NWM_ML_Eval.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "169234f7",
   "metadata": {},
   "source": [
    "![GEE_ML](./Images/NWM_ML_Hydrological_Cycle.png)\n",
    "\n",
    "# Evaluating Hydrological Model Performance\n",
    "\n",
    "\n",
    "Using this data, we processed it into a daily temporal resolution dataframe that supported the training and optimization of an XGBoost machine learning model.\n",
    "We trained a standard model on three years of the daily data (2020-2022) using a randomized train/test split of 66/33%, i.e., training the model on 66% of the data and evaluating model performance on the remaining 33%.\n",
    "For model performance, we used the metrics of KGE, Percent Bias, Mean Absolute Percentage Error, and root-mean-squared-error.\n",
    "\n",
    "We then used the GridSearch CV function to optimize the hyperparameters of the XGBoost algorithm, observing small improvements in model skill.\n",
    "\n",
    "In this final lesson, we will combine the skills from the previous two to develop a model prediction workflow, as well as evaluate the prediction skill of the models to unseen data.\n",
    "\n",
    "## Learning Objectives\n",
    "In this chapter we will:\n",
    "\n",
    "1. Build a new prediction dataframe to evaluate the model(s) to unseen data \n",
    "1. Process and prepare the new dataframe for the model\n",
    "1. Make predictions on the unseen data using each model\n",
    "1. Evaluate the performance of each model \n",
    "1. Investigate why or why not the model performs as it does"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9c8ea94b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\rjohnson18\\Anaconda3\\envs\\GEE_env\\lib\\site-packages\\dataretrieval\\nadp.py:44: UserWarning: GDAL not installed. Some functions will not work.\n",
      "  warnings.warn('GDAL not installed. Some functions will not work.')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.9.12\n"
     ]
    }
   ],
   "source": [
    "#Load necessary packages\n",
    "from hydrotools.nwis_client.iv import IVDataService\n",
    "from hydrotools.nwm_client import utils\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import Get_StreamStats\n",
    "from scipy import optimize\n",
    "from datetime import datetime, date\n",
    "import pickle\n",
    "import joblib\n",
    "from xgboost import XGBRegressor\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from hydrotools.nwis_client.iv import IVDataService\n",
    "from hydrotools.nwm_client import utils\n",
    "from progressbar import ProgressBar\n",
    "\n",
    "%matplotlib inline\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "def NWIS_retrieve(site, startDT, endDT):\n",
    "    service = IVDataService()\n",
    "    usgs_data = service.get(\n",
    "        sites=site,\n",
    "        startDT= startDT,\n",
    "        endDT= endDT\n",
    "        )\n",
    "    return usgs_data    \n",
    "\n",
    "from platform import python_version\n",
    "\n",
    "print(python_version())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42db8a3d",
   "metadata": {},
   "source": [
    "## Building the Prediction DataFrame\n",
    "We will be making streamflow predictions for our test location of Red Butte Creek from the beginning of 2023 to present.\n",
    "To do so, we need to load the stationary data (e.g., StreamStats) and retrieve NLDAS Precipitation and Temperature.\n",
    "Let's develop the workflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ab3356cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Get NWIS observations for the new time period\n",
    "start = '2020-01-01'\n",
    "end = '2020-12-31'\n",
    "NWISid = '10133600'\n",
    "\n",
    "flow = NWIS_retrieve(NWISid, start, end)\n",
    "\n",
    "#make the date the index for plotting and resampling\n",
    "flow['datetime'] = flow['value_time']\n",
    "flow.set_index('datetime', inplace = True)\n",
    "\n",
    "#clean the data\n",
    "flow = flow[flow['value'] > 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "da0fa5c4",
   "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>index</th>\n",
       "      <th>NWIS_site_id</th>\n",
       "      <th>Lat</th>\n",
       "      <th>Long</th>\n",
       "      <th>Drainage_area_mi2</th>\n",
       "      <th>Mean_Basin_Elev_ft</th>\n",
       "      <th>Perc_Forest</th>\n",
       "      <th>Perc_Develop</th>\n",
       "      <th>Perc_Imperv</th>\n",
       "      <th>Perc_Herbace</th>\n",
       "      <th>Perc_Slop_30</th>\n",
       "      <th>Mean_Ann_Precip_in</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>10133600</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index  NWIS_site_id        Lat        Long  Drainage_area_mi2  \\\n",
       "0      2      10133600  40.688039 -111.533719               16.3   \n",
       "\n",
       "   Mean_Basin_Elev_ft  Perc_Forest  Perc_Develop  Perc_Imperv  Perc_Herbace  \\\n",
