sistema_recomendacion.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Proyecto Final\n",
    "\n",
    "***Equipo 07***\n",
    "\n",
    "- Aide Jazmín González Cruz\n",
    "- Elena Villalobos Nolasco\n",
    "- Carolina Acosta Tovany\n",
    "\n",
    "#### Instrucciones\n",
    "\n",
    "El proyecto/examen final consistirá en:\n",
    "\n",
    "La implementación del algoritmo de filtrado colaborativo utilizando la metodología vista en clase (uso de otra metodología no se calificará).\n",
    "\n",
    "Todos los algoritmos de aprendizaje de máquina que se utilicen deberán haber sido creados por ustedes. Sólo podrán utilizar Transformers y funciones de apoyo de scikit-learn (para realizar la división de los datos en entrenamiento y prueba, o el procedimiento de validación cruzada, etc.) mas ningún estimator (regresión logística, máquina de vectores de soporte, k medias, etc.). \n",
    "\n",
    "Se deberá explicar como se obtuvo la k con la que se generó el resultado final.\n",
    "\n",
    "Se utilizarán los archivos con el conjunto pequeño de calificaciones y películas ubicado en la siguiente https://www.kaggle.com/rounakbanik/the-movies-dataset:\n",
    "\n",
    "- **links_small.csv**: Contains the TMDB and IMDB IDs of a small subset of 9,000 movies of the Full Dataset.\n",
    "\n",
    "- **ratings_small.csv**: The subset of 100,000 ratings from 700 users on 9,000 movies.\n",
    "\n",
    "Con el fin de mejorar la calificación (opcional, puntos extra), se podrán utilizar los algoritmos desarrollado en las tareas del curso y los datos relevantes (los que hacen match con los datos anteriores) contenidos en los archivos:\n",
    "\n",
    "- **movies_metadata.csv**: The main Movies Metadata file. Contains information on 45,000 movies featured in the Full MovieLens dataset. Features include posters, backdrops, budget, revenue, release dates, languages, production countries and companies.\n",
    "\n",
    "- **keywords.csv**: Contains the movie plot keywords for our MovieLens movies. Available in the form of a stringified JSON Object.\n",
    "\n",
    "- **credits.csv**: Consists of Cast and Crew Information for all our movies. Available in the form of a stringified JSON Object.\n",
    "\n",
    "La métrica con la que se determinará el desempeño del algoritmo es el NDCG \n",
    "\n",
    "(https://en.wikipedia.org/wiki/Discounted_cumulative_gain#Normalized_DCG)\n",
    "\n",
    "Una vez obtenida la matriz de calificaciones, el programa deberá ser capaz de regresar las 5 mejores recomendaciones del o de los usuarios que se consulten.\n",
    "\n",
    "El proyecto se entregará en un Jupyter notebook. El readme file debe contener las instrucciones para que se ejecute el código. Deben cerciorarse que siguiendo esas instrucciones el programa corre sin errores. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importación de paqueterías necesarias\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import random\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importación de paqueterías creadas por nosotras\n",
    "from sis_recom import load_data \n",
    "from sis_recom import create_train_test, fun_ECM\n",
    "from sis_recom import recomendaciones_id\n",
    "from sis_recom import desempenio_NDCG, obtain_ndcg_all_users"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "La función objetivo:\n",
    "    \n",
    "$$J(X) = \\frac{1}{2} \\displaystyle\\sum_{(a,i)\\in\\mathbb{D}} \\left(Y_{ai}-\\left [ UV^T \\right ]_{ai} \\right)^2 + \\frac{\\lambda}{2} \\displaystyle\\sum_{a=1}^n \\displaystyle\\sum_{j=1}^k U_{aj}^2 + \\frac{\\lambda}{2} \\displaystyle\\sum_{i=1}^m \\displaystyle\\sum_{j=1}^k V_{ij}^2$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ya derivadas e igualadas a cero dependiendo de los usuarios e items:\n",
    "\n",
    "$$x_u^T = r_uY\\left(Y^TY+\\lambda_xI\\right)^{-1}$$\n",
    "\n",
    "$$y_i^T = r_iX\\left(X^TX+\\lambda_yI\\right)^{-1}$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Cargamos datos ya limpios\n",
    "ratings, bases_nombres_id, arr_movies = load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>161830</th>\n",
       "      <th>161918</th>\n",
       "      <th>161944</th>\n",
       "      <th>162376</th>\n",
       "      <th>162542</th>\n",
       "      <th>162672</th>\n",
       "      <th>163056</th>\n",
       "      <th>163949</th>\n",
       "      <th>164977</th>\n",
       "      <th>164979</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 9125 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     1    2    3    4    5    6    7    8    9   10  ...  161830  161918  \\\n",
       "1  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  ...     0.0     0.0   \n",
       "2  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  4.0  ...     0.0     0.0   \n",
       "3  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  ...     0.0     0.0   \n",
       "4  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  4.0  ...     0.0     0.0   \n",
       "5  0.0  0.0  4.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  ...     0.0     0.0   \n",
       "\n",
       "   161944  162376  162542  162672  163056  163949  164977  164979  \n",
       "1     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0  \n",
       "2     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0  \n",
       "3     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0  \n",
       "4     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0  \n",
       "5     0.0     0.0     0.0     0.0     0.0     0.0     0.0     0.0  \n",
       "\n",
       "[5 rows x 9125 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_users, n_items = ratings.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(671, 9125)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_users, n_items"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Separando datos en train y test\n",
