 'shrinking':shrinking}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662701477906,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "0swzQvyKcaDE",
    "outputId": "215b9c5f-4c13-42bf-dfcd-97b1920e5015"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'kernel': ['linear', 'poly', 'rbf', 'sigmoid'],\n",
       "  'degree': array([2, 3, 4]),\n",
       "  'gamma': ['scale', 'auto'],\n",
       "  'shrinking': [True, False]}]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tune_parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "executionInfo": {
     "elapsed": 309,
     "status": "ok",
     "timestamp": 1662701481928,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "fGRFjLbZ3p60"
   },
   "outputs": [],
   "source": [
    "# Instantiate the grid search model\n",
    "grid_search = GridSearchCV(estimator = model_onec_svm,\n",
    "                           param_grid = tune_parameters,\n",
    "                           cv = 10,\n",
    "                           n_jobs=-1, verbose=1, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "executionInfo": {
     "elapsed": 454,
     "status": "ok",
     "timestamp": 1662701485485,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "JweW30iWfwcv"
   },
   "outputs": [],
   "source": [
    "# we fit on majority class\n",
    "trainX_copy = trainX.copy()\n",
    "trainX_copy['injury'] = trainy\n",
    "trainX_copy = trainX_copy[trainX_copy.injury == 0]\n",
    "\n",
    "trainy_copy = trainy[trainy ==0]\n",
    "trainX_copy = trainX_copy.drop(columns = 'injury')\n",
    "# trainy_copy = "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 384,
     "status": "ok",
     "timestamp": 1662701495705,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "h_H4OHsztflE",
    "outputId": "689f597b-c9da-43e9-ea5c-1e17b8257bfa"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((8436,), (8436, 71))"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainy_copy.shape, trainX_copy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 111252,
     "status": "ok",
     "timestamp": 1662703380413,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "SCp87OvPcC6s",
    "outputId": "082d0902-4565-424d-e7b4-0320c8b274ea"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 48 candidates, totalling 480 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_search.py:972: UserWarning: One or more of the test scores are non-finite: [nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan\n",
      " nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan nan\n",
      " nan nan nan nan nan nan nan nan nan nan nan nan]\n",
      "  category=UserWarning,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7.42 s, sys: 312 ms, total: 7.73 s\n",
      "Wall time: 12min 50s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=10, estimator=OneClassSVM(), n_jobs=-1,\n",
       "             param_grid=[{'degree': array([2, 3, 4]),\n",
       "                          'gamma': ['scale', 'auto'],\n",
       "                          'kernel': ['linear', 'poly', 'rbf', 'sigmoid'],\n",
       "                          'shrinking': [True, False]}],\n",
       "             scoring='accuracy', verbose=1)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time  # hyper parameter tuning to avoid over fitting\n",
    "grid_search.fit(trainX_copy, trainy_copy)\n",
    "# grid_search.fit(trainX_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 890,
     "status": "ok",
     "timestamp": 1662703388978,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ntSV7X71cHQ0",
    "outputId": "81eaf22c-21b6-4577-a8cc-2106f1f2b5eb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(nan, OneClassSVM(degree=2, kernel='linear'))"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_score_, grid_search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "10rqKYqcv09R"
   },
   "outputs": [],
   "source": [
    "oc_svm_best = grid_search_bg.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "executionInfo": {
     "elapsed": 305,
     "status": "ok",
     "timestamp": 1662703540106,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "CDtazhBR2Nnj"
   },
   "outputs": [],
   "source": [
    "# define outlier detection model\n",
    "model = OneClassSVM(gamma='scale', nu=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 382,
     "status": "ok",
     "timestamp": 1662703554499,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "xKg76-B92bbV",
    "outputId": "4c731607-88ac-4d9e-ea6d-d2ab82962442"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OneClassSVM(nu=0.01)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit on majority class\n",
    "\n",
    "model.fit(trainX_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {
    "executionInfo": {
     "elapsed": 847,
     "status": "ok",
     "timestamp": 1662703578422,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "XwJVzYjw2u0w"
   },
   "outputs": [],
   "source": [
    "# detect outliers in the test set\n",
    "yhat = model.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "executionInfo": {
     "elapsed": 267,
     "status": "ok",
     "timestamp": 1662703843149,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "edst2O5T2zOp"
   },
   "outputs": [],
   "source": [
    "# mark inliers 1, outliers -1\n",
    "\n",
    "testy = testy.replace(1, -1)\n",
    "\n",
    "testy = testy.replace(0, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 446,
     "status": "ok",
     "timestamp": 1662703846632,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "PBWkZVnF2045",
    "outputId": "2c457d10-1c27-47dc-80be-1205dcfb5dee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.019\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "score = f1_score(testy, yhat, pos_label= -1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "yiZHyrZb8jKp"
   },
   "source": [
    "# A -  Repeating Models with Standardized data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "executionInfo": {
     "elapsed": 299,
     "status": "ok",
     "timestamp": 1662732551796,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "eoJ39hYs3SLG"
   },
   "outputs": [],
   "source": [
    "df_day2 = df_day.drop(columns = 'athlete_id')\n",
    "\n",
    "std_scaler = StandardScaler()\n",
    "\n",
    "df_day_scaled = std_scaler.fit_transform(df_day2.drop(columns = 'injury').to_numpy())\n",
    "df_day_scaled = pd.DataFrame(df_day_scaled, columns=list(df_day2.drop(columns = 'injury')))\n",
    "df_day_scaled['injury'] = df_day.injury"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "C-Rza9gW-weF"
   },
   "source": [
    "## Model 2: Isolation Forest II"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "executionInfo": {
     "elapsed": 279,
     "status": "ok",
     "timestamp": 1662705839095,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "8QcX_fhm-weG"
   },
   "outputs": [],
   "source": [
    "def drop_a_column(df, cols):\n",
    "    \"\"\"\n",
    "    This functiion drops a column in a dataframe and returns the new df without\n",
    "    the dropped column.\n",
    "\n",
    "    cols is a list\n",
    "    df is a dataframe\n",
    "    \"\"\"\n",
    "\n",
    "    df_dr = df.drop(cols, axis = 1)\n",
    "\n",
    "    return df_dr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "executionInfo": {
     "elapsed": 311,
     "status": "ok",
     "timestamp": 1662705849287,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "BiOrbWwx-weG"
   },
   "outputs": [],
   "source": [
    "# inliers = df_day_scaled[df_day_scaled.injury == 'no']\n",
    "inliers = df_day_scaled[df_day_scaled.injury == 0] #  no injury data\n",
    "inliers_y = inliers.injury #  no injury reponse data\n",
    "inliers = inliers.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "executionInfo": {
     "elapsed": 293,
     "status": "ok",
     "timestamp": 1662705857779,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "BH4eQBtB-weG"
   },
   "outputs": [],
   "source": [
    "# outliers = df_day_scaled[df_day_scaled.injury == 'yes']\n",
    "outliers = df_day_scaled[df_day_scaled.injury == 1] #  injury data\n",
    "outliers_y = outliers.injury #  injury reponse data\n",
    "outliers = outliers.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 193
    },
    "executionInfo": {
     "elapsed": 291,
     "status": "ok",
     "timestamp": 1662705859794,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "9Mx3h7u9-weG",
    "outputId": "1ea2a73b-989e-4152-a3d8-20618813bc2c"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>-0.165685</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>0.011076</td>\n",
       "      <td>2.331009</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-0.535600</td>\n",
       "      <td>-0.949786</td>\n",
       "      <td>-0.083533</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>2.704305</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-0.574482</td>\n",
       "      <td>-0.949257</td>\n",
       "      <td>-0.137026</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>1.677385</td>\n",
       "      <td>-0.652858</td>\n",
       "      <td>-0.950227</td>\n",
       "      <td>-0.085303</td>\n",
       "      <td>0.29539</td>\n",
       "      <td>1.250975</td>\n",
       "      <td>3.996750</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>2.714195</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.535897</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.137909</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.29909</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>1.512013</td>\n",
       "      <td>-0.574761</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.242559</td>\n",
       "      <td>-1.521713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.428581</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-1.002056</td>\n",
       "      <td>-0.976938</td>\n",
       "      <td>-1.081857</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>2.698331</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-0.573368</td>\n",
       "      <td>-0.948660</td>\n",
       "      <td>-0.135464</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>1.670717</td>\n",
       "      <td>-0.652204</td>\n",
       "      <td>-0.949629</td>\n",