       "0              7890.0         69.5          14.3         4.08          1.62   \n",
       "\n",
       "   Perc_Slop_30  Mean_Ann_Precip_in  \n",
       "0          48.3                21.5  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Load StreamStats Dataframe from the Chapter 1, adjust as necessary for additional sites.\n",
    "StreamStats = pd.read_csv('./Data/StreamStats2.csv')\n",
    "#remove \"Unamed: 0\" column\n",
    "StreamStats.pop('Unnamed: 0')\n",
    "StreamStats = StreamStats[StreamStats['NWIS_site_id'] == int(NWISid)]\n",
    "StreamStats.reset_index(inplace = True)\n",
    "StreamStats"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32d89873",
   "metadata": {},
   "source": [
    "## Get climate information from GEE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8deb2d1f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<p>To authorize access needed by Earth Engine, open the following\n",
       "        URL in a web browser and follow the instructions:</p>\n",
       "        <p><a href=https://code.earthengine.google.com/client-auth?scopes=https%3A//www.googleapis.com/auth/earthengine%20https%3A//www.googleapis.com/auth/devstorage.full_control&request_id=37H7kd7zLw9CUDgtAHWeeFFrH1hYpfKSSflsietDjNg&tc=qT2PoDbvvh3_a8BTbM8ML6ffS56fNtElFGCs-RRubzY&cc=C8OjLm-34rFHgLSZaryzr02HmvFp73ovBCrTn9qWQfQ>https://code.earthengine.google.com/client-auth?scopes=https%3A//www.googleapis.com/auth/earthengine%20https%3A//www.googleapis.com/auth/devstorage.full_control&request_id=37H7kd7zLw9CUDgtAHWeeFFrH1hYpfKSSflsietDjNg&tc=qT2PoDbvvh3_a8BTbM8ML6ffS56fNtElFGCs-RRubzY&cc=C8OjLm-34rFHgLSZaryzr02HmvFp73ovBCrTn9qWQfQ</a></p>\n",
       "        <p>The authorization workflow will generate a code, which you should paste in the box below.</p>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Enter verification code: 4/1Adeu5BXCMUboXvTZAWi6T0AHq427h5WQb-51ZPGpJ1lQVvvGF2R5sBZ3hc0\n",
      "\n",
      "Successfully saved authorization token.\n"
     ]
    }
   ],
   "source": [
    "#Connect to Earth Engine to get NLDAS Temperature and Precipitation\n",
    "import ee\n",
    "import EE_funcs\n",
    "ee.Authenticate()\n",
    "ee.Initialize()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "125a71dc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Retrieving NLDAS Temperature and Precipitation from GEE\n",
      "Retrieving NLDAS Temperature and Precipitation for desired location(s) \n",
      "Converting object to DataFrame\n"
     ]
    }
   ],
   "source": [
    "#NLDAS temperature\n",
    "print('Retrieving NLDAS Temperature and Precipitation from GEE')\n",
    "temp = ee.ImageCollection('NASA/NLDAS/FORA0125_H002').select('temperature').filterDate(start, end)\n",
    "\n",
    "#NLDAS precipitation\n",
    "precip = ee.ImageCollection('NASA/NLDAS/FORA0125_H002').select('total_precipitation').filterDate(start, end)\n",
    "\n",
    "# Define slc location of interest, lets use the USGS guage location!\n",
    "lon = StreamStats['Long'][0]\n",
    "lat = StreamStats['Lat'][0]\n",
    "poi = ee.Geometry.Point(lon, lat)\n",
    "\n",
    "scale = 1000  # scale in meters\n",
    "\n",
    "# Get the data for the pixel intersecting the point at Red Butte Creek.\n",
    "print('Retrieving NLDAS Temperature and Precipitation for desired location(s) ')\n",
    "temp_poi = temp.getRegion(poi, scale).getInfo()\n",
    "precip_poi = precip.getRegion(poi, scale).getInfo()\n",
    "\n",
    "#Convert pointer object array to dataframe to work with\n",
    "print('Converting object to DataFrame')\n",
    "temp_RB = EE_funcs.ee_array_to_df(temp_poi,['temperature'])\n",
    "precip_RB = EE_funcs.ee_array_to_df(precip_poi,['total_precipitation'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "cc675145",
   "metadata": {},
   "outputs": [],
   "source": [
    "#resample to the desired temporal scale, take the respective statistic from the data of resampling\n",
    "temporal_resample = 'D'\n",
    "kg_in = 0.04\n",
    "\n",
    "#Temperature\n",
    "temp_RB.set_index('datetime', inplace = True)\n",
    "temp_RB = temp_RB.resample(temporal_resample).max()\n",
    "temp_RB.reset_index(inplace = True)\n",
    "\n",
    "#Precipitation\n",
    "precip_RB.set_index('datetime', inplace = True)\n",
    "precip_RB = precip_RB.resample(temporal_resample).sum()\n",
    "precip_RB.reset_index(inplace = True)\n",
    "\n",
    "#make columns for Fahrenheit\n",
    "temp_RB['temperature_F'] = (temp_RB['temperature']*9/5)+32 \n",
    "\n",
    "#make columns for inches\n",
    "precip_RB['precipitation_in'] = precip_RB['total_precipitation']*kg_in\n",
    "\n",
    "#need to add in time to precip data, same daterange as temperaure\n",
    "precip_RB['time'] = temp_RB['time']\n",
    "\n",
    "#ensure columns are in the correct order\n",