    "train, test = create_train_test(np.array(ratings))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [4., 0., 0., ..., 0., 0., 0.],\n",
       "       [5., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Matrix_Factorization:\n",
    "    \"\"\"\n",
    "    Factorización de matrices para predecir entradas vacías en una matriz\n",
    "    \n",
    "    Parametros\n",
    "    ----------\n",
    "    n_iters : numero de iteraciones para entregar algoritmo\n",
    "        \n",
    "    k : rank de la matriz\n",
    "        \n",
    "    _lambda : regularización\n",
    "    \"\"\"\n",
    "    \n",
    "    def __init__(self, n_iters, k, _lambda):\n",
    "        self._lambda = _lambda\n",
    "        self.n_iters = n_iters\n",
    "        self.k = k  \n",
    "        \n",
    "    def fit(self, train, test):\n",
    "        \"\"\"\n",
    "        partición de entrenamiento y test\n",
    "        y selección de vectores aleatorios\n",
    "        \"\"\"\n",
    "        self.n, self.m = train.shape\n",
    "        \n",
    "        self.Xu = np.random.random((self.n, self.k))\n",
    "        self.Yi = np.random.random((self.m, self.k))\n",
    "        \n",
    "        self.test_ecm_record  = []\n",
    "        self.train_ecm_record = []\n",
    "        \n",
    "        for _ in range(self.n_iters):\n",
    "            self.Xu = self._als_step(train, self.Xu, self.Yi)\n",
    "            self.Yi = self._als_step(train.T, self.Yi, self.Xu) \n",
    "            predictions = self.predict()\n",
    "            test_ecm = self.compute_ecm(test, predictions)\n",
    "            train_ecm = self.compute_ecm(train, predictions)\n",
    "            self.train_ecm_record.append(train_ecm)\n",
    "            self.test_ecm_record.append(test_ecm)\n",
    "        \n",
    "        return self    \n",
    "    \n",
    "    def _als_step(self, ratings, vec_res, vec_fij):\n",
    "        \"\"\"\n",
    "        función que obtiene los vectores X_u ó Y_i\n",
    "        \"\"\"\n",
    "        A = vec_fij.T.dot(vec_fij) + np.eye(self.k) * self._lambda\n",
    "        b = ratings.dot(vec_fij)\n",
    "        A_inv = np.linalg.inv(A)\n",
    "        vec_res = b.dot(A_inv)\n",
    "        return vec_res\n",
    "    \n",
    "    def predict(self):\n",
    "        \"\"\"Predicción de matriz completa XuYi (UV)\"\"\"\n",
    "        pred = self.Xu.dot(self.Yi.T)\n",
    "        return pred\n",
    "    \n",
    "    @staticmethod\n",
    "    def compute_ecm(y_true, y_pred):\n",
    "        mask = np.nonzero(y_true)\n",
    "        ecm = fun_ECM(y_true[mask], y_pred[mask])\n",
    "\n",
    "        return ecm\n",
    "    \n",
    "def best_K(train, test, K_list, n_iters=15, lambda_ = 0.01):\n",
    "    df_k = pd.DataFrame(0.0, index=range(1), columns=K_list)    \n",
    "    for k in K_list:\n",
    "        als = Matrix_Factorization(n_iters, k, lambda_)\n",
    "        als.fit(train, test)\n",
    "        error = np.mean(als.test_ecm_record)\n",
    "        df_k[k][0] = error\n",
    "\n",
    "    return df_k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<__main__.Matrix_Factorization at 0x7f07df0c6c50>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "als = Matrix_Factorization(n_iters = 15, k = 36, _lambda = 0.01)\n",
    "als.fit(train, test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEICAYAAABYoZ8gAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAj+ElEQVR4nO3deZwdVZn/8c+3l6SzkQAJELKQ4LAKIYGAgoLBBVFQHB0cEBwZdBgQCOISQVBRGUVlXFBHREBQHHQEGVB/IoggmQkCCSIkLMMMSaAhkk4gKyTp5fn9UXXDTaf79u3uW7e6+37fr1e97qlTdasems5T1adOnaOIwMzMakdd3gGYmVl1OfGbmdUYJ34zsxrjxG9mVmOc+M3MaowTv5lZjcks8Uu6VtJKSYuL6r4u6QlJj0i6RdK4rM5vZmZdU1b9+CUdBWwAfhwRB6R1xwB/iIg2SV8FiIhP93Ss8ePHx7Rp0zKJ08xsqFq0aNGqiJjQub4hqxNGxL2SpnWqu6No9U/A35VzrGnTprFw4cIKRmdmNvRJWt5VfZ5t/KcDv83x/GZmNSmXxC/pIqAN+GmJfc6QtFDSwpaWluoFZ2Y2xFU98Us6DTgeOCVKPGCIiKsiYnZEzJ4wYbsmKjMz66PM2vi7IulYYB7wpoh4uZrnNrPa0traSnNzM5s2bco7lMw1NTUxefJkGhsby9o/s8Qv6UZgDjBeUjPweeBCYDhwpySAP0XEmVnFYGa1q7m5mTFjxjBt2jTSfDMkRQSrV6+mubmZ6dOnl/WdLHv1nNxF9TVZnc/MrNimTZuGfNIHkMTOO+9Mb56FDv03d1tb847AzHIy1JN+QW//O4du4t+0CS68EGbOTMpmZgYM5cRfXw+33QaPPQZf+1re0ZhZjVm9ejUzZ85k5syZ7LbbbkyaNGnr+pYtW3r8/j333MOCBQsyiW3oJv7GRvje95LyV74CTz+dbzxmVlN23nlnHn74YR5++GHOPPNMzj///K3rw4YN6/H7Tvx9NWcOnHJK0tQzdy54fmEzy9GiRYt405vexCGHHMLb3/52VqxYAcAVV1zB/vvvz4wZMzjppJNYtmwZV155Jd/85jeZOXMm8+fPr2gcVe3Hn4vLL4df/Qp+85uk6eeEE/KOyMyqLauHvL24mYwIzj33XG699VYmTJjAz3/+cy666CKuvfZaLrvsMpYuXcrw4cNZs2YN48aN48wzz2T06NF88pOfrHjYQ/uOH2C33eBLX0rK550HL/u9MTOrvs2bN7N48WLe9ra3MXPmTC699FKam5sBmDFjBqeccgo33HADDQ3Z348P/cQP8NGPJr17li+Hf/mXvKMxs2qLyGbpVQjBa1/72q3t/I8++ih33JEMWPyb3/yGs88+m4ceeohDDz2Utra2LH4KW9VG4m9ogH/7t6T89a/Dk0/mG4+Z1Zzhw4fT0tLCfffdByRDSixZsoSOjg6effZZjj76aL761a+ydu1aNmzYwJgxY1i/fn0msdRG4gc4/HD48IeTF7rOOccPes2squrq6rjpppv49Kc/zUEHHcTMmTNZsGAB7e3tnHrqqRx44IHMmjWLuXPnMm7cON71rndxyy23ZPJwN7MZuCpp9uzZUZGJWFatgn32gRdfhJ//HN7//v4f08wGpMcff5z99tsv7zCqpqv/XkmLImJ2531r544fYPz4pE8/wPnnQ0Z/RpmZDWS1lfgBPvIROOwweP55+MIX8o7GzKzqai/x19UlD3ol+Na34NFH847IzKyqai/xAxxyCJx1FrS3w9ln+0GvmdWU2kz8AJdeChMmwPz58JOf5B2NmVnV1G7i33HHpE8/wKc+BWvW5BqOmVm11G7iB/iHf4Ajj4SVK+Hii/OOxsyGkP4My7xw4ULmzp2bWWy11Y+/K48+CrNmJe38Dz4IBx+czXnMrKoGUj/+Sy65ZLsB19ra2io6Lo/78ffGgQcmg7d1dCRj+nR05B2RmQ1Rp512GmeeeSave93rmDdvHg888ACHH344s2bN4ogjjuDJdDiZe+65h+OPPx5ILhqnn346c+bMYc899+SKK67odxyZDQMn6VrgeGBlRByQ1p0IXALsBxwWERndxvfSJZfAz34G998P11wD//RPeUdkZhWke+7J5LgxZ06vv9Pc3MyCBQuor69n3bp1zJ8/n4aGBn7/+9/zmc98hptvvnm77zzxxBPcfffdrF+/nn322YezzjqLxsbGPsed5R3/dcCxneoWA+8F7s3wvL03Zgx84xtJ+YILkqEdzMwycOKJJ1JfXw/A2rVrOfHEEznggAM4//zzWbJkSZffOe644xg+fDjjx49nl1124YUXXuhXDJnd8UfEvZKmdap7HHo/I3xVvP/9cPXV8PvfJ5O0//CHeUdkZhXSlzvzrIwaNWpr+bOf/SxHH300t9xyC8uWLWNON3EOHz58a7m+vr7fwza7jb9Agu9+N5mr9+qr4U9/yjsiMxvi1q5dy6RJkwC47rrrqnbeAZv4JZ0haaGkhS0tLdU56T77QOGp+0c/ChlPhmBmtW3evHlceOGFzJo1K/PJV4pl2p0zber5deHhblH9PcAny324m2l3zs42boT994dnnoErroBzz63Oec2sogZSd85qcHfO/hg1Cr797aR88cXw17/mG4+ZWYVllvgl3QjcB+wjqVnShyX9raRm4HDgN5J+l9