       "      <td>-0.084824</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>1.254004</td>\n",
       "      <td>4.002433</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>2.708791</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-0.536027</td>\n",
       "      <td>-0.950227</td>\n",
       "      <td>-0.137798</td>\n",
       "      <td>0.29539</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>1.540748</td>\n",
       "      <td>-0.574816</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.242790</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.247831</td>\n",
       "      <td>-0.29909</td>\n",
       "      <td>-0.044280</td>\n",
       "      <td>2.330865</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-0.574761</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.137609</td>\n",
       "      <td>-1.520473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   nr_sessions  total_km   km_z3-4  km_z5-t1-t2  km_sprinting  \\\n",
       "0     0.293512 -0.165685 -0.298314     0.011076      2.331009   \n",
       "1    -1.428581 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "\n",
       "   strength_training  hours_alternative  perceived_exertion  \\\n",
       "0          -0.356549          -0.297443           -0.535600   \n",
       "1          -0.356549          -0.297443           -1.002056   \n",
       "\n",
       "   perceived_trainingsuccess  perceived_recovery  nr_sessions_1  total_km_1  \\\n",
       "0                  -0.949786           -0.083533      -1.426339   -0.939949   \n",
       "1                  -0.976938           -1.081857       0.294215   -0.939949   \n",
       "\n",
       "   km_z3-4_1  km_z5-t1-t2_1  km_sprinting_1  strength_training_1  \\\n",
       "0  -0.298336      -0.318424       -0.149973            -0.357972   \n",
       "1  -0.298336      -0.318424       -0.149973             2.698331   \n",
       "\n",
       "   hours_alternative_1  perceived_exertion_1  perceived_trainingsuccess_1  \\\n",
       "0            -0.297451             -1.001754                    -0.975817   \n",
       "1            -0.297451             -0.573368                    -0.948660   \n",
       "\n",
       "   perceived_recovery_1  nr_sessions_2  total_km_2  km_z3-4_2  km_z5-t1-t2_2  \\\n",
       "0             -1.081285       0.293492   -0.942237  -0.298543       -0.31988   \n",
       "1             -0.135464      -1.428482   -0.942237  -0.298543       -0.31988   \n",
       "\n",
       "   km_sprinting_2  strength_training_2  hours_alternative_2  \\\n",
       "0       -0.150599             2.704305            -0.296629   \n",
       "1       -0.150599            -0.356656            -0.296629   \n",
       "\n",
       "   perceived_exertion_2  perceived_trainingsuccess_2  perceived_recovery_2  \\\n",
       "0             -0.574482                    -0.949257             -0.137026   \n",
       "1             -1.002254                    -0.976427             -1.081894   \n",
       "\n",
       "   nr_sessions_3  total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3  \\\n",
       "0      -1.428499   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "1       0.294078   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "\n",
       "   strength_training_3  hours_alternative_3  perceived_exertion_3  \\\n",
       "0             -0.35636            -0.296926             -1.002298   \n",
       "1             -0.35636             1.670717             -0.652204   \n",
       "\n",
       "   perceived_trainingsuccess_3  perceived_recovery_3  nr_sessions_4  \\\n",
       "0                    -0.976792             -1.081854       0.293177   \n",
       "1                    -0.949629             -0.084824       0.293177   \n",
       "\n",
       "   total_km_4  km_z3-4_4  km_z5-t1-t2_4  km_sprinting_4  strength_training_4  \\\n",
       "0   -0.943872  -0.299153      -0.319664        -0.15043            -0.356194   \n",
       "1    1.254004   4.002433      -0.319664        -0.15043             2.708791   \n",
       "\n",
       "   hours_alternative_4  perceived_exertion_4  perceived_trainingsuccess_4  \\\n",
       "0             1.677385             -0.652858                    -0.950227   \n",
       "1            -0.297461             -0.536027                    -0.950227   \n",
       "\n",
       "   perceived_recovery_4  nr_sessions_5  total_km_5  km_z3-4_5  km_z5-t1-t2_5  \\\n",
       "0             -0.085303        0.29539    1.250975   3.996750      -0.320213   \n",
       "1             -0.137798        0.29539   -0.944348  -0.299624      -0.320213   \n",
       "\n",
       "   km_sprinting_5  strength_training_5  hours_alternative_5  \\\n",
       "0       -0.148445             2.714195            -0.296384   \n",
       "1       -0.148445            -0.355767             1.540748   \n",
       "\n",
       "   perceived_exertion_5  perceived_trainingsuccess_5  perceived_recovery_5  \\\n",
       "0             -0.535897                    -0.949919             -0.137909   \n",
       "1             -0.574816                    -0.949919             -0.242790   \n",
       "\n",
       "   nr_sessions_6  total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6  \\\n",
       "0       0.295179   -0.943941   -0.29909      -0.319835       -0.150087   \n",
       "1       0.295179   -0.247831   -0.29909      -0.044280        2.330865   \n",
       "\n",
       "   strength_training_6  hours_alternative_6  perceived_exertion_6  \\\n",
       "0            -0.356184             1.512013             -0.574761   \n",
       "1            -0.356184            -0.292961             -0.574761   \n",
       "\n",
       "   perceived_trainingsuccess_6  perceived_recovery_6      date  \n",
       "0                    -0.949638             -0.242559 -1.521713  \n",
       "1                    -0.949638             -0.137609 -1.520473  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 385,
     "status": "ok",
     "timestamp": 1662705867676,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "dOJq2kRL-weH",
    "outputId": "03bad760-ca33-4faf-d355-8d176df1a544"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((42183, 71), (583, 71))"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.shape, outliers.shape # this is the respective dimension of our no-injury and injury data in our dataset respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "executionInfo": {
     "elapsed": 288,
     "status": "ok",
     "timestamp": 1662705869320,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "k64Rg5d7-weH"
   },
   "outputs": [],
   "source": [
    "inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "w0GS4lCM-weH"
   },
   "source": [
    "### Train the isolation Forest Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "executionInfo": {
     "elapsed": 314,
     "status": "ok",
     "timestamp": 1662705871314,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "K1WoHNoV-weH"
   },
   "outputs": [],
   "source": [
    "rng = np.random.RandomState(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "executionInfo": {
     "elapsed": 277,
     "status": "ok",
     "timestamp": 1662705874740,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "JVkQCRha-weH"
   },
   "outputs": [],
   "source": [
    "inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "executionInfo": {
     "elapsed": 1095,
     "status": "ok",
     "timestamp": 1662705898921,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "UCpF9D5u-weH"
   },
   "outputs": [],
   "source": [
    "model_iso = IsolationForest( max_samples=150, random_state=rng, contamination='auto')\n",
    "# model_iso = IsolationForest( max_samples=150, random_state=rng, contamination=0.01)\n",
    "model_iso.fit(inliers_train)\n",
    "inlier_pred_test = model_iso.predict(inliers_test)\n",
    "outlier_pred = model_iso.predict(outliers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662705899652,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "F5byreA--weH",
    "outputId": "769d198c-ddc3-4b89-a14d-a4b970410e1f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy in Detecting injury Cases: 0.9615977243095887\n",
      "Accuracy in Detecting no-injury Cases: 0.0823327615780446\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy in Detecting injury Cases:\", list(inlier_pred_test).count(1)/inlier_pred_test.shape[0])\n",
    "print(\"Accuracy in Detecting no-injury Cases:\", list(outlier_pred).count(-1)/outlier_pred.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 355,
     "status": "ok",
     "timestamp": 1662705902252,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ws2irXbr-weI",
    "outputId": "cac7a440-9195-428e-ceda-8dbe0ad088d5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1,  1])"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(outlier_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 296,
     "status": "ok",
     "timestamp": 1662705887500,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ax8tQfK7-weI",
    "outputId": "acdfd657-2ace-44af-c2b4-a2a2dfc03aff"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.152\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "\n",
    "# mark inliers 1, outliers -1\n",
    "\n",
    "outliers_y = outliers_y.replace(1, -1)\n",
    "\n",
    "outliers_y = outliers_y.replace(0, 1)\n",
    "\n",
    "score = f1_score(outliers_y, outlier_pred, pos_label=-1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iA1bvJMa__Uu"
   },
   "source": [
    "## Model 3: One Class SVM II"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "executionInfo": {
     "elapsed": 386,
     "status": "ok",
     "timestamp": 1662706082865,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "VdT1OFGo__Uv"
   },
   "outputs": [],
   "source": [
    "X = df_day_scaled.drop(columns = ['injury'])\n",
    "y = df_day_scaled['injury']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "id": "iBbO-M2M__Uv"
   },
   "outputs": [],
   "source": [
    "# split into train/test sets\n",
    "trainX, testX, trainy, testy = train_test_split(X, y, test_size = 0.80, random_state=2, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "executionInfo": {
     "elapsed": 277,
     "status": "ok",