    "colorder = ['datetime', 'time', 'total_precipitation', 'precipitation_in']\n",
    "precip_RB = precip_RB[colorder]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1be8bca1",
   "metadata": {},
   "source": [
    "# Get NWM predictions for site"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "529e36db",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function for getting NWM data\n",
    "def get_NHD_Model_info(NWIS_sites):   \n",
    "    print('Getting NHD reaches')\n",
    "   #Get NHD reach colocated with NWIS\n",
    "    NHD_reaches = []\n",
    "    NWIS_NHDPlus = pd.DataFrame(columns = ['NWISid','NHDPlusid'])\n",
    "    for site in NWIS_sites:\n",
    "        \n",
    "        \n",
    "        try:\n",
    "            NHD_NWIS_df = utils.crosswalk(usgs_site_codes=site)\n",
    "            NHD_segment = NHD_NWIS_df.nwm_feature_id.values[0]\n",
    "            NHD_reaches.append(NHD_segment)\n",
    "\n",
    "        except:\n",
    "            NHD_segment = np.nan\n",
    "            NHD_reaches.append(NHD_segment)\n",
    "    \n",
    "    NWIS_NHDPlus['NWISid'] = NWIS_sites\n",
    "    NWIS_NHDPlus['NHDPlusid'] = NHD_reaches\n",
    "    \n",
    "    print(NHD_reaches)\n",
    "    return NHD_reaches, NWIS_NHDPlus\n",
    "\n",
    "def get_NWM_data(NWIS_NHDPlus, start, end):\n",
    "    NWM_obs = {}\n",
    "    reaches = NWIS_NHDPlus['NHDPlusid'].to_list()\n",
    "    for reach in reaches:\n",
    "        if reach > 1:\n",
    "            r = str(int(reach))\n",
    "            reachid = f\"{r}_1980_2020.csv\"\n",
    "            filepath = f\"./Data/NWM/ut/{reachid}\"\n",
    "            NWM_obs[r] = pd.read_csv(filepath)\n",
    "            NWM_obs[r].rename(columns = {'Datetime':'datetime'}, inplace = True)\n",
    "            NWM_obs[r].set_index('datetime', inplace = True)\n",
    "            NWM_obs[r]=NWM_obs[r][start:end]\n",
    "            NWM_obs[r] = pd.DataFrame(NWM_obs[r]['NWM_flow'])\n",
    "            NWM_obs[r]['NWIS_site_id'] =  NWIS_NHDPlus['NWISid'][ NWIS_NHDPlus['NHDPlusid'] == reach].values[0]\n",
    "            NWM_obs[r].reset_index(inplace = True)\n",
    "        else:\n",
    "            r = str(reach)\n",
    "           \n",
    "    return NWM_obs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "e9b7b46a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% |########################################################################|\n"
     ]
    }
   ],
   "source": [
    "NWISids = [NWISid]\n",
    "NWM_obs = get_NWM_data(NWIS_NHDPlus, start, end)\n",
    "\n",
    "#merge into one DF\n",
    "NWM_obs_df = pd.DataFrame()\n",
    "pbar = ProgressBar()\n",
    "for site in pbar(NWM_obs):\n",
    "    NWM_obs_df = NWM_obs_df.append(NWM_obs[site])\n",
    "\n",
    "#convert to datetime to make sure it joins nicely with the other variables \n",
    "NWM_obs_df['datetime'] = pd.to_datetime(NWM_obs_df['datetime'])\n",
    "\n",
    "#Convert NWIS site id to int to make sure it joins nicely with the other variables\n",
    "NWM_obs_df['NWIS_site_id'] = NWM_obs_df['NWIS_site_id'].astype(int)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "edb5b1c4",
   "metadata": {},
   "source": [
    "## Create a DataFrame for input into ML model\n",
    "\n",
    "We have looked at daily resolution temperature and precipitation, so lets stay at this resolution. \n",
    "The NLDAS data is available at 1 hr temporal resolution with the ability to resample to an greater timestep using the [resample](https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.resample.html) function.\n",
    "\n",
    "The [Merge, Join, Concatenate and Compare](https://pandas.pydata.org/pandas-docs/stable/user_guide/merging.html) functions will be most useful here."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "83bdb6e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Lets begin with the temperature and precipitation data\n",
    "Pred_DF = temp_RB.copy()\n",
    "\n",
    "#remote unnecessary columns, time, temperature\n",
    "colrem = ['time', 'temperature']\n",
    "Pred_DF.drop(colrem, axis =1, inplace =True)\n",
    "\n",
    "#lets add the precipitation data\n",
    "Pred_DF['precipitation_in'] = precip_RB['precipitation_in']\n",
    "\n",
    "#set index to datetime\n",
    "Pred_DF.set_index('datetime', drop = True, inplace = True)\n",
    "\n",
    "#lets add streamflow, but we need to make it at the same temporal resolution (daily)\n",
    "cols = ['value']\n",
    "flow = flow[cols]\n",
    "flow = flow.resample(temporal_resample).mean()\n",
    "flow.rename(columns ={'value':'flow_cfs'}, inplace = True)\n",
    "\n",
    "#concat dataframe by date\n",
    "Pred_DF = pd.concat([Pred_DF, flow], axis = 1).dropna()\n",
    "\n",
    "#Add the NWIS id to match with StreamStats\n",
    "Pred_DF['NWIS_site_id'] = NWISid\n",
    "#convert object to int64 to match StreamStats\n",