X5++WEE+Cd74R165LhHMzMhpDMEn9EnBwREyOiMSImR8Q1EXFLWh4eEbtGxNuzOn+/SPCd70BTE9xwA/zxj3lHZGZ9MBhGJqiE3v53uqmnO3vumXTrhGTo5tbWfOMxs15pampi9erVQz75RwSrV6+mqamp7O9k1o9/SJg3D378Y1iyJGn3Lxpnw8wGtsmTJ9Pc3EzVegXmqKmpicmTJ5e9vwdp68ntt8M73pE89H3iCejFD9fMLE/u1dNXxx4L731v0s3z4x/POxozs35z4i/Ht74FI0fCL34Bd9yRdzRmZv3ixF+OKVPgc59LyuecA5s35xuPmVk/OPGX6/zzYb/94Kmn4PLL847GzKzPnPjLNWwYfO97SfnSS2Hp0nzjMTPrIyf+3jj6aDj5ZNi0KXmz96KLkl4/69blHZmZWdncnbO3VqyAQw5JPgvq6uCgg5KJ2wvLrrvmF6OZGd1353Ti74t165JhHObPT5aFC7cfwnmvvba9EOy5ZzIUhJlZlTjxZ+nll5P5egsXgvvuS/r9F9t9d3jjG1+9EBx4YPKXgplZRpz4q6m1FR5++NULwX/91/bz+I4bB294w6sXgtmzkwfIZmYV4sSfp4hkuIfChWD+fFi+fNt9mprgsMOSJqEJE7pedtklGTrCzKwMTvwDzbPPbnshWLKkvO+NGNH9haFwcSheHzPGzxbMapQT/0C3enXynOD556Glpftl06beHXfYMBg/PvlLYdSoZOiJkSNfLXf+7E3diBG+qJgNYN0l/rKGZZY0DNg7XX0yIjw4faXtvHPybkApEclD45YWWLmy9AWisGzcmFxMstLYmFxcKrk0NkJ9PTQ0JEt/ysXrdXVJua6u+3JP24vLvujZINVj4pc0B7geWAYImCLpQxFxb6aR2fYkGD06WaZPL+87L78ML76YXABefvnVz+JyV3XlbH/lleRBdmvr9r2YakV9ffL/pXAhKFwcelvX1Xap56Uv+0HX+/S3vni9N+Vyt2f52V25p+3d7dvb9VLb5sxJhoavoHLu+P8VOCYinkzi0d7AjcAhFY3EslFolslCRPL+wpYtlVs2b06OWVja27su92ZbYenoeHVpb9/2sy9lSNbNsiTlkvgbC0kfICL+R1JjRaOwwUlKmmUaG2uzt1HhQhCRLB0dr34Wl8ut66pczlLuvoWLVVfb+ltfvN6bcrnbs/zsrtzT9u727e16T/u+7nVUWjmJf6Gkq4Eb0vVTgCH+pNWsDIUmGrNBppzf2rOAx4C56fJYWleSpGslrZS0uKhuJ0l3Snoq/dyxr4GbmVnf9Jj4I2JzRHwjIt6bLt+MiHJmIrkOOLZT3QXAXRGxF3BXum5mZlXUbVOPpP+IiPdLehTYrrN/RMwodeCIuFfStE7VJwBz0vL1wD3Ap3sRr5mZ9VOpNv7z0s/jK3i+XSOiMJ7xX4Fuxy6WdAZwBsDUqVMrGIKZWW3rNvEXEnRELO9un/6IiJDU7WvDEXEVcBUkb+5mEYOZWS0q1dSzni6aeAoiYoc+nO8FSRMjYoWkicDKPhzDzMz6odQd/xgASV8CVgA/IXlz9xRgYh/PdxvwIeCy9PPWPh7HzMz6qJzunO+OiH+LiPURsS4ivk/ykLYkSTcC9wH7SGqW9GGShP82SU8Bb03Xzcysisp5gWujpFOAn5E0/ZwM9DgwS0Sc3M2mt5QfnpmZVVo5d/wfAN4PvJAuJ6Z1ZmY2CPV4xx8RyyijacfMzAaHHu/4Je0t6a7C0AuSZki6OPvQzMwsC+U09fwQuBBoBYiIR4CTsgzKzMyyU07iHxkRD3Sqa8siGDMzy145iX+VpNeQvswl6e9I+vWbmdkgVE53zrNJhk7YV9JzwFLg1EyjMjOzzJTTq+dp4K2SRgF1EbE++7DMzCwrpcbq+Xg39QBExDcyisnMzDJU6o7/cuBh4LfAZpJxeszMbJArlfhnkQzPcBywCLiRZPYsD5FsZjaIddurJyL+EhEXRMRM4BqSt3cfk/TuagVnZmaVV86buxNI7v4PBJrxGPpmZoNaqYe7p5MMztYE3AS8PyKc9M3MBrlSbfxXA4uB5cDbgWMKPXoAIsJNPmZmg1CpxH901aIwM7OqKTX14h+rGYiZmVVHOWP1mJnZEOLEb2ZWY8pO/JJGSxqdZTBmZpa9cvrxHyjpz8ASkhe4Fkk6oD8nlXSepMWSlkj6WH+OZWZmvVPOHf8PgI9HxB4RMRX4BMkwzX2SXjT+CTgMOAg4XtLf9PV4ZmbWO+Uk/lERcXdhJSLuAUb145z7AfdHxMsR0Qb8EXhvP45nZma9UE7if1rSZyVNS5eLgaf7cc7FwJGSdpY0EngnMKXzTpLOkLRQ0sKWlpZ+nM7MzIqVk/hPByYAvwRuBsandX0SEY8DXwXuAG4nGfq5vYv9roqI2RExe8KECX09nZmZdVLODFwvAXMredKIuIZkxE8kfZlk8DczM6uCcnr13ClpXNH6jpJ+15+TStol/ZxK0r7/7/05npmZla+cydbHR8SawkpEvFRI3P1ws6SdgVbg7OLjm5lZtspJ/B2SpkbEMwCS9gD6NQtXRBzZn++bmVnflZP4LwL+S9IfSebdPRI4I9OozMwsM+U83L1d0sHA69Oqj0XEqmzDMjOzrHT7cFfSvunnwcBU4Pl0mZrWmZnZIFTqjv/jJE06/9rFtgDenElEZmaWqVITsRTa8d8REZuKt0lqyjQqMzPLTDlv7i4os87MzAaBbu/4Je0GTAJGSJpF0qMHYAdgZBViMzOzDJRq4387cBowGfhGUf164DMZxmRmZhkq1cZ/PXC9pPdFxM1VjMnMzDJUTj/+myUdB7wWaCqq/2KWgZmZWTa6fLgr6ThJE9PylcAHSGbNqgNOBPaoWoRmZlZR3fXqWQpcI+kw4IiIOAVYERGXAIcDe1cpPjMzq7AuE39EPAa8C1gHvJJWt0maQDKi5sTqhGdmZpVW6uFuO/CEpF+n4/F/G3iEpFvnNdUJz8zMKq3kw11JdcBd6Xj5t0r6LTAiItZWIzgzM6u8km/uRkQH8L2i9S1O+mZmg1s5QzbcJel9ktTzrmZmNtCVk/j/GfgFsFnSOknrJa3LOC4zM8tIOS9wjalGIGZmVh2lBmkrOdlKRDxU+XDMzCxrpe74CxOwNAGzgb+QdOWcASwkeZGrTySdD3yEZEKXR4F/7Dzmv5mZZaPbNv6IODoijgZWAAdHxOyIOASYBTzX1xNKmgTMBWZHxAFAPXBSX49nZma9U87D3X0i4tHCSkQsBvbr53kbSMb5byAZ2//5fh7PzMzKVE7if0TS1ZLmpMsPSd7g7ZOIeA64HHiG5K+JtRFxR+f9JJ0haaGkhS0tLX09nZmZdVJO4v9HYAlwXro8ltb1iaQdgROA6cDuwChJp3beLyKuSpuXZk+YMKGvpzMzs07K6c65CfhmulTCW4GlEdECIOmXwBHADRU6vpmZldBj4pe0F/AVYH+2nYhlzz6e8xng9ZJGkoz8+RaSXkJmZlYF5TT1/Aj4PtAGHA38mH7cnUfE/cBNwEMkXTnrgKv6ejwzM+udchL/iIi4C1BELE8nYzmuPyeNiM9HxL4RcUBEfDAiNvfneGZmVr4em3pIxuipA56SdA5JH/7R2YZlZmZZKeeO/zySvvZzgUOADwIfyjIoMzPLTjm9eh5MixvoRzdOMzMbGMrp1XM3yZg624iIN2cSkZmZZaqcNv5PFpWbgPeR9PAxM7NBqJymnkWdqv5b0gMZxWNmZhkrp6lnp6LVOpIHvGMzi8jMzDJVTlPPIpI2fpE08SwFPpxlUGZmlp1ymnqmVyMQMzOrjm778Us6VdIHu6j/oKQPZBuWmZllpdQLXOcCt3RR/0vgE9mEY2ZmWSuV+BsjYkPnyojYCDRmF5KZmWWpVOIfIWlU50pJY4Bh2YVkZmZZKpX4rwFukrRHoULSNOBn6TYzMxuEuu3VExGXS9oA3CupMBrnBuCyiPh+VaIzM7OKK9mdMyKuBK5Mm3eIiPVVicrMzDJTzgtcTvhmZkNIOePxm5nZEFIy8Uuqk3REtYIxM7PslUz8EdEBfK9KsZiZWRWU09Rzl6T3SVIlTihpH0kPFy3rJH2sEsc2M7OelfNw95+BjwPtkl4hGaUzImKHvpwwIp4EZgJIqieZvL2roSHMzCwD5YzOOSbD878F+L+IWJ7hOczMrEhZ3TklvRs4Kl29JyJ+XaHznwTcWKFjmZlZGXps45d0GXAe8Fi6nCfpK/09saRhwLuBX3Sz/QxJCyUtbGlp6e/pzMwspYgovYP0CDAz7eFTaJf/c0TM6NeJpROAsyPimJ72nT17dixcuLA/pzMzqzmSFkXE7M715b7ANa6oXKn5dk/GzTxmZlVXThv/l4E/S7qbpEfPUcAF/TlpOtzz20h6DJmZWRWVTPyS6oAO4PXAoWn1pyPir/05aTqZy879OYaZmfVNT6NzdkiaFxH/AdxWpZjMzCxD5bTx/17SJyVNkbRTYck8sgpY29bGLe4RZGa2jXLa+P8+/Ty7qC6APSsfTuVsbG9n7/vvZ1VrK0sOPZR9R203i6SZWU3qcXRO4IKImN5pGdBJH2BUfT3vGT+eDuCSZcvyDsfMbMAoZ3TOT1Uploq7eI89GCbx85YWHtmwIe9wzMwGhCHdxj+lqYmzdt8dgM8tXZpzNGZmA0M5if/vSdr37wUWpcugeY32gqlTGVlXx62rV/PAunV5h2NmlrseE38X7fuDoo2/YLfhwzl30iQAPuu7fjOz7hO/pHlF5RM7bftylkFV2rypU9mhvp47XnqJe9esyTscM7NclbrjP6mofGGnbcdmEEtmdmps5ONTpgBw8dKl9DQwnZnZUFYq8aubclfrA97HJk9mp4YG5q9dy50vvZR3OGZmuSmV+KObclfrA97YhgbmTZ0K+K7fzGpbqcR/UDoR+npgRlourB9Ypfgq6pxJk9i1sZEH16/nV6tX5x2OmVkuuk38EVEfETtExJiIaEjLhfXGagZZKaPq6/nMHnsASQ+fDt/1m1kNKnciliHjjIkTmTx8OI9s3MgvPICbmdWgmkv8TfX1fC696//80qW0dXTkHJGZWXXVXOIHOG233dizqYknX3mFG154Ie9wzMyqqiYTf2NdHZdMmwbAF5YvZ4vv+s2shtRk4gf4wK67st/IkSzbtIlrV6zIOxwzs6qp2cRfL/HF9K7/S8uX80p7e74BmZlVSS6JX9I4STdJekLS45IOzyOO906YwMzRo3l+yxaufP75PEIwM6u6vO74vw3cHhH7AgcBj+cRRJ3El9K7/q888wwb2tryCMPMrKqqnvgljQWOAq4BiIgtEbGm2nEUHLfzzrx+hx1oaW3lO889l1cYZmZVk8cd/3SgBfiRpD9LulpSbjOhS+LS6dMB+Nqzz7KmtTWvUMzMqiKPxN8AHAx8PyJmARuBCzrvJOkMSQslLWzJ+A3bN48bx5xx41jT1sY3mpszPZeZWd7ySPzNQHNE3J+u30RyIdhGRFwVEbMjYvaECRMyDaj4rv+bzc2s2rIl0/OZmeWp6ok/Iv4KPCtpn7TqLcBj1Y6jszeMHcs7dtqJDe3tfO3ZZ/MOx8wsM3n16jkX+KmkR4CZwICYyvFL6V3/d597jhWbN+ccjZlZNnJJ/BHxcNqMMyMi3hMRA2JKrEPGjOFvx4/nlY4OvvzMM3mHY2aWiZp9c7c7X5w2DQE/eP55lm/alHc4ZmYV58TfyQGjR3PyLrvQGsGly5fnHY6ZWcU58Xfh89OmUQ/8aMUKnnr55bzDMTOrKCf+Luw9ciQf2m032oEvLFuWdzhmZhXlxN+Nz02bRqPEv69cyZKNG/MOx8ysYpz4u7FHUxNnTJxIkEzRaGY2VDjxl/CZPfagqa6Om1et4qH16/MOx8ysIpz4S9h9+HDOmTQJgM/6rt/Mhggn/h7MmzKF0fX1/L8XX2TB2rV5h2Nm1m9O/D2YMGwYH5s