     "timestamp": 1662706094209,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "jVUR7ant__Uv"
   },
   "outputs": [],
   "source": [
    "# we fit on majority class\n",
    "trainX_copy = trainX.copy()\n",
    "trainX_copy['injury'] = trainy\n",
    "trainX_copy = trainX_copy[trainX_copy.injury == 0]\n",
    "\n",
    "trainy_copy = trainy[trainy ==0]\n",
    "\n",
    "trainX_copy = trainX_copy.drop(columns = 'injury')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 393,
     "status": "ok",
     "timestamp": 1662706096889,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "P2j-Xifw__Uv",
    "outputId": "8c550058-f7ae-4dda-cb45-a7c3697c6422"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((8436,), (8436, 71))"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainy_copy.shape, trainX_copy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "executionInfo": {
     "elapsed": 289,
     "status": "ok",
     "timestamp": 1662706099936,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "R0xBBRmv__Uw"
   },
   "outputs": [],
   "source": [
    "# define outlier detection model\n",
    "model = OneClassSVM(gamma='scale', nu=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 271,
     "status": "ok",
     "timestamp": 1662706102023,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "3aBB3iYK__Uw",
    "outputId": "314370a4-e056-470c-fc3f-d07c3daabbee"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OneClassSVM(nu=0.01)"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit on majority class\n",
    "\n",
    "model.fit(trainX_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "executionInfo": {
     "elapsed": 697,
     "status": "ok",
     "timestamp": 1662706105616,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "KlSvZcXU__Uw"
   },
   "outputs": [],
   "source": [
    "# detect outliers in the test set\n",
    "yhat = model.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "executionInfo": {
     "elapsed": 282,
     "status": "ok",
     "timestamp": 1662706107851,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "4L87a1fx__Uw"
   },
   "outputs": [],
   "source": [
    "# mark inliers 1, outliers -1\n",
    "\n",
    "testy = testy.replace(1, -1)\n",
    "\n",
    "testy = testy.replace(0, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 276,
     "status": "ok",
     "timestamp": 1662706109967,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "N1lZ7udQ__Uw",
    "outputId": "676c2bae-49cd-41f5-f9a6-39625272ccfe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.024\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "score = f1_score(testy, yhat, pos_label= -1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "79sMx1_oAreC"
   },
   "source": [
    "The outcome for both scaled and unscaled data, the results are the same in terms of f1 score for IF, but for the SVM, there is an improvement with scaling."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kvoYYIXVi6o5"
   },
   "source": [
    "# B - Repeating Models with PCA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ciigzlJ0lMzO"
   },
   "source": [
    "## Model 2: Isolation Forest III"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1662733034893,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "JglPp2sble3E"
   },
   "outputs": [],
   "source": [
    "# Make an instance of the Model\n",
    "pca_if = PCA(.95) # retain 95% variance in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "executionInfo": {
     "elapsed": 242,
     "status": "ok",
     "timestamp": 1662733038129,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "w78q11XFlMzP"
   },
   "outputs": [],
   "source": [
    "# inliers = df_day_scaled[df_day_scaled.injury == 'no']\n",
    "inliers = df_day_scaled[df_day_scaled.injury == 0] #  no injury data\n",
    "\n",
    "\n",
    "####---------may comment-------------###\n",
    "# inliers_y = inliers.injury #  no injury reponse data \n",
    "# inliers = inliers.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "executionInfo": {
     "elapsed": 280,
     "status": "ok",
     "timestamp": 1662733110171,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "w0F1XLDglMzP"
   },
   "outputs": [],
   "source": [
    "# outliers = df_day_scaled[df_day_scaled.injury == 'yes']\n",
    "outliers = df_day_scaled[df_day_scaled.injury == 1] #  injury data\n",
    "outliers_y = outliers.injury #  injury reponse data\n",
    "outliers = outliers.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 193
    },
    "executionInfo": {
     "elapsed": 6,
     "status": "ok",
     "timestamp": 1662733111890,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "dXwGIGsulMzP",
    "outputId": "73ec0452-7e47-49e2-f4fc-2be3c57fe515",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>-0.165685</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>0.011076</td>\n",
       "      <td>2.331009</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-0.535600</td>\n",
       "      <td>-0.949786</td>\n",
       "      <td>-0.083533</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>2.704305</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-0.574482</td>\n",
       "      <td>-0.949257</td>\n",
       "      <td>-0.137026</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>1.677385</td>\n",
       "      <td>-0.652858</td>\n",
       "      <td>-0.950227</td>\n",
       "      <td>-0.085303</td>\n",
       "      <td>0.29539</td>\n",
       "      <td>1.250975</td>\n",
       "      <td>3.996750</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>2.714195</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.535897</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.137909</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.29909</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>1.512013</td>\n",
       "      <td>-0.574761</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.242559</td>\n",
       "      <td>-1.521713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-1.428581</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-1.002056</td>\n",
       "      <td>-0.976938</td>\n",
       "      <td>-1.081857</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>2.698331</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-0.573368</td>\n",
       "      <td>-0.948660</td>\n",
       "      <td>-0.135464</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>1.670717</td>\n",
       "      <td>-0.652204</td>\n",
       "      <td>-0.949629</td>\n",
       "      <td>-0.084824</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>1.254004</td>\n",
       "      <td>4.002433</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>2.708791</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-0.536027</td>\n",
       "      <td>-0.950227</td>\n",
       "      <td>-0.137798</td>\n",
       "      <td>0.29539</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>1.540748</td>\n",
       "      <td>-0.574816</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.242790</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.247831</td>\n",
       "      <td>-0.29909</td>\n",
       "      <td>-0.044280</td>\n",
       "      <td>2.330865</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-0.574761</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.137609</td>\n",
       "      <td>-1.520473</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   nr_sessions  total_km   km_z3-4  km_z5-t1-t2  km_sprinting  \\\n",
       "0     0.293512 -0.165685 -0.298314     0.011076      2.331009   \n",
       "1    -1.428581 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "\n",
       "   strength_training  hours_alternative  perceived_exertion  \\\n",
       "0          -0.356549          -0.297443           -0.535600   \n",
       "1          -0.356549          -0.297443           -1.002056   \n",
       "\n",
       "   perceived_trainingsuccess  perceived_recovery  nr_sessions_1  total_km_1  \\\n",
       "0                  -0.949786           -0.083533      -1.426339   -0.939949   \n",
       "1                  -0.976938           -1.081857       0.294215   -0.939949   \n",
       "\n",
       "   km_z3-4_1  km_z5-t1-t2_1  km_sprinting_1  strength_training_1  \\\n",
       "0  -0.298336      -0.318424       -0.149973            -0.357972   \n",
       "1  -0.298336      -0.318424       -0.149973             2.698331   \n",
       "\n",
       "   hours_alternative_1  perceived_exertion_1  perceived_trainingsuccess_1  \\\n",
       "0            -0.297451             -1.001754                    -0.975817   \n",
       "1            -0.297451             -0.573368                    -0.948660   \n",
       "\n",
       "   perceived_recovery_1  nr_sessions_2  total_km_2  km_z3-4_2  km_z5-t1-t2_2  \\\n",
       "0             -1.081285       0.293492   -0.942237  -0.298543       -0.31988   \n",
       "1             -0.135464      -1.428482   -0.942237  -0.298543       -0.31988   \n",
       "\n",
       "   km_sprinting_2  strength_training_2  hours_alternative_2  \\\n",
       "0       -0.150599             2.704305            -0.296629   \n",
       "1       -0.150599            -0.356656            -0.296629   \n",
       "\n",
       "   perceived_exertion_2  perceived_trainingsuccess_2  perceived_recovery_2  \\\n",
       "0             -0.574482                    -0.949257             -0.137026   \n",
       "1             -1.002254                    -0.976427             -1.081894   \n",
       "\n",
       "   nr_sessions_3  total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3  \\\n",
       "0      -1.428499   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "1       0.294078   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "\n",
       "   strength_training_3  hours_alternative_3  perceived_exertion_3  \\\n",
       "0             -0.35636            -0.296926             -1.002298   \n",
       "1             -0.35636             1.670717             -0.652204   \n",
       "\n",
       "   perceived_trainingsuccess_3  perceived_recovery_3  nr_sessions_4  \\\n",
       "0                    -0.976792             -1.081854       0.293177   \n",
       "1                    -0.949629             -0.084824       0.293177   \n",
       "\n",