    "Pred_DF['NWIS_site_id'] = Pred_DF['NWIS_site_id'].astype('int64')\n",
    "\n",
    "#Lets join the StreamStats Features, we can merge by NWIS_site_id\n",
    "DF = Pred_DF.join(StreamStats.set_index(\"NWIS_site_id\"), on=\"NWIS_site_id\")\n",
    "\n",
    "#reset the index\n",
    "DF.reset_index(inplace = True)\n",
    "\n",
    "#Add in NWM predictions\n",
    "#connect NWM flow with streamstats and climate\n",
    "DF = DF.merge(NWM_obs_df,on=['NWIS_site_id', 'datetime'])\n",
    "\n",
    "#Remove NWIS_site_id\n",
    "DF.pop('NWIS_site_id')\n",
    "\n",
    "#Add Day of year (DOY)\n",
    "DF['DOY'] = DF.datetime.dt.dayofyear\n",
    "\n",
    "#remove old datetime column\n",
    "DF.pop('datetime')\n",
    "\n",
    "#remove non applicable columns\n",
    "DF.pop('index')\n",
    "\n",
    "#save DataFrame for Model Development\n",
    "DF.to_csv('./Data/Processed/NWM_ML_Pred_DF1.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "a877c2ea",
   "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>temperature_F</th>\n",
       "      <th>precipitation_in</th>\n",
       "      <th>flow_cfs</th>\n",
       "      <th>Lat</th>\n",
       "      <th>Long</th>\n",
       "      <th>Drainage_area_mi2</th>\n",
       "      <th>Mean_Basin_Elev_ft</th>\n",
       "      <th>Perc_Forest</th>\n",
       "      <th>Perc_Develop</th>\n",
       "      <th>Perc_Imperv</th>\n",
       "      <th>Perc_Herbace</th>\n",
       "      <th>Perc_Slop_30</th>\n",
       "      <th>Mean_Ann_Precip_in</th>\n",
       "      <th>NWM_flow</th>\n",
       "      <th>DOY</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>26.43800</td>\n",
       "      <td>0.332920</td>\n",
       "      <td>7.143667</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.534960</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>26.22200</td>\n",
       "      <td>0.880368</td>\n",
       "      <td>9.437500</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.534960</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>26.27600</td>\n",
       "      <td>0.209520</td>\n",
       "      <td>8.792396</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.402530</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>29.44400</td>\n",
       "      <td>0.072208</td>\n",
       "      <td>9.184584</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.181813</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>26.40200</td>\n",
       "      <td>0.088600</td>\n",
       "      <td>9.007292</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.181813</td>\n",
       "      <td>5</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>70.03400</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.569896</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>11.300693</td>\n",
       "      <td>270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>270</th>\n",
       "      <td>62.09607</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.403021</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>11.197692</td>\n",
       "      <td>271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>271</th>\n",
       "      <td>58.74800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.627604</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>272</th>\n",
       "      <td>68.12600</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.605729</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>273</th>\n",
       "      <td>68.68400</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.456979</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>274</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>274 rows × 15 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     temperature_F  precipitation_in  flow_cfs        Lat        Long  \\\n",
       "0         26.43800          0.332920  7.143667  40.688039 -111.533719   \n",
       "1         26.22200          0.880368  9.437500  40.688039 -111.533719   \n",
       "2         26.27600          0.209520  8.792396  40.688039 -111.533719   \n",
       "3         29.44400          0.072208  9.184584  40.688039 -111.533719   \n",
       "4         26.40200          0.088600  9.007292  40.688039 -111.533719   \n",
       "..             ...               ...       ...        ...         ...   \n",
       "269       70.03400          0.000000  5.569896  40.688039 -111.533719   \n",
       "270       62.09607          0.000000  5.403021  40.688039 -111.533719   \n",
       "271       58.74800          0.000000  5.627604  40.688039 -111.533719   \n",
       "272       68.12600          0.000000  5.605729  40.688039 -111.533719   \n",
       "273       68.68400          0.000000  5.456979  40.688039 -111.533719   \n",
       "\n",
       "     Drainage_area_mi2  Mean_Basin_Elev_ft  Perc_Forest  Perc_Develop  \\\n",
       "0                 16.3              7890.0         69.5          14.3   \n",