8GYCLfddvZkOAE38ZPjF5MuMaGrh7zRr+8NKAGF3CzKzPnPjLMK6xkU9NmQIkd/0RkXNEZmZ958RfprmTJjG+sZH71q3jty++mHc4ZmZ95sRfptENDVw4dSqQ3PV3+K7fzAYpJ/5eOGv33dl92DD+vGEDt6xalXc4ZmZ94sTfCyPq67l4jz0AOOmxx3j9okV86v/+j1+tWsWLnqTdzAaJhrwDGGw+PHEid770EreuWsX969dz//r1XJ5O1XjAqFEcOXYsR40dy5HjxjFp+PCcozUz254GQw+V2bNnx8KFC/MOYxvr2tpYsHYt89PlgXXr2NzpZzm9qSm5EIwbx5Fjx7LXiBFIyiliM6s1khZFxOzt6p34K2NTezsPrl+/9ULw32vXsr69fZt9dm1s5I3pXwNHjR3LjNGjqfeFwMwy4sRfZe0RPLJhA/euXcv8NWuYv3YtKzs9B9ihvp4jxo7d2jx06A47MLzOj13MrDKc+HMWETz1yivcm14E5q9dy9JOc/oOk5g4bBg7NzayU2MjOzU0JOWGBnZqbNy2nH7u2NBAoy8WZtaF7hJ/Lg93JS0D1gPtQFtXgQ01kth75Ej2HjmSj+y+OwDPbd7M/DVrkr8K1q5l8caNLN+8meWbN/fq2DvU15e8UOzY0MCI+npG1NUxoq6OpvSzULd1va6O4XV1fg5hNsTl2avn6Iio6c7wk4YP56Rdd+WkXXcFYENbGytbW3mxtZUX29pYnX6+2Nq6bTn9LKyva29nXXs7yyoUV/GFoPgi0bm+qa6ORonGwqfEsKJy8bZhnfYrta0hXeqLPuth+7pCuWibL1pmPXN3zgFkdEMDoxsa2HPEiLK/0xHBura2kheKl9raeKWjg1c6OtjU0cEr7e3brhfVbYlgU1o/GIejE50uELC1XJeu10nUpfWFcqlt9UX7dLVNRdvqCutFder0/S737+L7Ktq/u3Jh/+7K3W0vHKd4fZvtJfbt6TjlrlPB71Bm3db6Evt1Psc2652+392+9LA/vfjOyLo6RjdUNlXnlfgDuENSAD+IiKs67yDpDOAMgKnpUAm2vTqJcY2NjGts7NUFozsdaeIvXAy2lru4SGzu6KA1YuuypXi9zG1bOu3X2tFBW0ShDZD2dNlaLqrfbjvJL1bhWGZDwaenTOGy17ymosfMK/G/MSKek7QLcKekJyLi3uId0ovBVZA83M0jyFpUJzGyvp6R9fXQ2Jh3OL3W0c0FoS2Cjgg60n3a03J7F3Ud6Xc6ivfpZv92kgf3AVvrO9K6bdaLyh2F/Tuvd1EXZZR73F4UTxTFG12sd5TYVrze0cNxelqngt+hzLpCR5ZS+3U+ByW+392+9LA/vfzOqPp6Ki2XxB8Rz6WfKyXdAhwG3Fv6W2Y9q0ufGZhZ96reD1DSKEljCmXgGGBxteMwM6tVedzx7wrckj7UaAD+PSJuzyEOM7OaVPXEHxFPAwdV+7xmZpbwK59mZjXGid/MrMY48ZuZ1RgnfjOzGuPEb2ZWYwbFsMySWoDlffz6eGAwDQY3mOIdTLHC4Ip3MMUKgyvewRQr9C/ePSJiQufKQZH4+0PSwsE07PNgincwxQqDK97BFCsMrngHU6yQTbxu6jEzqzFO/GZmNaYWEv92Qz4PcIMp3sEUKwyueAdTrDC44h1MsUIG8Q75Nn4zM9tWLdzxm5lZkSGd+CUdK+lJSf8r6YK84+mOpCmS7pb0mKQlks7LO6aeSKqX9GdJv847lp5IGifpJklPSHpc0uF5x1SKpPPT34PFkm6U1JR3TAWSrpW0UtLiorqdJN0p6an0c8c8YyzWTbxfT38XHpF0i6RxOYa4VVexFm37hKSQNL4S5xqyiV9SPfA94B3A/sDJkvbPN6putQGfiIj9gdcDZw/gWAvOAx7PO4gyfRu4PSL2JRkZdsDGLWkSMBeYHREHAPXASflGtY3rgGM71V0A3BURewF3pesDxXVsH++dwAERMQP4H+DCagfVjevYPlYkTSGZt+SZSp1oyCZ+klm9/jcino6ILcDPgBNyjqlLEbEiIh5Ky+tJEtOkfKPqnqTJwHHA1XnH0hNJY4GjgGsAImJLRKzJNaieNQAjJDUAI4Hnc45nq3SK1Bc7VZ8AXJ+WrwfeU82YSukq3oi4IyLa0tU/AZOrHlgXuvnZAnwTmMerszH221BO/JOAZ4vWmxnAybRA0jRgFnB/zqGU8i2SX8SOnOMox3SgBfhR2jR1dTrz24CUTkt6Ocnd3QpgbUTckW9UPdo1Ilak5b+STLY0WJwO/DbvILoj6QTguYj4SyWPO5QT/6AjaTRwM/CxiFiXdzxdkXQ8sDIiFuUdS5kagIOB70fELGAjA6spYhtp+/gJJBes3YFRkk7NN6ryRdJNcFB0FZR0EUkz60/zjqUrkkYCnwE+V+ljD+XE/xwwpWh9clo3IElqJEn6P42IX+YdTwlvAN4taRlJ89mbJd2Qb0glNQPNEVH4C+omkgvBQPVWYGlEtEREK/BL4IicY+rJC5ImAqSfK3OOp0eSTgOOB06Jgdun/TUkNwB/Sf+9TQYekrRbfw88lBP/g8BekqZLGkbygOy2nGPqkpIJiK8BHo+Ib+QdTykRcWFETI6IaSQ/0z9ExIC9I42IvwLPStonrXoL8FiOIfXkGeD1kkamvxdvYQA/jE7dBnwoLX8IuDXHWHok6ViSpsp3R8TLecfTnYh4NCJ2iYhp6b+3ZuDg9He6X4Zs4k8f3pwD/I7kH85/RMSSfKPq1huAD5LcPT+cLu/MO6gh5Fzgp5IeAWYCX843nO6lf5ncBDwEPEryb3TAvGkq6UbgPmAfSc2SPgxcBrxN0lMkf7FclmeMxbqJ97vAGODO9N/albkGmeom1mzONXD/yjEzsywM2Tt+MzPrmhO/mVmNceI3M6sxTvxmZjXGid/MrMY48duQJGlB+jlN0geqcL53D+QRYM2KuTunDWmS5gCfjIjje/GdhqJBvMyGHN/x25AkaUNavAw4Mn1R5/x0HoGvS3owHY/9n9P950iaL+k20jd7Jf2npEXp2PhnFB37WEkPSfqLpLvSutMkfTctT5P0h/T4d0mamtZfJ+kKSQskPS3p74qO+amimL6Q1o2S9Jv0PIsl/X0VfnRWAxryDsAsYxdQdMefJvC1EXGopOHAf0sqjH55MMk47UvT9dMj4kVJI4AHJd1McrP0Q+CoiFgqaacuzvkd4PqIuF7S6cAVvDpU8UTgjcC+JEMd3CTpGGAvkqHEBdwm6ShgAvB8RByXxj62Uj8Uq21O/FZrjgFmFN1tjyVJuluAB4qSPsBcSX+blqek+00A7i3sFxFdjZ9+OPDetPwT4GtF2/4zIjqAxyQVhi8+Jl3+nK6PTs81H/hXSV8Ffh0R8/vyH2zWmRO/1RoB50bE77apTJ4FbOy0/lbg8Ih4WdI9QCWmQNzcKZbC51ci4gfbBSsdDLwTuFTSXRHxxQrEYDXObfw21K0nGZCr4HfAWekw2Ejau5uJWcYCL6VJf1+SKTEhmbHpKEnT0+931dSzgFenSzyF5M69lN8Bp6fzMSBpkqRdJO0OvBwRNwBfZ2APJ22DiO/4bah7BGiX9BeSOU2/DUwjGddcJLNzvaeL790OnCnpceBJkoRPRLSkzwl+KamOZOz5t3X67rkkM359Kj3+P5YKMCLukLQfcF8SEhuAU4G/Ab4uqQNoBc7q1X+5WTfcndPMrMa4qcfMrMY48ZuZ1RgnfjOzGuPEb2ZWY5z4zcxqjBO/mVmNceI3M6sxTvxmZjXm/wMXaLnf01j3mwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(als.test_ecm_record, label = 'Test', linewidth = 2,color='r')\n",
    "plt.plot(als.train_ecm_record, label = 'Train', linewidth = 2,color='c')\n",
    "plt.xlabel('iteraciones')\n",
    "plt.ylabel('Error Cuadrático Medio')\n",
    "plt.legend(['Test','Train'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Observamos que con muy pocas iteraciones el ECM baja, por lo que en las siguientes evaluaciones utilizaremos 15 iteraciones.\n",
    "\n",
    "### Buscando la mejor K\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "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>10</th>\n",
       "      <th>71</th>\n",
       "      <th>171</th>\n",
       "      <th>271</th>\n",
       "      <th>371</th>\n",
       "      <th>471</th>\n",
       "      <th>571</th>\n",
       "      <th>671</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10.725289</td>\n",
       "      <td>10.570593</td>\n",
       "      <td>11.781813</td>\n",
       "      <td>12.72651</td>\n",
       "      <td>13.365499</td>\n",
       "      <td>13.78757</td>\n",
       "      <td>14.04776</td>\n",
       "      <td>14.170276</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         10         71         171       271        371       471       571  \\\n",
       "0  10.725289  10.570593  11.781813  12.72651  13.365499  13.78757  14.04776   \n",
       "\n",
       "         671  \n",
       "0  14.170276  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.random.seed(20201117)\n",
    "K_list = [10,71,171,271,371,471,571,671]\n",
    "df_k = best_K(train, test, K_list)\n",
    "df_k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "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>1</th>\n",
       "      <th>10</th>\n",
       "      <th>20</th>\n",
       "      <th>30</th>\n",
       "      <th>40</th>\n",
       "      <th>50</th>\n",
       "      <th>60</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12.277179</td>\n",
       "      <td>10.765336</td>\n",
       "      <td>10.470641</td>\n",
       "      <td>10.382057</td>\n",
       "      <td>10.337564</td>\n",