       "   total_km_4  km_z3-4_4  km_z5-t1-t2_4  km_sprinting_4  strength_training_4  \\\n",
       "0   -0.943872  -0.299153      -0.319664        -0.15043            -0.356194   \n",
       "1    1.254004   4.002433      -0.319664        -0.15043             2.708791   \n",
       "\n",
       "   hours_alternative_4  perceived_exertion_4  perceived_trainingsuccess_4  \\\n",
       "0             1.677385             -0.652858                    -0.950227   \n",
       "1            -0.297461             -0.536027                    -0.950227   \n",
       "\n",
       "   perceived_recovery_4  nr_sessions_5  total_km_5  km_z3-4_5  km_z5-t1-t2_5  \\\n",
       "0             -0.085303        0.29539    1.250975   3.996750      -0.320213   \n",
       "1             -0.137798        0.29539   -0.944348  -0.299624      -0.320213   \n",
       "\n",
       "   km_sprinting_5  strength_training_5  hours_alternative_5  \\\n",
       "0       -0.148445             2.714195            -0.296384   \n",
       "1       -0.148445            -0.355767             1.540748   \n",
       "\n",
       "   perceived_exertion_5  perceived_trainingsuccess_5  perceived_recovery_5  \\\n",
       "0             -0.535897                    -0.949919             -0.137909   \n",
       "1             -0.574816                    -0.949919             -0.242790   \n",
       "\n",
       "   nr_sessions_6  total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6  \\\n",
       "0       0.295179   -0.943941   -0.29909      -0.319835       -0.150087   \n",
       "1       0.295179   -0.247831   -0.29909      -0.044280        2.330865   \n",
       "\n",
       "   strength_training_6  hours_alternative_6  perceived_exertion_6  \\\n",
       "0            -0.356184             1.512013             -0.574761   \n",
       "1            -0.356184            -0.292961             -0.574761   \n",
       "\n",
       "   perceived_trainingsuccess_6  perceived_recovery_6      date  \n",
       "0                    -0.949638             -0.242559 -1.521713  \n",
       "1                    -0.949638             -0.137609 -1.520473  "
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 379,
     "status": "ok",
     "timestamp": 1662733129658,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "C-4Nc5qAlMzP",
    "outputId": "931b5a77-eb30-4b3b-ef1c-040093a79082"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((42183, 72), (583, 71))"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers.shape, outliers.shape # this is the respective dimension of our no-injury and injury data in our dataset respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "executionInfo": {
     "elapsed": 339,
     "status": "ok",
     "timestamp": 1662733148108,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "ea-dzE_XlMzQ"
   },
   "outputs": [],
   "source": [
    "inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "executionInfo": {
     "elapsed": 300,
     "status": "ok",
     "timestamp": 1662733151029,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "O-nFHph8mq9m"
   },
   "outputs": [],
   "source": [
    "inliers_train_y = inliers_train.injury #  no injury reponse data\n",
    "inliers_train = inliers_train.drop(['injury'], axis = 1)\n",
    "\n",
    "inliers_test_y = inliers_test.injury\n",
    "inliers_test = inliers_test.drop(['injury'], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 505
    },
    "executionInfo": {
     "elapsed": 480,
     "status": "ok",
     "timestamp": 1662733153768,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "qtionQ6TmTID",
    "outputId": "9d47ac11-f89b-444e-91c5-1094b4f9238e",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nr_sessions</th>\n",
       "      <th>total_km</th>\n",
       "      <th>km_z3-4</th>\n",
       "      <th>km_z5-t1-t2</th>\n",
       "      <th>km_sprinting</th>\n",
       "      <th>strength_training</th>\n",
       "      <th>hours_alternative</th>\n",
       "      <th>perceived_exertion</th>\n",
       "      <th>perceived_trainingsuccess</th>\n",
       "      <th>perceived_recovery</th>\n",
       "      <th>nr_sessions_1</th>\n",
       "      <th>total_km_1</th>\n",
       "      <th>km_z3-4_1</th>\n",
       "      <th>km_z5-t1-t2_1</th>\n",
       "      <th>km_sprinting_1</th>\n",
       "      <th>strength_training_1</th>\n",
       "      <th>hours_alternative_1</th>\n",
       "      <th>perceived_exertion_1</th>\n",
       "      <th>perceived_trainingsuccess_1</th>\n",
       "      <th>perceived_recovery_1</th>\n",
       "      <th>nr_sessions_2</th>\n",
       "      <th>total_km_2</th>\n",
       "      <th>km_z3-4_2</th>\n",
       "      <th>km_z5-t1-t2_2</th>\n",
       "      <th>km_sprinting_2</th>\n",
       "      <th>strength_training_2</th>\n",
       "      <th>hours_alternative_2</th>\n",
       "      <th>perceived_exertion_2</th>\n",
       "      <th>perceived_trainingsuccess_2</th>\n",
       "      <th>perceived_recovery_2</th>\n",
       "      <th>nr_sessions_3</th>\n",
       "      <th>total_km_3</th>\n",
       "      <th>km_z3-4_3</th>\n",
       "      <th>km_z5-t1-t2_3</th>\n",
       "      <th>km_sprinting_3</th>\n",
       "      <th>strength_training_3</th>\n",
       "      <th>hours_alternative_3</th>\n",
       "      <th>perceived_exertion_3</th>\n",
       "      <th>perceived_trainingsuccess_3</th>\n",
       "      <th>perceived_recovery_3</th>\n",
       "      <th>nr_sessions_4</th>\n",
       "      <th>total_km_4</th>\n",
       "      <th>km_z3-4_4</th>\n",
       "      <th>km_z5-t1-t2_4</th>\n",
       "      <th>km_sprinting_4</th>\n",
       "      <th>strength_training_4</th>\n",
       "      <th>hours_alternative_4</th>\n",
       "      <th>perceived_exertion_4</th>\n",
       "      <th>perceived_trainingsuccess_4</th>\n",
       "      <th>perceived_recovery_4</th>\n",
       "      <th>nr_sessions_5</th>\n",
       "      <th>total_km_5</th>\n",
       "      <th>km_z3-4_5</th>\n",
       "      <th>km_z5-t1-t2_5</th>\n",
       "      <th>km_sprinting_5</th>\n",
       "      <th>strength_training_5</th>\n",
       "      <th>hours_alternative_5</th>\n",
       "      <th>perceived_exertion_5</th>\n",
       "      <th>perceived_trainingsuccess_5</th>\n",
       "      <th>perceived_recovery_5</th>\n",
       "      <th>nr_sessions_6</th>\n",
       "      <th>total_km_6</th>\n",
       "      <th>km_z3-4_6</th>\n",
       "      <th>km_z5-t1-t2_6</th>\n",
       "      <th>km_sprinting_6</th>\n",
       "      <th>strength_training_6</th>\n",
       "      <th>hours_alternative_6</th>\n",
       "      <th>perceived_exertion_6</th>\n",
       "      <th>perceived_trainingsuccess_6</th>\n",
       "      <th>perceived_recovery_6</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>13649</th>\n",
       "      <td>2.015605</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>2.710874</td>\n",
       "      <td>1.067043</td>\n",
       "      <td>2.340882</td>\n",
       "      <td>1.005165</td>\n",
       "      <td>1.492768</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>0.295390</td>\n",
       "      <td>0.126541</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>0.359229</td>\n",
       "      <td>1.114713</td>\n",
       "      <td>0.701143</td>\n",
       "      <td>2.019481</td>\n",
       "      <td>0.394732</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>0.710710</td>\n",
       "      <td>0.680629</td>\n",
       "      <td>2.223775</td>\n",
       "      <td>0.356823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41478</th>\n",
       "      <td>-1.428581</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-1.002056</td>\n",
       "      <td>-0.976938</td>\n",
       "      <td>-1.081857</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-1.002919</td>\n",
       "      <td>-0.977086</td>\n",
       "      <td>-1.081842</td>\n",
       "      <td>-1.429124</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-1.003252</td>\n",
       "      <td>-0.976809</td>\n",
       "      <td>-1.082163</td>\n",
       "      <td>1.255199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30039</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>-0.138923</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-0.963185</td>\n",
       "      <td>1.303838</td>\n",
       "      <td>0.336814</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>-0.030001</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>0.124657</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>1.763280</td>\n",
       "      <td>-0.079642</td>\n",
       "      <td>1.335813</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>2.002612</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>0.008843</td>\n",
       "      <td>1.061293</td>\n",
       "      <td>0.125437</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>1.736176</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>1.889736</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>0.009085</td>\n",
       "      <td>1.033242</td>\n",
       "      <td>0.177553</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>0.128263</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-0.419196</td>\n",
       "      <td>1.114323</td>\n",
       "      <td>-0.137798</td>\n",
       "      <td>0.295390</td>\n",
       "      <td>-0.141181</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.185631</td>\n",
       "      <td>1.196212</td>\n",
       "      <td>-0.242790</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>0.140384</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>0.286387</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>1.333969</td>\n",
       "      <td>-0.596413</td>\n",
       "      <td>2.171300</td>\n",
       "      <td>0.019777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25844</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>2.710874</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>0.047471</td>\n",
       "      <td>1.249534</td>\n",
       "      <td>-0.451337</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>0.103814</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>0.124657</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>0.010794</td>\n",