       "1                 16.3              7890.0         69.5          14.3   \n",
       "2                 16.3              7890.0         69.5          14.3   \n",
       "3                 16.3              7890.0         69.5          14.3   \n",
       "4                 16.3              7890.0         69.5          14.3   \n",
       "..                 ...                 ...          ...           ...   \n",
       "269               16.3              7890.0         69.5          14.3   \n",
       "270               16.3              7890.0         69.5          14.3   \n",
       "271               16.3              7890.0         69.5          14.3   \n",
       "272               16.3              7890.0         69.5          14.3   \n",
       "273               16.3              7890.0         69.5          14.3   \n",
       "\n",
       "     Perc_Imperv  Perc_Herbace  Perc_Slop_30  Mean_Ann_Precip_in   NWM_flow  \\\n",
       "0           4.08          1.62          48.3                21.5   9.534960   \n",
       "1           4.08          1.62          48.3                21.5   9.534960   \n",
       "2           4.08          1.62          48.3                21.5   9.402530   \n",
       "3           4.08          1.62          48.3                21.5   9.181813   \n",
       "4           4.08          1.62          48.3                21.5   9.181813   \n",
       "..           ...           ...           ...                 ...        ...   \n",
       "269         4.08          1.62          48.3                21.5  11.300693   \n",
       "270         4.08          1.62          48.3                21.5  11.197692   \n",
       "271         4.08          1.62          48.3                21.5  10.947546   \n",
       "272         4.08          1.62          48.3                21.5  10.947546   \n",
       "273         4.08          1.62          48.3                21.5  10.947546   \n",
       "\n",
       "     DOY  \n",
       "0      1  \n",
       "1      2  \n",
       "2      3  \n",
       "3      4  \n",
       "4      5  \n",
       "..   ...  \n",
       "269  270  \n",
       "270  271  \n",
       "271  272  \n",
       "272  273  \n",
       "273  274  \n",
       "\n",
       "[274 rows x 15 columns]"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DF"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e8a35c8f",
   "metadata": {},
   "source": [
    "## Scale the DataFrame and make predictions\n",
    "\n",
    "Set up the DataFrame for predictions, i.e., remote flow observations and scale the model input variables."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "ea205cef",
   "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>NWM_ML</th>\n",
       "      <th>flow_cfs_pred</th>\n",
       "      <th>flow_cfs</th>\n",
       "      <th>temperature_F</th>\n",
       "      <th>precipitation_in</th>\n",
       "      <th>Lat</th>\n",
       "      <th>Long</th>\n",
       "      <th>Drainage_area_mi2</th>\n",
       "      <th>Mean_Basin_Elev_ft</th>\n",
       "      <th>Perc_Forest</th>\n",
       "      <th>Perc_Develop</th>\n",
       "      <th>Perc_Imperv</th>\n",
       "      <th>Perc_Herbace</th>\n",
       "      <th>Perc_Slop_30</th>\n",
       "      <th>Mean_Ann_Precip_in</th>\n",
       "      <th>NWM_flow</th>\n",
       "      <th>DOY</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9.817655</td>\n",
       "      <td>9.487590</td>\n",
       "      <td>7.143667</td>\n",
       "      <td>26.43800</td>\n",
       "      <td>0.332920</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.534960</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10.108067</td>\n",
       "      <td>8.597317</td>\n",
       "      <td>9.437500</td>\n",
       "      <td>26.22200</td>\n",
       "      <td>0.880368</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.534960</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>6.920573</td>\n",
       "      <td>7.925487</td>\n",
       "      <td>8.792396</td>\n",
       "      <td>26.27600</td>\n",
       "      <td>0.209520</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.402530</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8.736672</td>\n",
       "      <td>9.245140</td>\n",
       "      <td>9.184584</td>\n",
       "      <td>29.44400</td>\n",
       "      <td>0.072208</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.181813</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.888804</td>\n",
       "      <td>6.834861</td>\n",
       "      <td>9.007292</td>\n",
       "      <td>26.40200</td>\n",
       "      <td>0.088600</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>9.181813</td>\n",
       "      <td>5</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>5.632184</td>\n",
       "      <td>5.112414</td>\n",
       "      <td>5.569896</td>\n",