       "      <td>10.445505</td>\n",
       "      <td>10.476118</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          1          10         20         30         40         50         60\n",
       "0  12.277179  10.765336  10.470641  10.382057  10.337564  10.445505  10.476118"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "K_list2 = [1,10,20,30,40,50,60]\n",
    "df_k2 = best_K(train, test, K_list2)\n",
    "df_k2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEGCAYAAABy53LJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAsP0lEQVR4nO3deXyddZ33/9c7W9M2SdM2C90XKG0KQoGIKIu0CYrosCmOuKEywzg3jjjKODr+5uE4t97jOCh6j9uNgDC33rigDCi4tJVNZbFIgdKFlrK0tE3SPd3Spvn8/jhXyiFkOSk5OTnJ+/l4XI9zru91XSefq4R8zvVdFRGYmZllqiDXAZiZWX5x4jAzs35x4jAzs35x4jAzs35x4jAzs34pynUAg6GqqipmzpyZ6zDMzPLKY489tjUiqruWj4jEMXPmTJYtW5brMMzM8oqkF7ord1WVmZn1ixOHmZn1ixOHmZn1ixOHmZn1ixOHmZn1ixOHmZn1ixOHmZn1ixNHL+5b08y371uX6zDMzIaUrCUOSTdLapa0Iq3sPyStlvSkpDskVXZz3TRJ90paKelpSdekHfsXSS9JWp5sF2QrfoCHnt3G9YufofXAoWz+GDOzvJLNJ45bgPO7lC0GToyIk4BngM92c1078KmImA+cAVwtaX7a8esjYkGy3ZOFuI9oqKvl0OHgwbVbs/ljzMzyStYSR0Q8AGzvUvbbiGhPdh8GpnZz3eaI+HPyvhVYBUzJVpy9OXV6JZVjilmyqikXP97MbEjKZRvHR4Bf9XaCpJnAKcAjacUfS6q6bpY0vpdrr5K0TNKylpaWowqwqLCAhXNruG9NC4c7vMSumRnkKHFI+hypKqkf9nJOGfAz4BMRsTsp/g5wLLAA2Ax8tafrI+KGiKiPiPrq6ldN7pixhroatu89yOMv7jjqzzAzG04GPXFI+hDwDuB9EdHt13hJxaSSxg8j4ued5RHRFBGHI6ID+B5werbjPef4aooKxGJXV5mZAYOcOCSdD3wauDAi9vVwjoCbgFUR8bUuxyal7V4CrCDLKkqLecPsCSxd1ZztH2Vmlhey2R33NuAhYK6kjZKuBL4JlAOLk+60303OnSyps4fUmcAHgEXddLv9iqSnJD0JLAT+Plvxp2usq2Vd8x5e2LZ3MH6cmdmQlrWFnCLi8m6Kb+rh3E3ABcn73wPq4bwPDFiA/dBYV8sXfrGSJauaufKsWbkIwcxsyPDI8QxMmzCG42vLWOp2DjMzJ45MNdTV8uhz29ntUeRmNsI5cWSosa6G9o7g/jVHNybEzGy4cOLI0IJp45kwtsSjyM1sxHPiyFBhgY6MIm8/3JHrcMzMcsaJox/Om1/Drv2HWPaCR5Gb2cjlxNEPZ8+ppqSwwL2rzGxEc+Loh7Gjijjj2IkeRW5mI5oTRz811tWwfute1rfsyXUoZmY54cTRT4vm1QD4qcPMRiwnjn6aOn4M844p92y5ZjZiOXEchca6Wh57YQc79x3MdShmZoPOieMoNM6v5XBHcJ9HkZvZCOTEcRROmjKO6vJRHkVuZiOSE8dRKCgQi+bWcP8zLRzyKHIzG2GcOI5SQ10NrQfa+dNz23MdipnZoMpq4pB0s6RmSSvSyv5D0mpJT0q6Q1JlD9eeL2mNpHWSPpNWPkvSI0n5jyWVZPMeenLWnCpKigpY4m65ZjbCZPuJ4xbg/C5li4ETI+Ik4Bngs10vklQIfAt4GzAfuFzS/OTwvwPXR8RxwA7gyuyE3rsxJUWceexElqxqIiJyEYKZWU5kNXFExAPA9i5lv42I9mT3YWBqN5eeDqyLiPURcRD4EXCRJAGLgNuT824FLs5G7JloqKvlxe37WNfsUeRmNnLkuo3jI8CvuimfAmxI29+YlE0EdqYlns7yV5F0laRlkpa1tGSn22xDXWoUuaurzGwkyVnikPQ5oB34YTY+PyJuiIj6iKivrq7Oxo9g0rjRnDilwrPlmtmIkpPEIelDwDuA90X3DQQvAdPS9qcmZduASklFXcpzpmFeLX9+cQfb93oUuZmNDIOeOCSdD3wauDAi9vVw2p+AOUkPqhLgPcBdSZK5F3hXct4VwJ3Zjrk3jXW1dATcu9rVVWY2MmS7O+5twEPAXEkbJV0JfBMoBxZLWi7pu8m5kyXdA5C0YXwM+A2wCvhJRDydfOw/Ap+UtI5Um8dN2byHvpw4pYLaCo8iN7ORo6jvU45eRFzeTXG3f+gjYhNwQdr+PcA93Zy3nlSvqyFBEovm1XLX8pdoaz/MqKLCXIdkZpZVue5VNSw01tWw9+BhHlnvUeRmNvw5cQyAM4+rorTYa5Gb2cjgxDEASosLOeu4apasavYocjMb9pw4BkhjXQ0v7dzPmqbWXIdiZpZVThwDxGuRm9lI4cQxQGoqSjl56jgWr3Q7h5kNb04cA6ihrpYnNu6kpbUt16GYmWWNE8cAaqirITyK3MyGOSeOATR/UgWTx5V6FLmZDWsZJQ5JJZJOTLbibAeVryTRUFfLg2u3cuDQ4VyHY2aWFX0mDknnAmtJrcj3beAZSedkN6z81VBXw/5Dh3lo/bZch2JmlhWZPHF8FXhLRLw5Is4B3gpcn92w8tcZsycypqTQo8jNbNjKJHEUR8Sazp2IeAZwdVUPSosLOXtOFUs9itzMhqlMEscySTdKOjfZvgcsy3Zg+ayhrpbNuw7w9KbduQ7FzGzAZZI4/hZYCXw82VYmZdaDRfNqkDyK3MyGpz4TR0S0RcTXIuLSZLs+IjzCrRdVZaNYMK2SpavdzmFmw0+PiUPST5LXpyQ92XUbvBDzU2NdLU9u3EXT7gO5DsXMbED19sRxTfL6DuAvutl6JelmSc2SVqSVXSbpaUkdkup7uG5usqRs57Zb0ieSY/8i6aW0Yxd09xlDQWNdLQC/8yhyMxtmekwcEbE5eX2huy2Dz74FOL9L2QrgUuCBXn7umohYEBELgNOAfcAdaadc33k8WV52SDq+toyp40e7W66ZDTs9rjkuqRXosT9pRFT09sER8YCkmV3KViWfnWl8DcCzGSaqIUUSjXW13Pboi+w/eJjRJV6L3MyGh96eOMqT5PAN4DPAFGAq8I/A1wclOngPcFuXso8l7Sw3Sxrf04WSrpK0TNKylpaW7EbZg4a6GtraO/jDuq05+flmZtmQSXfcCyPi2xHRGhG7I+I7wEXZDkxSCXAh8NO04u8AxwILgM2kRrV3KyJuiIj6iKivrq7OZqg9esOsiZSNKnLvKjMbVjJJHHslvU9SoaQCSe8D9mY7MOBtwJ8j4shf3YhoiojDEdEBfA84fRDiOGolRQWcc3xqFHlHh0eRm9nwkEnieC/wbqAp2S5LyrLtcrpUU0malLZ7CanG9iGtsa6W5tY2VmzaletQzMwGRI+N450i4nmOompK0m3AuUCVpI3A54HtwH8C1cDdkpZHxFslTQZujIgLkmvHAucBf9PlY78iaQGpRvvnuzk+5CycW0OBYMmqZk6aWpnrcMzMXrM+E4ek40m1LdRGxImSTiLV7vHF3q6LiMt7OHRH14KI2ARckLa/F5jYzXkf6CveoWb82BJOmzGepaua+OR5x+c6HDOz1yyTqqrvAZ8FDgFExJOkejtZhhrqanl602427dyf61DMzF6zTBLHmIh4tEtZezaCGa4a62oAWOpR5GY2DGSSOLZKOpZkMKCkd5HqCmsZOra6jBkTx3gUuZkNC322cQBXAzcA8yS9BDwHvD+rUQ0zkmiYV8sPHnmBfQfbGVOSyT+7mdnQlMm06usjopFUT6h5EXFW0tPK+qFxfg0H2zt4cK1HkZtZfuttrqpP9lAOQER8LUsxDUuvnzmB8tIilq5q4q0nHJPrcMzMjlpvdSbXAcuBXwFtQMYzE9qrFRcWcO7cGn63OjWKvKDA/5xmlp96SxynkBq9/XbgMVKjuJdGhOfOOEqNdTX84olNLN+4k1On9zg/o5nZkNbb7LhPRMRnknUxbiI1enylpAsHK7jh5tzjaygskHtXmVle67NxXFI1qaeP1wEbAQ9GOErjxhRTP2M8S1f5n9DM8ldva45/RNKvSU1rLuDdEXFeRDw8aNENQ411taze0srGHftyHYqZ2VHp7YnjRmAy0Aq8FbhR0l2d26BENww1zk+tRe6nDjPLV701ji8ctChGkFlVY5ldPZYlq5q44k0zcx2OmVm/9Zg4IuL+wQxkJGmsq+X7f3iO1gOHKC8tznU4Zmb9kslcVTbAGubVcOhweBS5meUlJ44cOG3GeMaNLmaJu+WaWR7KOHFIKpNU1o/zb5bULGlFWtllkp6W1CGpvpdrn5f0lKTlkpallU+QtFjS2uQ1L0fRFRUWsHBuNfetaeGw1yI3szyTyTiO10l6HHia1ADAxySdmMFn3wKc36VsBXAp8EAG1y+MiAURkZ5gPkNq9PocYGmyn5ca6mrZvvcgj7+4I9ehmJn1SyZPHP8H+GREzIiI6cCnSE2z3quIeIDUGuPpZasiYs1RRZpyEXBr8v5W4OLX8Fk59ea51RQViCXulmtmeSaTxDE2Iu7t3ImI+4CxWYso+THAb5Onm6vSymsjonMRqS1AbU8fIOkqScskLWtpaclmrEelorSYN8ye4OlHzCzvZJI41kv6Z0kzk+3/A9ZnOa6zIuJU4G3A1ZLO6XpCMtlijw0EEXFDRNRHRH11dXUWQz16DfNqWdu8hxe27c11KGZmGcskcXyE1CJOPwd+BlQlZVkTES8lr83AHcDpyaEmSZMAkte8rudprEs9MLm6yszySSYrAO6IiI9HxKkRcVpEfCIistaiK2mspPLO98BbSDWqA9wFXJG8vwK4M1txDIbpE8cwp6bM1VVmllcy6VW1WFJl2v54Sb/J4LrbgIeAuZI2SrpS0iWSNgJvBO7u/BxJkyXdk1xaC/xe0hPAo8DdEfHr5NiXgfMkrQUak/281lBXy6PPbWf3gUO5DsXMLCO9zVXVqSoidnbuRMQOSTV9XRQRl/dw6I5uzt0EXJC8Xw+c3MNnbgMaMog5bzTW1fDd+5/l/jUt/MXJk3MdjplZnzJp4+iQNL1zR9IMemmUtv45Zfp4JowtcXWVmeWNTJ44Pkeq6uh+UutynA1c1fsllqnCArFwbg1LVjXRfriDokLPAmNmQ1smjeO/Bk4Ffgz8CDgtIvps47DMNdbVsGv/IZa94FHkZjb09bYC4Lzk9VRgOrAp2aYnZTZAzj6+mpLCAldXmVle6K2q6pOkqqS+2s2xABZlJaIRqGxUUTKKvJnPvX1+rsMxM+tVbws5dbZjvC0iDqQfk1Sa1ahGoMa6Wj5/19Osb9nD7OqMJyE2Mxt0mbTE/jHDMnsNGupSPZy9FrmZDXW9tXEcI+k0YLSkUySdmmznAmMGK8CRYur4Mcw7ptyLO5nZkNdbG8dbgQ8BU4GvpZW3Av+UxZhGrMa6Wr5z/7Ps3HeQyjEluQ7HzKxbPT5xRMStEbEQ+FBELEzbLoyInw9ijCNGQ10NhzuC+9YMvWngzcw69TkAMCJ+JuntwAlAaVr5v2YzsJHo5KmVVJWNYsmqJi4+ZUquwzEz61a3TxyS3p42ffl3gfcCf52cfxkwY9AiHEEKCsSiedXc/0wLhw535DocM7Nu9VRV9Rxwk6TTgTdFxPuAzRHxL6Rmtj1+kOIbcRrqamk90M6fntve98lmZjnQbeKIiJXAXwC7gf1JcbukauAQMGlwwht5zp5TRUlRgRd3MrMhq7fG8cMRsRr4ZbIexzeAJ4EXSM1bZVkwpqSIM4+dyNLVTaRWxzUzG1p6HQAoqQBYGhE7I+JOUm0bcyPic4MS3QjVUFfLC9v28WzLnlyHYmb2Kr0mjojoAL6Vtn8wInZlPaoRrnMU+eKVrq4ys6EnkylHlkp6pyT154Ml3SypWdKKtLLLJD0tqUNSfQ/XTZN0r6SVybnXpB37F0kvSVqebBf0J6Z8MWncaE6YXOHZcs1sSMokcfwN8FOgTdJuSa2Sdmdw3S3A+V3KVgCXAg/0cl078KmImA+cAVwtKX3K2OsjYkGy3dP9R+S/hrpa/vziDrbvPZjrUMzMXiGThZzKI6IgIkoioiLZr8jgugeA7V3KVkXEmj6u2xwRf07etwKrgBE3Gq6xroaOgHtXu7rKzIaWHkeO97VYU+cf92ySNBM4BXgkrfhjkj4ILCP1ZNLtsnmSriJZ4nb69OndnTKknTh5HDXlo1i6uol3njY11+GYmR3R25QjnQs4lQL1wBOk1hw/idQf7TdmMzBJZcDPgE9ERGfV2HeA/0lqIan/mcT4ke6uj4gbgBsA6uvr865fa0GBaKir5RdPbOJgewclRV6L3MyGht7GcSxMJjncDJwaEfURcRqpJ4CXshmUpGJSSeOH6RMqRkRTMr6kA/gecHo248i1xroa9rS188hz23IdipnZEZl8jZ0bEU917kTECqAuWwElvbduAlZFxNe6HEsfsX4Jqcb2YevM46ooLS5gyUr3rjKzoSOTxPGkpBslnZts3yM1grxXkm4DHgLmStoo6UpJl0jaSKqa625Jv0nOnSyps4fUmcAHgEXddLv9iqSnJD0JLAT+vn+3m19Kiws567gqlqxq9ihyMxsy+pxWHfgw8LdA53iKB0i1NfQqIi7v4dAd3Zy7Cbggef97Um0p3X3mBzKId1hpqKtlyapm1jS1Mu+YPjuzmZllXSbrcRwArk82G2QN815ei9yJw8yGgj6rqiTNkXR7MpJ7fec2GMEZ1FSUctLUcV6L3MyGjEzaOL5PqmqqnVS7wn8BP8hmUPZKjXW1LN+wk5bWtlyHYmaWUeIYHRFLAUXEC8liTm/PbliWrqGuhvAocjMbIjJJHG3J9OprJX1M0iVAWZbjsjTzJ1UweVypq6vMbEjIJHFcA4wBPg6cRqqr7BXZDMpeSRKL6mp4cO1WDhw6nOtwzGyEy2SSwz9FxJ6I2BgRH46ISyPi4cEIzl7WUFfL/kOHeWi9R5GbWW712R1X0r2k5oZ6hYhYlJWIrFtvnD2RMSWFLF3VxMK5NbkOx8xGsEwGAF6b9r4UeCepHlY2iDpHkf9uVTNxUdDPdbXMzAZMJgMAH+tS9AdJj2YpHutF4/xafruyiZWbd3PC5HG5DsfMRqhMqqompO0WkGog91+tHFg0rwYJlqxsduIws5zJpKrqMVJtHCJVRfUccGU2g7LuVZWNYsG0SpaubuKaxjm5DsfMRqhMqqpmDUYglpnGulr+4zdraNp9gNqK0lyHY2YjUI/dcSW9X9KrZqOV9AFJ781uWNaThrpUj6rfeRS5meVIb+M4/o5upkAHfg58KjvhWF/m1pYzpXI0Sz2K3MxypLfEURwRe7oWRsReoDh7IVlvJNFYV8Pv13kUuZnlRm+JY7SksV0LJZUDJZl8uKSbJTVLWpFWdpmkpyV1SKrv5drzJa2RtE7SZ9LKZ0l6JCn/saSMYhlOGufXcuBQB39YtzXXoZjZCNRb4rgJuF3SjM4CSTOBHyXHMnELcH6XshXApaRWEuyWpELgW8DbgPnA5ZLmJ4f/Hbg+Io4DdjACe3i9YdZEykYVedJDM8uJHhNHRFwH3Ak8IGmbpG3A/cAvI+I/MvnwiHgA2N6lbFVErOnj0tOBdRGxPiIOkkpWFyk1XHoRcHty3q3AxZnEMpyUFBVwzvFVLF3VTEeH1yI3s8HV6ySHEfHdiJgBzARmRsSMiOhzvfEBMAXYkLa/MSmbCOyMiPYu5a8i6SpJyyQta2lpyWqwudAwr5bm1jZWbNqV61DMbITJZFp1IqI1IlqzHcxAiogbIqI+Iuqrq6tzHc6AWzivhgLBklXulmtmgyujxJEDLwHT0vanJmXbgEpJRV3KR5wJY0s4dfp4d8s1s0HXa+KQVCDpTYMVTJo/AXOSHlQlwHuAuyIigHuBdyXnXUGqHWZEaqir5elNu9m8a3+uQzGzEaSvNo4OUr2bjoqk24CHgLmSNkq6UtIlkjYCbwTulvSb5NzJku5Jfm478DHgN8Aq4CcR8XTysf8IfFLSOlJtHpn28Bp2zpufGkW+1NVVZjaIMpnkcKmkdwI/T77xZywiLu/h0KtGpEfEJuCCtP17gHu6OW89qV5XI96x1WXMmDiGJauaeP8ZM/q+wMxsAGTSxvE3wE+Bg5J2S2qVtDvLcVkGJNEwr5Y/PruNfQe9tpaZDY5M1hwvj4iCiCiOiIpkv2IwgrO+NdbVcLC9gwfXehS5mQ2OjHpVSbpQ0nXJ9o5sB2WZe/2sCZSXFrl3lZkNmj4Th6QvA9cAK5PtGkn/lu3ALDPFhQW8+fhqfre6xaPIzWxQZPLEcQFwXkTcHBE3k5p76u3ZDcv6o7Gulq172nhi485ch2JmI0CmAwAr0957sesh5ty51RQWyN1yzWxQZJI4/hfwuKRbJN1Kag3yL2U3LOuPyjEl1M8Y79lyzWxQ9DlyHOgAziC18t/PgDdGxI8HITbrh8a6WlZvaWXjjn25DsXMhrlMRo5/OiI2R8RdybZlkGKzfuhci9zVVWaWbZlUVS2RdK2kaZImdG5Zj8z6ZXZ1GbOrxrq6ysyyLpMpR/4yeb06rSyA2QMfjr0WDXU13PrHF9jT1k7ZqEz+05qZ9V8mbRyfiYhZXTYnjSGooa6Wg4c7ePCZ4bdwlZkNHZm0cfzDIMVir1H9jPGMG13sxZ3MLKvcxjGMFBUWsHBuNfeuaeawR5GbWZZkkjj+klT7xgOkxnA8BizLZlB29Brqatm+9yCPv7gj16GY2TDVZwtqRMwajEBsYLx5bjVFBWLJqmbqZ/rB0MwGXo9PHJI+nfb+si7H/lc2g7KjV1FazOmzJni2XDPLmt6qqt6T9v6zXY6d39cHS7pZUrOkFWllEyQtlrQ2eR3fzXULJS1P2w5Iujg5douk59KOLegrjpGooa6Wtc17WNfcmutQzGwY6i1xqIf33e135xZenWA+AyyNiDnA0mT/FSLi3ohYEBELgEXAPuC3aaf8Q+fxiFieQRwjzltPqGVUUQGXfOuP3Pjgeg4d7sh1SGY2jPSWOKKH993tv/riiAeA7V2KLwJuTd7fClzcx8e8C/hVRHgCpn6YOn4Md3/8bE6dMZ4v3r2K87/+APd7bIeZDZDeEsfJnWuMAycl7zv3X3eUP682IjYn77cAtX2c/x7gti5lX5L0pKTrJY3q6UJJV0laJmlZS8vI+6N5XE0Zt3z49dx0RT2HO4Irbn6Uv7r1Tzy/dW+uQzOzPKeI7PX3lzQT+GVEnJjs74yIyrTjOyLiVe0cybFJwJPA5Ig4lFa2BSgBbgCejYh/7SuO+vr6WLZs5PYgbms/zPf/8Dz/uXQthw4HHzlrFh9bdJynJTGzXkl6LCLqu5ZnupDTQGlK/vh3JoHehji/G7ijM2kAJLP0RkS0Ad8HTs9qtMPEqKJCPvrmY7n32nN5x8mT+O79z7Louvv42WMbvdysmfXbYCeOu4ArkvdXAHf2cu7ldKmmSks6ItU+suLVl1lPaipK+dq7F3DH/3gTkypH86mfPsGl3/kjyzfszHVoZpZHspY4JN0GPATMlbRR0pXAl4HzJK0FGpN9JNVLujHt2pnANOD+Lh/7Q0lPAU8BVcAXsxX/cHbK9PHc8bdv4rrLTmbjjv1c/K0/cO1Pn6C59UCuQzOzPJDVNo6hYqS3cfSm9cAhvvm7ddz8h+cYVVTI3y06jg+fOYuSosF+GDWzoWaotHHYEFNeWsxnL6jjN584h9NnTeDffrWat379AX632iPPzax7ThwGpFYQvPlDr+f7H349Aj5yyzI+9P1HebZlT65DM7MhxonDXmHh3Bp+/Ylz+NwFdSx7fgdvvf4BvnT3SloPHOr7YjMbEZw47FVKigr463Nmc++153LpqVO48ffPsfC6+/jJnza4+66ZOXFYz6rLR/GVd53MnVefyfQJY/j0z57k4m//gcde8FofZiOZE4f16aSpldz+0Tdx/V+ezJZdB3jnd/7I3/94OU273X3XbCRy4rCMFBSIS06Zyr3Xnsv/OPdY7n5yMwuvu49v37eOtvbDuQ7PzAaRE4f1y9hRRXz6/Hks/uQ5nHlcFV/59Rrecv0DLF7ZxEgYE2RmThx2lGZMHMv3PljPf33kdIoLC/jr/1rGB29+1ItHmY0AThz2mpxzfDW/uuZs/vkd81m+YSdv/fqDfOEXT7Nrv7vvmg1XThz2mhUXFnDlWbO479pzeXf9VG754/MsvO4+bnv0RQ67+67ZsOPEYQNmYtko/u3Sk/jFx85idtVYPvvzp7jwm7/nT893XQjSzPKZE4cNuBOnjOOnH30j33jPArbtOchl332Ij9/2OJt37c91aGY2AJw4LCskcdGCKfzu2jfzd4uO49dPb2HRdffzn0vXcuCQu++a5TMnDsuqMSVFfOotc1n6yTfz5uOr+eriZ2j82v38esVmd981y1NOHDYopk0Yw3c/cBo//Ks3MKakkI/+4M+878ZHWLPF3XfN8k1WE4ekmyU1S1qRVjZB0mJJa5PX8T1ce1jS8mS7K618lqRHJK2T9GNJJdm8BxtYZx5XxT0fP5svXHgCT2/azQX/+0E+f+cKdu47mOvQzCxD2X7iuAU4v0vZZ4ClETEHWJrsd2d/RCxItgvTyv8duD4ijgN2AFcOcMyWZUWFBVzxppnce+25XH76NP7vwy+w8Lr7+L8Pv+Duu2Z5IKuJIyIeALr2xbwIuDV5fytwcaafJ0nAIuD2o7nehpYJY0v44sWv45d/dzZzasv55/9ewdv/94M8vH5brkMzs17koo2jNiI2J++3ALU9nFcqaZmkhyVdnJRNBHZGRHuyvxGY0t3Fkq5Krl/W0tIyULFbFsyfXMGPrzqDb773FHbvP8R7bniYq3/4Zzbu2Jfr0MysGzltHI9Ut5qe6iZmJIukvxf4uqRj+/nZN0REfUTUV1dXv9ZQLcsk8Y6TJrP0U+dyTcMclqxqouGr93P94mfYf9Ddd82GkqIc/MwmSZMiYrOkSUBzdydFxEvJ63pJ9wGnAD8DKiUVJU8dU4GXBiluGwSjSwr5+/OO57L6qfzbPav5xtK13P7YRq5eeBwLplVybM1YRhUV5jpMs0EREew9eJjWA4fYvb899XrgEK0H2tm9/xC7D7S/Yr+1m/3vfbCes+ZUDWhcuUgcdwFXAF9OXu/sekLS02pfRLRJqgLOBL4SESHpXuBdwI96ut7y39TxY/jW+07l/c9u4wu/eJp/uuMpAIoKxKyqscw9ppx5x5RzfG05846pYOr40RQUKMdRm73SwfaO5I99e0Z//I+c0/byuX31FykpLKBidBHlpcVUlKZeJ40rpXxUMeWlRRwzbtSA35eyOQhL0m3AuUAV0AR8Hvhv4CfAdOAF4N0RsV1SPfDRiPgrSW8C/g/QQao67esRcVPymbNJJY0JwOPA+yOirbc46uvrY9myZQN/gzYoOjqCtc17WL1lN2u2tPJMUyurt7SyccfLU5iMLSlkTm16Miln7jHlTCwb+P9pbGSICPa0tXf7Lb73b/0vJ4oDhzr6/DnlpUVUlBa/8nV05vulxdl7Apf0WNJk8MrykTB614ljeGo9cIhnmvbwTFMra7a0HkksO/a9PKV7VdmoI0lkbm3q9fjackaXuLrLUslh064DLH9xJ09s3MnyDTvZtHM/u/cfYk9be5/f9kcVFbz8TX906vUVf9xH9fJHf3QxZSVFQ/pJuafEkYuqKrMBUV5azGkzxnPajJfHkEYELa1trDmSTFKvP3j4BdraU9/+JJgxYUxaMqlg7jHlzJw4hqJCT6YwnO0+cIgnN+xi+YYdLN+wi+UbdrJ1T6rCoqSogBMmV3D6zAkZf+Mfqe1tThw2rEiipqKUmopSzp7zcm+6wx3Bi9v3sWbL7iPJZM2WVhavbDryrbKkqIA5NWVHnkxS7SgV1FaMIjWEyPLJwfYO1mxpTUsSO3i2Ze+R48dWj+Wc46s4ZVolJ0+rZN4xFZQU+YtDJlxVZSPagUOHWde8J0kmqaTyTFMrTbtfbjYbN7q4SzIp5/hjyqkoLc5h5JYuItiwfT+Pb9jBE0mSWLFpNweTp8yqshIWTKtkQZIkTppaybjR/u/XF1dVmXWjtLiQE6eM48Qp415RvmPvwSPVXZ2vdzz+Enva2o+cM3lcaZJMKo60o8yudnfhwbBz30GWb9h5JEk8sXEX2/em5jsrLS7gdVPGccUbZ3BykiymVI72U+MAcuIw68b4sSWcMXsiZ8yeeKQsInhp5/5XJJM1W1r5/bqtHDqcenLv2l24M6lMqXR34aPV1n6YlZt288SGVOP18g07eX5balYBCebUlNFYV3MkScytLXdbVZa5qsrsNTrY3sFzW/cmyWT3kUb5nroLzz2mnMmVo6kqG0V12SgmlpUwdpS/w0EqOT+3dW+qh9OLqSSxcvPuI4m5tmJUUuU0npOnjeOkqZWU+d8ua9wd14nDBtmetvYjXYV76i7caXRxIRPLSqgqG0VV8jrxyOvLZVVlo6gcXTxsnl627WlLqpx28viGnTy5cRe79qf+fcaWFPK6qeNYMG08C6alXo8ZV5rjiEcWt3GYDbKyUUWcOn08p05/ZXfhrXsO0rT7AC172ti25yBb97SxbU8bW5P3L+08cKTOvrtp5gsLxISxJUwcW0J1+Sgmju0+wUwsK2FiWcmQaXM5cOgwK17adaS66YmNO9mwPfVUViCYe0wFF7xu0pEkcVxNGYXDJEEON04cZoNIEtXlo6gu73tEe0dHsHP/IbbtaeuSZFKvnYnm+W172dp6kP09rOVeXlp0pEqs65NMdVlJknBS5eWjigakEbmjI3i2Zc8rksTqza20J4lwSuVoFkyr5ANnzGDBtPGcOKWCMSX+c5Qv/F/KbIgqSJ4sJowtYU5teZ/n7zvYztbWg2zd28bW1ja27T145LUleapZ27yHh9e3dVtdBqmxLFVjS6jq40mmqmwUE8aWHHkiaN594EiSWL5hJ09t3EVr0gOtfFQRJ0+r5G/ePPtI20RNuauc8pkTh9kwMaakiOkTi5g+cUyf5x463MGOIwml+yeZlj1trNrcyra9bUcap9NJMGFMKnk0t6bGvRQViLpJFVx0yuSkbaKS2VVjh02bjKU4cZiNQMWFBUdG2PclIti9v/2VTzJpCebAocOcMHkcC6ZVcsLkiqxOumdDgxOHmfVKEuPGFDNuTDHHVpflOhwbAjxKxszM+sWJw8zM+sWJw8zM+sWJw8zM+iVriUPSzZKaJa1IK5sgabGktcnr+G6uWyDpIUlPS3pS0l+mHbtF0nOSlifbgmzFb2Zm3cvmE8ctwPldyj4DLI2IOcDSZL+rfcAHI+KE5PqvS6pMO/4PEbEg2ZYPeNRmZtarrCWOiHgA2N6l+CLg1uT9rcDF3Vz3TESsTd5vApqB6q7nmZlZbgx2G0dtRGxO3m8Bans7WdLpQAnwbFrxl5IqrOsl9Tjhj6SrJC2TtKylpeU1B25mZilZnVZd0kzglxFxYrK/MyIq047viIhXtXMkxyYB9wFXRMTDaWVbSCWTG4BnI+JfM4ijBXghg5CrgK0ZnJcvhtP9DKd7geF1P8PpXsD3k25GRLyqxmewR443SZoUEZuTJNDc3UmSKoC7gc91Jg2AtKeVNknfB67N5Id2d+M9/Nxl3c09n6+G0/0Mp3uB4XU/w+lewPeTicGuqroLuCJ5fwVwZ9cTJJUAdwD/FRG3dzk2KXkVqfaRFV2vNzOz7Mpmd9zbgIeAuZI2SroS+DJwnqS1QGOyj6R6STcml74bOAf4UDfdbn8o6SngKVKPX1/MVvxmZta9rFVVRcTlPRxq6ObcZcBfJe9/APygh89cNGABdu+GLH/+YBtO9zOc7gWG1/0Mp3sB30+fRsSa42ZmNnA85YiZmfWLE4eZmfWLE0dC0vmS1khaJ6m7qVCGrKOdF2yokjRN0r2SViZzll2TlOfdPUkqlfSopCeSe/lCUj5L0iPJ79uPk96EeUNSoaTHJf0y2c/L+5H0vKSnkk44y5KyvPs96ySpUtLtklZLWiXpjdm4HycOUv8TAN8C3gbMBy6XND+3UfXLLRzdvGBDVTvwqYiYD5wBXJ3898jHe2oDFkXEycAC4HxJZwD/DlwfEccBO4ArcxfiUbkGWJW2n8/3szCZ+65zrEM+/p51+gbw64iYB5xM6r/RwN9PRIz4DXgj8Ju0/c8Cn811XP28h5nAirT9NcCk5P0kYE2uY3wN93YncF6+3xMwBvgz8AZSI3mLkvJX/P4N9Q2YmvwBWgT8ElC+3g/wPFDVpSwvf8+AccBzJJ2esnk/fuJImQJsSNvfmJTls37NCzZUJdPWnAI8Qp7eU1Kts5zUTAmLSc29tjMi2pNT8u337evAp4GOZH8i+Xs/AfxW0mOSrkrK8vL3DJgFtADfT6oRb5Q0lizcjxPHCBCprxp51+9aUhnwM+ATEbE7/Vg+3VNEHI6IBaS+qZ8OzMttREdP0juA5oh4LNexDJCzIuJUUtXUV0s6J/1gPv2ekRqXdyrwnYg4BdhLl2qpgbofJ46Ul4BpaftTk7J81pQ2RUuP84INVZKKSSWNH0bEz5PivL6niNgJ3EuqKqdSUucA3Hz6fTsTuFDS88CPSFVXfYM8vZ+IeCl5bSY11dHp5O/v2UZgY0Q8kuzfTiqRDPj9OHGk/AmYk/QMKQHeQ2perXzW57xgQ1UyF9lNwKqI+Fraoby7J0nVShYikzSaVFvNKlIJ5F3JaXlxLwAR8dmImBoRM0n9f/K7iHgfeXg/ksZKKu98D7yF1Px3efd7BhARW4ANkuYmRQ3ASrJwPx45npB0Aam620Lg5oj4Um4jylwyL9i5pObvagI+D/w38BNgOqkp5d8dEV0X1hqSJJ0FPEhqTrLOevR/ItXOkVf3JOkkUouWFZL6ovaTiPhXSbNJfWOfADwOvD8i2nIXaf9JOhe4NiLekY/3k8R8R7JbBPy/iPiSpInk2e9Zp2RevxtJLT2xHvgwye8dA3g/ThxmZtYvrqoyM7N+ceIwM7N+ceIwM7N+ceIwM7N+ceIwM7N+ceIwywFJM9NnMzbLJ04cZmbWL04cZjkmaXYyKd3rcx2LWSaK+j7FzLIlmR7iR8CHIuKJXMdjlgknDrPcqSY1b9ClEbEy18GYZcpVVWa5swt4ETgr14GY9YefOMxy5yBwCfAbSXsi4v/lOiCzTDhxmOVQROxNFkdanCSPfJ/O30YAz45rZmb94jYOMzPrFycOMzPrFycOMzPrFycOMzPrFycOMzPrFycOMzPrFycOMzPrl/8f90qShdeKUCwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(K_list2,df_k2.iloc[0])\n",
    "plt.xlabel('k')\n",
    "plt.ylabel('Error Cuadrático Medio')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "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>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10.368308</td>\n",
       "      <td>10.374968</td>\n",
       "      <td>10.363542</td>\n",
       "      <td>10.357243</td>\n",
       "      <td>10.351268</td>\n",
       "      <td>10.307344</td>\n",
       "      <td>10.392848</td>\n",
       "      <td>10.36641</td>\n",
       "      <td>10.349925</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          31         32         33         34         35         36  \\\n",
       "0  10.368308  10.374968  10.363542  10.357243  10.351268  10.307344   \n",
       "\n",
       "          37        38         39  \n",
       "0  10.392848  10.36641  10.349925  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "K_list3 = [31,32,33,34,35,36,37,38,39]\n",
    "df_k3 = best_K(train, test, K_list3)\n",
    "df_k3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Con los resultados de hasta ahora vemos que **K = 36** tiene un mejor desempeño, el cuál varía dependiendo de los números random que se generan en la inicialización del vector U y V"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Validación"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Una vez elegido el modelo realizaremos la validación: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<__main__.Matrix_Factorization at 0x7f07df114bd0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "k_optima = 36\n",
    "\n",
    "als = Matrix_Factorization(n_iters = 15, k = k_optima, _lambda = 0.01)\n",
    "als.fit(train, test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Esta es la matriz completa para poder hacer recomendaciones\n",
    "MCompleta = als.predict()\n",
    "MCompleta = pd.DataFrame(MCompleta, \n",
    "                         columns=ratings.columns, \n",
    "                         index=ratings.index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Recomendaciones"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Ingrese un usuario nuevo aquí para mostrar las recomendaciones**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Aquí puede cambiar el número de usuario para obtener sus recomendaciones\n",