       "      <td>1.576923</td>\n",
       "      <td>0.022173</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>-0.272953</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>0.008843</td>\n",
       "      <td>1.387328</td>\n",
       "      <td>0.072944</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>0.195595</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>0.509084</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>2.265246</td>\n",
       "      <td>0.082550</td>\n",
       "      <td>0.334979</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>0.295390</td>\n",
       "      <td>0.206858</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>3.136517</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.224549</td>\n",
       "      <td>1.332043</td>\n",
       "      <td>-0.347672</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>1.465670</td>\n",
       "      <td>3.143522</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-0.107317</td>\n",
       "      <td>1.305565</td>\n",
       "      <td>-0.242559</td>\n",
       "      <td>1.351852</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2854</th>\n",
       "      <td>-1.428581</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-1.002056</td>\n",
       "      <td>-0.976938</td>\n",
       "      <td>-1.081857</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-1.002919</td>\n",
       "      <td>-0.977086</td>\n",
       "      <td>-1.081842</td>\n",
       "      <td>-1.429124</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-1.003252</td>\n",
       "      <td>-0.976809</td>\n",
       "      <td>-1.082163</td>\n",
       "      <td>-0.430031</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6265</th>\n",
       "      <td>-1.428581</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-1.002056</td>\n",
       "      <td>-0.976938</td>\n",
       "      <td>-1.081857</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>-0.270870</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-0.651257</td>\n",
       "      <td>-0.948660</td>\n",
       "      <td>-0.713466</td>\n",
       "      <td>-1.428482</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-1.002254</td>\n",
       "      <td>-0.976427</td>\n",
       "      <td>-1.081894</td>\n",
       "      <td>-1.428499</td>\n",
       "      <td>-0.943095</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-1.002298</td>\n",
       "      <td>-0.976792</td>\n",
       "      <td>-1.081854</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>0.295390</td>\n",
       "      <td>-0.007320</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.613734</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.714757</td>\n",
       "      <td>-1.429124</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-1.003252</td>\n",
       "      <td>-0.976809</td>\n",
       "      <td>-1.082163</td>\n",
       "      <td>-1.328407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11284</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>0.195609</td>\n",
       "      <td>1.643325</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-0.380114</td>\n",
       "      <td>-0.949786</td>\n",
       "      <td>-0.188620</td>\n",
       "      <td>-1.426339</td>\n",
       "      <td>-0.939949</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-1.001754</td>\n",
       "      <td>-0.975817</td>\n",
       "      <td>-1.081285</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>0.664044</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-0.496706</td>\n",
       "      <td>-0.949257</td>\n",
       "      <td>-0.137026</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>0.048235</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>1.337475</td>\n",
       "      <td>0.679094</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-0.341009</td>\n",
       "      <td>-0.949629</td>\n",
       "      <td>-0.137299</td>\n",
       "      <td>0.293177</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>2.708791</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-0.574970</td>\n",
       "      <td>-0.950227</td>\n",
       "      <td>-0.137798</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-1.002919</td>\n",
       "      <td>-0.977086</td>\n",
       "      <td>-1.081842</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.943941</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>2.414500</td>\n",
       "      <td>-0.574761</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.137609</td>\n",
       "      <td>-1.159884</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38158</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>0.182228</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>1.917337</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>0.980384</td>\n",
       "      <td>0.842253</td>\n",
       "      <td>0.284271</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>0.331301</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>-0.318424</td>\n",
       "      <td>2.328444</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>0.984397</td>\n",
       "      <td>0.599278</td>\n",
       "      <td>0.705266</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>-0.942237</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>2.704305</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>0.086620</td>\n",
       "      <td>0.871105</td>\n",
       "      <td>0.545378</td>\n",
       "      <td>0.294078</td>\n",
       "      <td>1.789761</td>\n",
       "      <td>3.674940</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>0.553676</td>\n",
       "      <td>0.897429</td>\n",
       "      <td>0.649830</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>0.295390</td>\n",
       "      <td>0.287175</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>0.507820</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>1.954887</td>\n",
       "      <td>-0.243598</td>\n",
       "      <td>2.379247</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>0.890041</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>0.048498</td>\n",
       "      <td>0.707801</td>\n",
       "      <td>1.279222</td>\n",
       "      <td>1.210590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>860</th>\n",
       "      <td>2.015605</td>\n",
       "      <td>3.607835</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>-0.356549</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>-0.457857</td>\n",
       "      <td>-0.949786</td>\n",
       "      <td>-0.188620</td>\n",
       "      <td>2.014770</td>\n",
       "      <td>1.669460</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>2.173910</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>-0.495480</td>\n",
       "      <td>-0.948660</td>\n",
       "      <td>-0.135464</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>0.797901</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>-0.691147</td>\n",
       "      <td>-0.949257</td>\n",
       "      <td>-0.084534</td>\n",
       "      <td>2.016654</td>\n",
       "      <td>2.807884</td>\n",
       "      <td>3.156576</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>-0.150824</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>-0.535506</td>\n",
       "      <td>-0.949629</td>\n",
       "      <td>-0.137299</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>2.020453</td>\n",
       "      <td>1.063569</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-0.341305</td>\n",
       "      <td>-0.949919</td>\n",
       "      <td>-0.085468</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>0.260865</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>-0.319835</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>-0.613715</td>\n",
       "      <td>-0.949638</td>\n",
       "      <td>-0.295035</td>\n",
       "      <td>-1.400277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15795</th>\n",
       "      <td>0.293512</td>\n",
       "      <td>-0.941799</td>\n",
       "      <td>-0.298314</td>\n",
       "      <td>-0.320065</td>\n",
       "      <td>-0.151023</td>\n",
       "      <td>2.710874</td>\n",
       "      <td>-0.297443</td>\n",
       "      <td>1.135869</td>\n",
       "      <td>0.136298</td>\n",
       "      <td>2.806353</td>\n",
       "      <td>0.294215</td>\n",
       "      <td>-0.030001</td>\n",
       "      <td>-0.298336</td>\n",
       "      <td>0.678509</td>\n",
       "      <td>-0.149973</td>\n",
       "      <td>-0.357972</td>\n",
       "      <td>-0.297451</td>\n",
       "      <td>1.880113</td>\n",
       "      <td>1.006630</td>\n",
       "      <td>1.230722</td>\n",
       "      <td>0.293492</td>\n",
       "      <td>0.530187</td>\n",
       "      <td>-0.298543</td>\n",
       "      <td>-0.31988</td>\n",
       "      <td>-0.150599</td>\n",
       "      <td>-0.356656</td>\n",
       "      <td>-0.296629</td>\n",
       "      <td>0.242173</td>\n",
       "      <td>1.414497</td>\n",
       "      <td>-0.137026</td>\n",
       "      <td>2.016654</td>\n",
       "      <td>-0.045539</td>\n",
       "      <td>-0.299187</td>\n",
       "      <td>-0.319307</td>\n",
       "      <td>1.509012</td>\n",
       "      <td>-0.35636</td>\n",
       "      <td>-0.296926</td>\n",
       "      <td>1.487260</td>\n",
       "      <td>0.761616</td>\n",
       "      <td>1.384484</td>\n",
       "      <td>-1.430026</td>\n",
       "      <td>-0.943872</td>\n",
       "      <td>-0.299153</td>\n",
       "      <td>-0.319664</td>\n",
       "      <td>-0.15043</td>\n",
       "      <td>-0.356194</td>\n",
       "      <td>-0.297461</td>\n",
       "      <td>-1.003349</td>\n",
       "      <td>-0.977392</td>\n",
       "      <td>-1.082694</td>\n",
       "      <td>-1.429674</td>\n",
       "      <td>-0.944348</td>\n",
       "      <td>-0.299624</td>\n",
       "      <td>-0.320213</td>\n",
       "      <td>-0.148445</td>\n",
       "      <td>-0.355767</td>\n",
       "      <td>-0.296384</td>\n",
       "      <td>-1.002919</td>\n",
       "      <td>-0.977086</td>\n",
       "      <td>-1.081842</td>\n",
       "      <td>0.295179</td>\n",
       "      <td>-0.274604</td>\n",
       "      <td>-0.299090</td>\n",