       "      <td>70.03400</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>11.300693</td>\n",
       "      <td>270</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>270</th>\n",
       "      <td>6.750904</td>\n",
       "      <td>6.822995</td>\n",
       "      <td>5.403021</td>\n",
       "      <td>62.09607</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>11.197692</td>\n",
       "      <td>271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>271</th>\n",
       "      <td>7.888011</td>\n",
       "      <td>6.901912</td>\n",
       "      <td>5.627604</td>\n",
       "      <td>58.74800</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>272</th>\n",
       "      <td>5.909693</td>\n",
       "      <td>5.235436</td>\n",
       "      <td>5.605729</td>\n",
       "      <td>68.12600</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>273</th>\n",
       "      <td>5.988474</td>\n",
       "      <td>5.130267</td>\n",
       "      <td>5.456979</td>\n",
       "      <td>68.68400</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.688039</td>\n",
       "      <td>-111.533719</td>\n",
       "      <td>16.3</td>\n",
       "      <td>7890.0</td>\n",
       "      <td>69.5</td>\n",
       "      <td>14.3</td>\n",
       "      <td>4.08</td>\n",
       "      <td>1.62</td>\n",
       "      <td>48.3</td>\n",
       "      <td>21.5</td>\n",
       "      <td>10.947546</td>\n",
       "      <td>274</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>274 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        NWM_ML  flow_cfs_pred  flow_cfs  temperature_F  precipitation_in  \\\n",
       "0     9.817655       9.487590  7.143667       26.43800          0.332920   \n",
       "1    10.108067       8.597317  9.437500       26.22200          0.880368   \n",
       "2     6.920573       7.925487  8.792396       26.27600          0.209520   \n",
       "3     8.736672       9.245140  9.184584       29.44400          0.072208   \n",
       "4     4.888804       6.834861  9.007292       26.40200          0.088600   \n",
       "..         ...            ...       ...            ...               ...   \n",
       "269   5.632184       5.112414  5.569896       70.03400          0.000000   \n",
       "270   6.750904       6.822995  5.403021       62.09607          0.000000   \n",
       "271   7.888011       6.901912  5.627604       58.74800          0.000000   \n",
       "272   5.909693       5.235436  5.605729       68.12600          0.000000   \n",
       "273   5.988474       5.130267  5.456979       68.68400          0.000000   \n",
       "\n",
       "           Lat        Long  Drainage_area_mi2  Mean_Basin_Elev_ft  \\\n",
       "0    40.688039 -111.533719               16.3              7890.0   \n",
       "1    40.688039 -111.533719               16.3              7890.0   \n",
       "2    40.688039 -111.533719               16.3              7890.0   \n",
       "3    40.688039 -111.533719               16.3              7890.0   \n",
       "4    40.688039 -111.533719               16.3              7890.0   \n",
       "..         ...         ...                ...                 ...   \n",
       "269  40.688039 -111.533719               16.3              7890.0   \n",
       "270  40.688039 -111.533719               16.3              7890.0   \n",
       "271  40.688039 -111.533719               16.3              7890.0   \n",
       "272  40.688039 -111.533719               16.3              7890.0   \n",
       "273  40.688039 -111.533719               16.3              7890.0   \n",
       "\n",
       "     Perc_Forest  Perc_Develop  Perc_Imperv  Perc_Herbace  Perc_Slop_30  \\\n",
       "0           69.5          14.3         4.08          1.62          48.3   \n",
       "1           69.5          14.3         4.08          1.62          48.3   \n",
       "2           69.5          14.3         4.08          1.62          48.3   \n",
       "3           69.5          14.3         4.08          1.62          48.3   \n",
       "4           69.5          14.3         4.08          1.62          48.3   \n",
       "..           ...           ...          ...           ...           ...   \n",
       "269         69.5          14.3         4.08          1.62          48.3   \n",
       "270         69.5          14.3         4.08          1.62          48.3   \n",
       "271         69.5          14.3         4.08          1.62          48.3   \n",
       "272         69.5          14.3         4.08          1.62          48.3   \n",
       "273         69.5          14.3         4.08          1.62          48.3   \n",
       "\n",
       "     Mean_Ann_Precip_in   NWM_flow  DOY  \n",
       "0                  21.5   9.534960    1  \n",
       "1                  21.5   9.534960    2  \n",
       "2                  21.5   9.402530    3  \n",
       "3                  21.5   9.181813    4  \n",