    "id_user = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "rec_nvas, rec_todas = recomendaciones_id(np.array(ratings)[id_user-1], \n",
    "                                         np.array(MCompleta)[id_user-1], \n",
    "                                         arr_movies, \n",
    "                                         bases_nombres_id, \n",
    "                                         id_user)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estás son las nuevas películas que te recomendamos usuario no.  2\n"
     ]
    },
    {
     "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>movieId</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>318</td>\n",
       "      <td>The Shawshank Redemption</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1135</th>\n",
       "      <td>316</td>\n",
       "      <td>Stargate</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1312</th>\n",
       "      <td>344</td>\n",
       "      <td>Ace Ventura: Pet Detective</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2613</th>\n",
       "      <td>595</td>\n",
       "      <td>Beauty and the Beast</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6655</th>\n",
       "      <td>380</td>\n",
       "      <td>True Lies</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      movieId                       title\n",
       "200       318    The Shawshank Redemption\n",
       "1135      316                    Stargate\n",
       "1312      344  Ace Ventura: Pet Detective\n",
       "2613      595        Beauty and the Beast\n",
       "6655      380                   True Lies"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Listado de películas que no ha visto el usuario.\n",
    "print(\"Estás son las nuevas películas que te recomendamos usuario no. \", id_user)\n",
    "bases_nombres_id[bases_nombres_id['movieId'].isin(rec_nvas)][[\"movieId\", \"title\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Estás películas te pueden interesar usuario no.  2\n"
     ]
    },
    {
     "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>movieId</th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>356</td>\n",
       "      <td>Forrest Gump</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>593</td>\n",
       "      <td>The Silence of the Lambs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>288</th>\n",
       "      <td>527</td>\n",
       "      <td>Schindler's List</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>384</th>\n",
       "      <td>590</td>\n",
       "      <td>Dances with Wolves</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>477</th>\n",
       "      <td>296</td>\n",
       "      <td>Pulp Fiction</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     movieId                     title\n",
       "5        356              Forrest Gump\n",
       "196      593  The Silence of the Lambs\n",
       "288      527          Schindler's List\n",
       "384      590        Dances with Wolves\n",
       "477      296              Pulp Fiction"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Listado de películas que incluyen las que ya vio el usuario\n",
    "print(\"Estás películas te pueden interesar usuario no. \", id_user)\n",
    "bases_nombres_id[bases_nombres_id['movieId'].isin(rec_todas)][[\"movieId\", \"title\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.51"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Probemos el NDCG para este usuario\n",
    "desempenio_NDCG(MCompleta.iloc[id_user])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Obtengamos  el desempeño para todos los usuarios"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Para la matriz obtenida"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "ndcg = obtain_ndcg_all_users(MCompleta)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Para la matriz original"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "ndcg_Y_0 = obtain_ndcg_all_users(ratings)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Algunas comparaciones:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "user: 1 - ndcg: 0.73\n",
      "user: 2 - ndcg: 0.63\n",
      "user: 3 - ndcg: 0.51\n",
      "user: 4 - ndcg: 0.59\n",
      "user: 5 - ndcg: 0.57\n",
      "user: 6 - ndcg: 0.56\n",
      "user: 7 - ndcg: 0.61\n",
      "user: 8 - ndcg: 0.45\n"
     ]
    }
   ],
   "source": [
    "# Matriz final\n",
    "for x in list(ndcg)[0:8]:\n",
    "    print (\"user: {} - ndcg: {}\".format(x, ndcg[x]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "user: 1 - ndcg: 0.26\n",
      "user: 2 - ndcg: 0.53\n",
      "user: 3 - ndcg: 0.36\n",
      "user: 4 - ndcg: 0.56\n",
      "user: 5 - ndcg: 0.45\n",
      "user: 6 - ndcg: 0.32\n",
      "user: 7 - ndcg: 0.53\n",
      "user: 8 - ndcg: 0.45\n"
     ]
    }
   ],
   "source": [
    "# Matriz inicial\n",
    "for x in list(ndcg_Y_0)[0:8]:\n",
    "    print (\"user: {} - ndcg: {}\".format(x, ndcg_Y_0[x]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Transformando diccionario a arreglo\n",
    "# Datos predichos\n",
    "ndcg_ar = list(ndcg.values())\n",
    "# Datos originales\n",
    "ndcg_Y_0_ar = list(ndcg_Y_0.values())\n",
    "# Usuarios\n",
    "list_fisrt_users = list(range(1, 51))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,10))\n",
    "plt.plot(list_fisrt_users, ndcg_ar[:50], label = 'NDCG_Fin', linewidth = 1,color='b')\n",
    "plt.plot(list_fisrt_users, ndcg_Y_0_ar[:50], label = 'NDCG_Ini', linewidth = 1,color='g')\n",
    "plt.xlabel('Usuarios')\n",
    "plt.ylabel('NDCG')\n",
    "plt.legend(['NDCG_Fin','NDCG_Ini'],loc='upper left', bbox_to_anchor=(1.0, 0.5))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Se puede observar que se mejoró en el desempeño"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}