       "      <td>0.231276</td>\n",
       "      <td>-0.150087</td>\n",
       "      <td>-0.356184</td>\n",
       "      <td>-0.292961</td>\n",
       "      <td>1.762459</td>\n",
       "      <td>1.196881</td>\n",
       "      <td>0.072292</td>\n",
       "      <td>1.400178</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>33746 rows × 71 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       nr_sessions  total_km   km_z3-4  km_z5-t1-t2  km_sprinting  \\\n",
       "13649     2.015605 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "41478    -1.428581 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "30039     0.293512 -0.138923 -0.298314    -0.320065     -0.151023   \n",
       "25844     0.293512 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "2854     -1.428581 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "...            ...       ...       ...          ...           ...   \n",
       "6265     -1.428581 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "11284     0.293512  0.195609  1.643325    -0.320065     -0.151023   \n",
       "38158     0.293512  0.182228 -0.298314    -0.320065      1.917337   \n",
       "860       2.015605  3.607835 -0.298314    -0.320065     -0.151023   \n",
       "15795     0.293512 -0.941799 -0.298314    -0.320065     -0.151023   \n",
       "\n",
       "       strength_training  hours_alternative  perceived_exertion  \\\n",
       "13649           2.710874           1.067043            2.340882   \n",
       "41478          -0.356549          -0.297443           -1.002056   \n",
       "30039          -0.356549          -0.297443           -0.963185   \n",
       "25844           2.710874          -0.297443            0.047471   \n",
       "2854           -0.356549          -0.297443           -1.002056   \n",
       "...                  ...                ...                 ...   \n",
       "6265           -0.356549          -0.297443           -1.002056   \n",
       "11284          -0.356549          -0.297443           -0.380114   \n",
       "38158          -0.356549          -0.297443            0.980384   \n",
       "860            -0.356549          -0.297443           -0.457857   \n",
       "15795           2.710874          -0.297443            1.135869   \n",
       "\n",
       "       perceived_trainingsuccess  perceived_recovery  nr_sessions_1  \\\n",
       "13649                   1.005165            1.492768      -1.426339   \n",
       "41478                  -0.976938           -1.081857      -1.426339   \n",
       "30039                   1.303838            0.336814       0.294215   \n",
       "25844                   1.249534           -0.451337       0.294215   \n",
       "2854                   -0.976938           -1.081857      -1.426339   \n",
       "...                          ...                 ...            ...   \n",
       "6265                   -0.976938           -1.081857       0.294215   \n",
       "11284                  -0.949786           -0.188620      -1.426339   \n",
       "38158                   0.842253            0.284271       0.294215   \n",
       "860                    -0.949786           -0.188620       2.014770   \n",
       "15795                   0.136298            2.806353       0.294215   \n",
       "\n",
       "       total_km_1  km_z3-4_1  km_z5-t1-t2_1  km_sprinting_1  \\\n",
       "13649   -0.939949  -0.298336      -0.318424       -0.149973   \n",
       "41478   -0.939949  -0.298336      -0.318424       -0.149973   \n",
       "30039   -0.030001  -0.298336       0.124657       -0.149973   \n",
       "25844    0.103814  -0.298336       0.124657       -0.149973   \n",
       "2854    -0.939949  -0.298336      -0.318424       -0.149973   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.270870  -0.298336      -0.318424       -0.149973   \n",
       "11284   -0.939949  -0.298336      -0.318424       -0.149973   \n",
       "38158    0.331301  -0.298336      -0.318424        2.328444   \n",
       "860      1.669460  -0.298336       2.173910       -0.149973   \n",
       "15795   -0.030001  -0.298336       0.678509       -0.149973   \n",
       "\n",
       "       strength_training_1  hours_alternative_1  perceived_exertion_1  \\\n",
       "13649            -0.357972            -0.297451             -1.001754   \n",
       "41478            -0.357972            -0.297451             -1.001754   \n",
       "30039            -0.357972            -0.297451              1.763280   \n",
       "25844            -0.357972            -0.297451              0.010794   \n",
       "2854             -0.357972            -0.297451             -1.001754   \n",
       "...                    ...                  ...                   ...   \n",
       "6265             -0.357972            -0.297451             -0.651257   \n",
       "11284            -0.357972            -0.297451             -1.001754   \n",
       "38158            -0.357972            -0.297451              0.984397   \n",
       "860              -0.357972            -0.297451             -0.495480   \n",
       "15795            -0.357972            -0.297451              1.880113   \n",
       "\n",
       "       perceived_trainingsuccess_1  perceived_recovery_1  nr_sessions_2  \\\n",
       "13649                    -0.975817             -1.081285      -1.428482   \n",
       "41478                    -0.975817             -1.081285      -1.428482   \n",
       "30039                    -0.079642              1.335813       0.293492   \n",
       "25844                     1.576923              0.022173       0.293492   \n",
       "2854                     -0.975817             -1.081285      -1.428482   \n",
       "...                            ...                   ...            ...   \n",
       "6265                     -0.948660             -0.713466      -1.428482   \n",
       "11284                    -0.975817             -1.081285       0.293492   \n",
       "38158                     0.599278              0.705266       0.293492   \n",
       "860                      -0.948660             -0.135464       0.293492   \n",
       "15795                     1.006630              1.230722       0.293492   \n",
       "\n",
       "       total_km_2  km_z3-4_2  km_z5-t1-t2_2  km_sprinting_2  \\\n",
       "13649   -0.942237  -0.298543       -0.31988       -0.150599   \n",
       "41478   -0.942237  -0.298543       -0.31988       -0.150599   \n",
       "30039    2.002612  -0.298543       -0.31988       -0.150599   \n",
       "25844   -0.272953  -0.298543       -0.31988       -0.150599   \n",
       "2854    -0.942237  -0.298543       -0.31988       -0.150599   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.942237  -0.298543       -0.31988       -0.150599   \n",
       "11284    0.664044  -0.298543       -0.31988       -0.150599   \n",
       "38158   -0.942237  -0.298543       -0.31988       -0.150599   \n",
       "860      0.797901  -0.298543       -0.31988       -0.150599   \n",
       "15795    0.530187  -0.298543       -0.31988       -0.150599   \n",
       "\n",
       "       strength_training_2  hours_alternative_2  perceived_exertion_2  \\\n",
       "13649            -0.356656            -0.296629             -1.002254   \n",
       "41478            -0.356656            -0.296629             -1.002254   \n",
       "30039            -0.356656            -0.296629              0.008843   \n",
       "25844            -0.356656            -0.296629              0.008843   \n",
       "2854             -0.356656            -0.296629             -1.002254   \n",
       "...                    ...                  ...                   ...   \n",
       "6265             -0.356656            -0.296629             -1.002254   \n",
       "11284            -0.356656            -0.296629             -0.496706   \n",
       "38158             2.704305            -0.296629              0.086620   \n",
       "860              -0.356656            -0.296629             -0.691147   \n",
       "15795            -0.356656            -0.296629              0.242173   \n",
       "\n",
       "       perceived_trainingsuccess_2  perceived_recovery_2  nr_sessions_3  \\\n",
       "13649                    -0.976427             -1.081894      -1.428499   \n",
       "41478                    -0.976427             -1.081894      -1.428499   \n",
       "30039                     1.061293              0.125437       0.294078   \n",
       "25844                     1.387328              0.072944       0.294078   \n",
       "2854                     -0.976427             -1.081894      -1.428499   \n",
       "...                            ...                   ...            ...   \n",
       "6265                     -0.976427             -1.081894      -1.428499   \n",
       "11284                    -0.949257             -0.137026       0.294078   \n",
       "38158                     0.871105              0.545378       0.294078   \n",
       "860                      -0.949257             -0.084534       2.016654   \n",
       "15795                     1.414497             -0.137026       2.016654   \n",
       "\n",
       "       total_km_3  km_z3-4_3  km_z5-t1-t2_3  km_sprinting_3  \\\n",
       "13649   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "41478   -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "30039    1.736176  -0.299187       1.889736       -0.150824   \n",
       "25844    0.195595  -0.299187       0.509084       -0.150824   \n",
       "2854    -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.943095  -0.299187      -0.319307       -0.150824   \n",
       "11284    0.048235  -0.299187       1.337475        0.679094   \n",
       "38158    1.789761   3.674940      -0.319307       -0.150824   \n",
       "860      2.807884   3.156576      -0.319307       -0.150824   \n",
       "15795   -0.045539  -0.299187      -0.319307        1.509012   \n",
       "\n",
       "       strength_training_3  hours_alternative_3  perceived_exertion_3  \\\n",
       "13649             -0.35636            -0.296926             -1.002298   \n",
       "41478             -0.35636            -0.296926             -1.002298   \n",