       "4                  21.5   9.181813    5  \n",
       "..                  ...        ...  ...  \n",
       "269                21.5  11.300693  270  \n",
       "270                21.5  11.197692  271  \n",
       "271                21.5  10.947546  272  \n",
       "272                21.5  10.947546  273  \n",
       "273                21.5  10.947546  274  \n",
       "\n",
       "[274 rows x 17 columns]"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Split the data into features and target\n",
    "y = DF['flow_cfs']\n",
    "\n",
    "#remove target from features\n",
    "DF.pop('flow_cfs')\n",
    "\n",
    "X = DF\n",
    "\n",
    "#scale the predictor inputs\n",
    "#load the scalers\n",
    "scaler_x = pickle.load(open('./Model/NWM_ML/XGBoost/scaler_x1.pkl', 'rb'))\n",
    "scaler_y = pickle.load(open('./Model/NWM_ML/XGBoost/scaler_y1.pkl', 'rb'))\n",
    "\n",
    "#apply the scaler to the X_testing data\n",
    "X_scaled = scaler_x.transform(X)\n",
    "\n",
    "#Load the original model\n",
    "model = pickle.load(open('./Model/NWM_ML/XGBoost/Model_1.dat', \"rb\"))\n",
    "\n",
    "# make a prediction\n",
    "y_pred = model.predict(X_scaled)\n",
    "    \n",
    "#scale the predictions back to streamflow in cfs and convert to pandas DataFrame\n",
    "y_pred = pd.DataFrame(scaler_y.inverse_transform(pd.DataFrame(y_pred)),columns = ['flow_cfs_pred'])\n",
    "y = pd.DataFrame(y, columns = ['flow_cfs'])\n",
    "\n",
    "#Load the optimized model, remember to change it ot the correction one!\n",
    "model = pickle.load(open('./Model/NWM_ML/XGBoost/Model_HyperParam.dat', \"rb\"))\n",
    "\n",
    "# make a prediction\n",
    "y_pred_Hyp = model.predict(X_scaled)\n",
    "    \n",
    "#scale the predictions back to streamflow in cfs and convert to pandas DataFrame\n",
    "y_pred_Hyp = pd.DataFrame(scaler_y.inverse_transform(pd.DataFrame(y_pred_Hyp)),columns = ['NWM_ML'])\n",
    "\n",
    "\n",
    "#Combine X_test, y_pred, and y_test DataFrames to evaluate the performance of the model\n",
    "#reset indexes\n",
    "DFs = [y_pred_Hyp, y_pred, y, X]\n",
    "for df in DFs:\n",
    "    df.reset_index(drop = True, inplace = True)\n",
    "    \n",
    "#merge\n",
    "Eval_DF = pd.concat(DFs, axis=1)\n",
    "\n",
    "Eval_DF\n",
    "    \n",
    "#merge\n",
    "Eval_DF = pd.concat(DFs, axis=1)\n",
    "\n",
    "Eval_DF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "7a0570a9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'Evaluation_funcs' from 'C:\\\\Users\\\\rjohnson18\\\\Box\\\\CIROH\\\\EarthEngine\\\\GEE_Workshop\\\\GEE_ML_Workflow\\\\Evaluation_funcs.py'>"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import importlib\n",
    "importlib.reload(Evaluation_funcs)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "810771e6",
   "metadata": {},
   "source": [
    "## Comparision of hydrological models\n",
    "\n",
    "We can find the evaluation functions from Chapter 2 in the Evaluation_funcs.py file and import them in the interactive python file to A) clean up our notebook and B) evaluate the performance of different models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "92cc73b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMSE for  NWM_ML  is  12.060587  cfs\n",
      "RMSE for  NWM_flow  is  14.465549133012166  cfs\n",
      "Mean Absolute Percentage Error for  NWM_ML  is  38.01 %\n",
      "Mean Absolute Percentage Error for  NWM_flow  is  79.89 %\n",
      "Percentage Bias for  NWM_ML  is  -41.91 %\n",
      "Percentage Bias for  NWM_flow  is  -64.56 %\n",
      "Kling-Glutz Efficiency for  NWM_ML  is  -0.43\n",
      "Kling-Glutz Efficiency for  NWM_flow  is  -0.61\n"
     ]
    }
   ],
   "source": [
    "import Evaluation_funcs\n",
    "\n",
    "#Evaluate model performance of the different models, 'flow_cfs_pred', \n",
    "prediction_columns = ['NWM_ML', 'NWM_flow']\n",
    "\n",
    "#plot the predictions\n",
    "filename = 'NWM_ML_parity'\n",
    "Evaluation_funcs.Model_Evaluation_Plots(Eval_DF, prediction_columns, NWISid, filename)\n",
    "\n",
    "#Get RMSE from the model\n",
    "Evaluation_funcs.RMSE(Eval_DF, prediction_columns)\n",
    "\n",
    "#Get Mean Absolute Percentage Error from the model\n",
    "Evaluation_funcs.MAPE(Eval_DF, prediction_columns)\n",
    "\n",
    "#Get Percent Bias from the model\n",
    "Evaluation_funcs.PBias(Eval_DF, prediction_columns)\n",
    "\n",
    "#Get Kling-Gutz Efficiency from the model\n",
    "Evaluation_funcs.KGE(Eval_DF, prediction_columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "832b2aad",
   "metadata": {},
   "source": [
    "## That was not the model performance we were expecting\n",
    "Let's investigate!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "a58e7c1c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "filename = 'NWM_ML_Hydrograph'\n",
    "Evaluation_funcs.Hydrograph_Evaluation_Plots(Eval_DF, prediction_columns,NWISid, filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52cec553",
   "metadata": {},
   "source": [
    "## Why is the model not capturing the high flow rates?\n",
    "Did we miss something in the training data?\n",
    "\n",
    "Lets look at the hydrograph from the training data to see if we can identify a reason for the poor predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "d8a83425",
   "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>flow_cfs</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>datetime</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2021-01-01</th>\n",
       "      <td>1.022609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-02</th>\n",
       "      <td>1.020000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-03</th>\n",
       "      <td>1.026146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-04</th>\n",
       "      <td>1.033021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021-01-05</th>\n",
       "      <td>1.080000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-12-27</th>\n",
       "      <td>1.100500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-12-28</th>\n",
       "      <td>1.337188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-12-29</th>\n",
       "      <td>1.141875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-12-30</th>\n",
       "      <td>1.090625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022-12-31</th>\n",
       "      <td>1.130000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>730 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            flow_cfs\n",
       "datetime            \n",
       "2021-01-01  1.022609\n",
       "2021-01-02  1.020000\n",
       "2021-01-03  1.026146\n",
       "2021-01-04  1.033021\n",
       "2021-01-05  1.080000\n",
       "...              ...\n",
       "2022-12-27  1.100500\n",
       "2022-12-28  1.337188\n",
       "2022-12-29  1.141875\n",
       "2022-12-30  1.090625\n",
       "2022-12-31  1.130000\n",
       "\n",
       "[730 rows x 1 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Get NWIS observations for the new time period\n",
    "start = '2021-1-1'\n",
    "end = '2022-12-31'\n",
    "NWISid = '10172200'\n",
    "\n",
    "flow = NWIS_retrieve(NWISid, start, end)\n",
    "\n",
    "#make the date the index for plotting and resampling\n",
    "flow['datetime'] = flow['value_time']\n",
    "flow.set_index('datetime', inplace = True)\n",
    "\n",
    "#clean the data\n",
    "flow = flow[flow['value'] > 0]\n",
    "\n",
    "#resample to daily resolution\n",
    "flow.rename(columns={'value':'flow_cfs'}, inplace = True)\n",
    "flow = pd.DataFrame(flow['flow_cfs'])\n",
    "flow = flow.resample(temporal_resample).mean()\n",
    "\n",
    "flow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "0d5786bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Plot the Hydrograph\n",
    "def hydrograph_plot(DF, flow):\n",
    "    fig, ax = plt.subplots(1,1, figsize=(8, 7))\n",
    "    fig.patch.set_facecolor('white')\n",
    "\n",
    "    #set min/max for y-axis of the predictions/observations\n",
    "    ymin = min(DF['flow_cfs'])*1.1\n",
    "    ymax = max(DF['flow_cfs'])*1.1\n",
    "\n",
    "\n",
    "    ax.plot(DF.index, DF['flow_cfs'],\n",
    "                   c='orange', label= 'Observed Streamflow')\n",
    "\n",
    "     # Add some parameters.\n",
    "    ax.set_title('Hydrograph', fontsize=16)\n",
    "    ax.set_xlabel('Time (DOY)', fontsize=14)\n",
    "    ax.set_ylabel('Streamflow (cfs)', fontsize=14,)\n",
    "    ax.set_ylim(0, ymax)\n",
    "    ax.legend(fontsize=14, loc='upper right')\n",
    "\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "7ac35fa2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hydrograph_plot(flow, 'flow_cfs')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21c152b2",
   "metadata": {},
   "source": [
    "## Are there any significant differences between the flow observations used in training vs. the 2023 water year?\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46392fa4",
   "metadata": {},
   "source": [
    "## Exercise\n",
    "We applied the GEE-ML model workflow to one stream. Apply the workflow to the ML model you developed in the previous chapter's exercise. \n",
    "Take 30 minutes to get new data from the respective sites, connect to your saved dataframe files, and predict!\n",
    "How did you model perform?\n",
    "\n",
    "\n",
    "**Make a copy of this Notebook with your initials_GEE_Workshop_CH3, e.g., RJ_GEE_Workshop_CH3. Save to the Exercises folder.**\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c959bcae",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "GEE_env",
   "language": "python",
   "name": "gee_env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}