       "30039             -0.35636            -0.296926              0.009085   \n",
       "25844             -0.35636            -0.296926              2.265246   \n",
       "2854              -0.35636            -0.296926             -1.002298   \n",
       "...                    ...                  ...                   ...   \n",
       "6265              -0.35636            -0.296926             -1.002298   \n",
       "11284             -0.35636            -0.296926             -0.341009   \n",
       "38158             -0.35636            -0.296926              0.553676   \n",
       "860               -0.35636            -0.296926             -0.535506   \n",
       "15795             -0.35636            -0.296926              1.487260   \n",
       "\n",
       "       perceived_trainingsuccess_3  perceived_recovery_3  nr_sessions_4  \\\n",
       "13649                    -0.976792             -1.081854      -1.430026   \n",
       "41478                    -0.976792             -1.081854      -1.430026   \n",
       "30039                     1.033242              0.177553       0.293177   \n",
       "25844                     0.082550              0.334979      -1.430026   \n",
       "2854                     -0.976792             -1.081854      -1.430026   \n",
       "...                            ...                   ...            ...   \n",
       "6265                     -0.976792             -1.081854      -1.430026   \n",
       "11284                    -0.949629             -0.137299       0.293177   \n",
       "38158                     0.897429              0.649830      -1.430026   \n",
       "860                      -0.949629             -0.137299      -1.430026   \n",
       "15795                     0.761616              1.384484      -1.430026   \n",
       "\n",
       "       total_km_4  km_z3-4_4  km_z5-t1-t2_4  km_sprinting_4  \\\n",
       "13649   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "41478   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "30039    0.128263  -0.299153      -0.319664        -0.15043   \n",
       "25844   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "2854    -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "11284   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "38158   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "860     -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "15795   -0.943872  -0.299153      -0.319664        -0.15043   \n",
       "\n",
       "       strength_training_4  hours_alternative_4  perceived_exertion_4  \\\n",
       "13649            -0.356194            -0.297461             -1.003349   \n",
       "41478            -0.356194            -0.297461             -1.003349   \n",
       "30039            -0.356194            -0.297461             -0.419196   \n",
       "25844            -0.356194            -0.297461             -1.003349   \n",
       "2854             -0.356194            -0.297461             -1.003349   \n",
       "...                    ...                  ...                   ...   \n",
       "6265             -0.356194            -0.297461             -1.003349   \n",
       "11284             2.708791            -0.297461             -0.574970   \n",
       "38158            -0.356194            -0.297461             -1.003349   \n",
       "860              -0.356194            -0.297461             -1.003349   \n",
       "15795            -0.356194            -0.297461             -1.003349   \n",
       "\n",
       "       perceived_trainingsuccess_4  perceived_recovery_4  nr_sessions_5  \\\n",
       "13649                    -0.977392             -1.082694       0.295390   \n",
       "41478                    -0.977392             -1.082694      -1.429674   \n",
       "30039                     1.114323             -0.137798       0.295390   \n",
       "25844                    -0.977392             -1.082694       0.295390   \n",
       "2854                     -0.977392             -1.082694      -1.429674   \n",
       "...                            ...                   ...            ...   \n",
       "6265                     -0.977392             -1.082694       0.295390   \n",
       "11284                    -0.950227             -0.137798      -1.429674   \n",
       "38158                    -0.977392             -1.082694       0.295390   \n",
       "860                      -0.977392             -1.082694       2.020453   \n",
       "15795                    -0.977392             -1.082694      -1.429674   \n",
       "\n",
       "       total_km_5  km_z3-4_5  km_z5-t1-t2_5  km_sprinting_5  \\\n",
       "13649    0.126541  -0.299624      -0.320213       -0.148445   \n",
       "41478   -0.944348  -0.299624      -0.320213       -0.148445   \n",
       "30039   -0.141181  -0.299624      -0.320213       -0.148445   \n",
       "25844    0.206858  -0.299624      -0.320213        3.136517   \n",
       "2854    -0.944348  -0.299624      -0.320213       -0.148445   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.007320  -0.299624      -0.320213       -0.148445   \n",
       "11284   -0.944348  -0.299624      -0.320213       -0.148445   \n",
       "38158    0.287175  -0.299624       0.507820       -0.148445   \n",
       "860      1.063569  -0.299624      -0.320213       -0.148445   \n",
       "15795   -0.944348  -0.299624      -0.320213       -0.148445   \n",
       "\n",
       "       strength_training_5  hours_alternative_5  perceived_exertion_5  \\\n",
       "13649            -0.355767            -0.296384              0.359229   \n",
       "41478            -0.355767            -0.296384             -1.002919   \n",
       "30039            -0.355767            -0.296384             -0.185631   \n",
       "25844            -0.355767            -0.296384             -0.224549   \n",
       "2854             -0.355767            -0.296384             -1.002919   \n",
       "...                    ...                  ...                   ...   \n",
       "6265             -0.355767            -0.296384             -0.613734   \n",
       "11284            -0.355767            -0.296384             -1.002919   \n",
       "38158            -0.355767            -0.296384              1.954887   \n",
       "860              -0.355767            -0.296384             -0.341305   \n",
       "15795            -0.355767            -0.296384             -1.002919   \n",
       "\n",
       "       perceived_trainingsuccess_5  perceived_recovery_5  nr_sessions_6  \\\n",
       "13649                     1.114713              0.701143       2.019481   \n",
       "41478                    -0.977086             -1.081842      -1.429124   \n",
       "30039                     1.196212             -0.242790       0.295179   \n",
       "25844                     1.332043             -0.347672       0.295179   \n",
       "2854                     -0.977086             -1.081842      -1.429124   \n",
       "...                            ...                   ...            ...   \n",
       "6265                     -0.949919             -0.714757      -1.429124   \n",
       "11284                    -0.977086             -1.081842       0.295179   \n",
       "38158                    -0.243598              2.379247       0.295179   \n",
       "860                      -0.949919             -0.085468       0.295179   \n",
       "15795                    -0.977086             -1.081842       0.295179   \n",
       "\n",
       "       total_km_6  km_z3-4_6  km_z5-t1-t2_6  km_sprinting_6  \\\n",
       "13649    0.394732  -0.299090      -0.319835       -0.150087   \n",
       "41478   -0.943941  -0.299090      -0.319835       -0.150087   \n",
       "30039    0.140384  -0.299090       0.286387       -0.150087   \n",
       "25844    1.465670   3.143522      -0.319835       -0.150087   \n",
       "2854    -0.943941  -0.299090      -0.319835       -0.150087   \n",
       "...           ...        ...            ...             ...   \n",
       "6265    -0.943941  -0.299090      -0.319835       -0.150087   \n",
       "11284   -0.943941  -0.299090      -0.319835       -0.150087   \n",
       "38158    0.890041  -0.299090      -0.319835       -0.150087   \n",
       "860      0.260865  -0.299090      -0.319835       -0.150087   \n",
       "15795   -0.274604  -0.299090       0.231276       -0.150087   \n",
       "\n",
       "       strength_training_6  hours_alternative_6  perceived_exertion_6  \\\n",
       "13649            -0.356184            -0.292961              0.710710   \n",
       "41478            -0.356184            -0.292961             -1.003252   \n",
       "30039            -0.356184            -0.292961              1.333969   \n",
       "25844            -0.356184            -0.292961             -0.107317   \n",
       "2854             -0.356184            -0.292961             -1.003252   \n",
       "...                    ...                  ...                   ...   \n",
       "6265             -0.356184            -0.292961             -1.003252   \n",
       "11284            -0.356184             2.414500             -0.574761   \n",
       "38158            -0.356184            -0.292961              0.048498   \n",
       "860              -0.356184            -0.292961             -0.613715   \n",
       "15795            -0.356184            -0.292961              1.762459   \n",
       "\n",
       "       perceived_trainingsuccess_6  perceived_recovery_6      date  \n",
       "13649                     0.680629              2.223775  0.356823  \n",
       "41478                    -0.976809             -1.082163  1.255199  \n",
       "30039                    -0.596413              2.171300  0.019777  \n",
       "25844                     1.305565             -0.242559  1.351852  \n",
       "2854                     -0.976809             -1.082163 -0.430031  \n",
       "...                            ...                   ...       ...  \n",
       "6265                     -0.976809             -1.082163 -1.328407  \n",
       "11284                    -0.949638             -0.137609 -1.159884  \n",
       "38158                     0.707801              1.279222  1.210590  \n",
       "860                      -0.949638             -0.295035 -1.400277  \n",
       "15795                     1.196881              0.072292  1.400178  \n",
       "\n",
       "[33746 rows x 71 columns]"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RJ27gfNKl624"
   },
   "source": [
    "### Apply PCA on data splits"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kE_kzAEnmIHq"
   },
   "source": [
    "Fit PCA on training set on training set alone."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 276,
     "status": "ok",
     "timestamp": 1662733392134,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "dc7UMXelmJdl",
    "outputId": "803095be-1c8c-4c8a-f1a2-6d8addb97d15",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(n_components=0.95)"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca_if.fit(inliers_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NkNC1c-ynzww"
   },
   "source": [
    "Apply the mapping (transform) to both the training set and the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {
    "colab": {
     "background_save": true,
     "output_embedded_package_id": "1zH0UOo5HYgO6vNJw3Jy0p32CsYf3haEu"
    },
    "id": "w3M5BnW2o0kt",
    "outputId": "3a70fd94-7020-44f6-be89-4edcaec7ac91",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(33746, 71)"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inliers_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 399
    },
    "executionInfo": {
     "elapsed": 598,
     "status": "error",
     "timestamp": 1662733394315,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "EtyJb4SRmAf6",
    "outputId": "b497c651-cfea-4929-b2cd-d7399fc06d10"
   },
   "outputs": [],
   "source": [
    "inliers_train = pca_if.transform(inliers_train)\n",
    "inliers_test = pca_if.transform(inliers_test)\n",
    "outliers = pca_if.transform(outliers)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dNGH17ctlMzQ"
   },
   "source": [
    "### Train the isolation Forest Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "executionInfo": {
     "elapsed": 375,
     "status": "ok",
     "timestamp": 1662733299118,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "1E32wNCKlMzQ"
   },
   "outputs": [],
   "source": [
    "rng = np.random.RandomState(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "id": "lb8go5aolMzQ"
   },
   "outputs": [],
   "source": [
    "# inliers_train, inliers_test = train_test_split(inliers, test_size=0.20, random_state = 42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 419
    },
    "executionInfo": {
     "elapsed": 1856,
     "status": "error",
     "timestamp": 1662733331512,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "BJm-swXZlMzQ",
    "outputId": "26807a54-b4f6-4677-9e61-c0e4285277ab"
   },
   "outputs": [],
   "source": [
    "model_iso = IsolationForest( max_samples=150, random_state=rng, contamination='auto')\n",
    "# model_iso = IsolationForest( max_samples=150, random_state=rng, contamination=0.01)\n",
    "model_iso.fit(inliers_train)\n",
    "inlier_pred_test = model_iso.predict(inliers_test)\n",
    "outlier_pred = model_iso.predict(outliers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1662705899652,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "yFPTyb5ClMzQ",
    "outputId": "769d198c-ddc3-4b89-a14d-a4b970410e1f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy in Detecting injury Cases: 0.9502192722531706\n",
      "Accuracy in Detecting no-injury Cases: 0.0548885077186964\n"
     ]
    }
   ],
   "source": [
    "print(\"Accuracy in Detecting injury Cases:\", list(inlier_pred_test).count(1)/inlier_pred_test.shape[0])\n",
    "print(\"Accuracy in Detecting no-injury Cases:\", list(outlier_pred).count(-1)/outlier_pred.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 355,
     "status": "ok",
     "timestamp": 1662705902252,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "IqQfrEx2lMzR",
    "outputId": "cac7a440-9195-428e-ceda-8dbe0ad088d5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-1,  1])"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(outlier_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 296,
     "status": "ok",
     "timestamp": 1662705887500,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "cUGyDlH1lMzR",
    "outputId": "acdfd657-2ace-44af-c2b4-a2a2dfc03aff"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.104\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "\n",
    "# mark inliers 1, outliers -1\n",
    "\n",
    "outliers_y = outliers_y.replace(1, -1)\n",
    "\n",
    "outliers_y = outliers_y.replace(0, 1)\n",
    "\n",
    "score = f1_score(outliers_y, outlier_pred, pos_label=-1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The use of PCA reduced our prediction accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0-UxUllslMzR"
   },
   "source": [
    "## Model 3: One Class SVM III"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "id": "Ql06Rjd0lMzR"
   },
   "outputs": [],
   "source": [
    "X = df_day_scaled.drop(columns = ['injury'])\n",
    "y = df_day_scaled['injury']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "id": "hbLgHrunlMzR"
   },
   "outputs": [],
   "source": [
    "# split into train/test sets\n",
    "trainX, testX, trainy, testy = train_test_split(X, y, test_size=0.80, random_state=2, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "id": "boDYbiTwlMzR"
   },
   "outputs": [],
   "source": [
    "# we fit on majority class\n",
    "trainX_copy = trainX.copy()\n",
    "trainX_copy['injury'] = trainy\n",
    "trainX_copy = trainX_copy[trainX_copy.injury == 0]\n",
    "\n",
    "trainy_copy = trainy[trainy ==0]\n",
    "trainX_copy = trainX_copy.drop(columns = 'injury')\n",
    "# trainy_copy = "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "RJ27gfNKl624"
   },
   "source": [
    "### Apply PCA on data splits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1662733034893,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "JglPp2sble3E"
   },
   "outputs": [],
   "source": [
    "# Make an instance of the Model\n",
    "pca_oc_svm = PCA(.95) # retain 95% variance in the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kE_kzAEnmIHq"
   },
   "source": [
    "Fit PCA on training set on training set alone."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 276,
     "status": "ok",
     "timestamp": 1662733392134,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "dc7UMXelmJdl",
    "outputId": "803095be-1c8c-4c8a-f1a2-6d8addb97d15",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PCA(n_components=0.95)"
      ]
     },
     "execution_count": 129,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pca_oc_svm.fit(trainX_copy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NkNC1c-ynzww"
   },
   "source": [
    "Apply the mapping (transform) to both the training set and the test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 399
    },
    "executionInfo": {
     "elapsed": 598,
     "status": "error",
     "timestamp": 1662733394315,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "EtyJb4SRmAf6",
    "outputId": "b497c651-cfea-4929-b2cd-d7399fc06d10"
   },
   "outputs": [],
   "source": [
    "trainX_copy = pca_oc_svm.transform(trainX_copy)\n",
    "testX = pca_oc_svm.transform(testX)\n",
    "# outliers = pca_oc_svm.transform(outliers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 393,
     "status": "ok",
     "timestamp": 1662706096889,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "PAxeTFj-lMzR",
    "outputId": "8c550058-f7ae-4dda-cb45-a7c3697c6422"
   },
   "outputs": [],
   "source": [
    "# trainy_copy.shape, trainX_copy.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "id": "73iYZqdrlMzS"
   },
   "outputs": [],
   "source": [
    "# define outlier detection model\n",
    "model_one_class_svm_pca = OneClassSVM(gamma='scale', nu=0.01)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 271,
     "status": "ok",
     "timestamp": 1662706102023,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "aQrGnofilMzS",
    "outputId": "314370a4-e056-470c-fc3f-d07c3daabbee"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OneClassSVM(nu=0.01)"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# fit on majority class\n",
    "\n",
    "model_one_class_svm_pca.fit(trainX_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {
    "id": "PDVXWyunlMzS"
   },
   "outputs": [],
   "source": [
    "# detect outliers in the test set\n",
    "yhat = model_one_class_svm_pca.predict(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "id": "PbSMi-v7lMzS"
   },
   "outputs": [],
   "source": [
    "# mark inliers 1, outliers -1\n",
    "\n",
    "testy = testy.replace(1, -1)\n",
    "\n",
    "testy = testy.replace(0, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 276,
     "status": "ok",
     "timestamp": 1662706109967,
     "user": {
      "displayName": "Olusegun Ajibola",
      "userId": "14016280963587867119"
     },
     "user_tz": -120
    },
    "id": "siy27vfclMzW",
    "outputId": "676c2bae-49cd-41f5-f9a6-39625272ccfe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score: 0.022\n"
     ]
    }
   ],
   "source": [
    "# calculate score\n",
    "score = f1_score(testy, yhat, pos_label= -1)\n",
    "print('F1 Score: %.3f' % score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UmiKJ1PKjalu"
   },
   "source": [
    "The accuracy further decreased  by 0.002 when we use PCA in One Class SVM."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "LHDzPS1PAyAy"
   },
   "source": [
    "# Conclusion I\n",
    "\n",
    "- All models i.e Autoenconders, Isolation Forest and One Class SVM have a high accuracy.  \n",
    "- Isolation Forest has the best f1 score and we conclude it to be our best model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "pEMjRllpBriS"
   },
   "source": [
    "# Weekly Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "authorship_tag": "ABX9TyNatMkjs60/4PGct5YcRMAf",
   "collapsed_sections": [],
   "mount_file_id": "1PJm3hWtdtvX-hDKbbw5rNUA1gbKfIyzn",
   "name": "",
   "toc_visible": true,
   "version": ""
  },
  "gpuClass": "standard",
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  },
  "toc-autonumbering": false
 },
 "nbformat": 4,
 "nbformat_minor": 4
}