analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "stupid-minority",
   "metadata": {},
   "source": [
    "# Analysis & Bias Correction of Forecast Data\n",
    "\n",
    "\n",
    "###### Contents\n",
    "    1. Exploratory Analysis\n",
    "    2. Data Preprocessing\n",
    "    3. Analysis of Best Model Forecast\n",
    "    4. Bias Correction of Forecasts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "falling-brush",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import seaborn as sns\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from analysis_functions import Preprocessing, Bias_Correction\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "fallen-shade",
   "metadata": {},
   "outputs": [],
   "source": [
    "# load in datasets\n",
    "# set index to observational time stamp\n",
    "BOS_obs = pd.read_csv(\"data/KBOS/obs.csv\",index_col=0,parse_dates=True)\n",
    "HLC_obs = pd.read_csv(\"data/KHLC/obs.csv\",index_col=0,parse_dates=True)\n",
    "\n",
    "# set indexes to forecast time stamp | convert DatetimeIndex tzUTC-obj to normal DatetimeIndex\n",
    "BOS_hrrr = pd.read_csv(\"data/KBOS/HRRR.csv\",index_col=2,parse_dates=True).tz_convert(None)\n",
    "BOS_gfs = pd.read_csv(\"data/KBOS/GFS.csv\",index_col=2,parse_dates=True).tz_convert(None)\n",
    "BOS_ecmwf = pd.read_csv(\"data/KBOS/ECMWF.csv\",index_col=2,parse_dates=True).tz_convert(None)\n",
    "HLC_hrrr = pd.read_csv(\"data/KHLC/HRRR.csv\",index_col=2,parse_dates=True).tz_convert(None)\n",
    "HLC_gfs = pd.read_csv(\"data/KHLC/GFS.csv\",index_col=2,parse_dates=True).tz_convert(None)\n",
    "HLC_ecmwf = pd.read_csv(\"data/KHLC/ECMWF.csv\",index_col=2,parse_dates=True).tz_convert(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "recreational-career",
   "metadata": {},
   "outputs": [],
   "source": [
    "# combine both KHLC and KBOS data for analysis\n",
    "obs = pd.concat([BOS_obs,HLC_obs])\n",
    "hrrr = pd.concat([BOS_hrrr,HLC_hrrr])\n",
    "gfs = pd.concat([BOS_gfs,HLC_gfs])\n",
    "ecmwf = pd.concat([BOS_ecmwf,HLC_ecmwf])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "forward-barbados",
   "metadata": {},
   "source": [
    "### 1. Exploratory Analysis of the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "sunset-trout",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Observational Date Range: 2021-06-01 00:00:00 to 2021-07-01 00:00:00\n"
     ]
    }
   ],
   "source": [
    "# find max/min range of observational dataset\n",
    "obs_max = obs.index.max()\n",
    "obs_min = obs.index.min()\n",
    "\n",
    "print(f\"Observational Date Range: {obs_min} to {obs_max}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "healthy-spencer",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing hours observational data: []\n",
      "Missing hours HRRR data: []\n",
      "Missing hours GFS data: ['2021-06-01T00:00:00.000000000' '2021-06-01T01:00:00.000000000']\n",
      "Missing hours ECMWF data: ['2021-06-01T00:00:00.000000000']\n"
     ]
    }
   ],
   "source": [
    "# check for missing hours from the data (only use the date-range of the observational data)\n",
    "missing_hours_obs = pd.date_range(start = '2021-06-1', end = '2021-07-1',freq='H').difference(obs.index)\n",
    "missing_hours_hrrr = pd.date_range(start = '2021-06-1', end = '2021-07-1',freq='H').difference(hrrr.index)\n",
    "missing_hours_gfs = pd.date_range(start = '2021-06-1', end = '2021-07-1',freq='H').difference(gfs.index)\n",
    "missing_hours_ecmwf = pd.date_range(start = '2021-06-1', end = '2021-07-1',freq='H').difference(ecmwf.index)\n",
    "\n",
    "print(f\"Missing hours observational data: {missing_hours_obs.values}\")\n",
    "print(f\"Missing hours HRRR data: {missing_hours_hrrr.values}\")\n",
    "print(f\"Missing hours GFS data: {missing_hours_gfs.values}\")\n",
    "print(f\"Missing hours ECMWF data: {missing_hours_ecmwf.values}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "transsexual-bottle",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "icao                0\n",
       "temperature_k      20\n",
       "dewpoint_k         21\n",
       "wind_speed_mps     24\n",
       "sfc_press_pa        2\n",
       "cloud_cover       849\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sum missing vals\n",
    "obs.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "equivalent-mechanism",
   "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_k</th>\n",
       "      <th>dewpoint_k</th>\n",
       "      <th>wind_speed_mps</th>\n",
       "      <th>sfc_press_pa</th>\n",
       "      <th>cloud_cover</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1416.000000</td>\n",
       "      <td>1415.000000</td>\n",
       "      <td>1412.000000</td>\n",
       "      <td>1434.000000</td>\n",
       "      <td>587.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>297.524364</td>\n",
       "      <td>288.580628</td>\n",
       "      <td>4.212921</td>\n",
       "      <td>97498.629470</td>\n",
       "      <td>64.650767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5.904731</td>\n",
       "      <td>4.258666</td>\n",
       "      <td>2.264981</td>\n",
       "      <td>3989.612292</td>\n",
       "      <td>23.862909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>284.150000</td>\n",
       "      <td>277.111111</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>92622.940000</td>\n",
       "      <td>43.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>292.150000</td>\n",
       "      <td>285.250000</td>\n",
       "      <td>2.570480</td>\n",
       "      <td>93528.560000</td>\n",
       "      <td>43.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>297.150000</td>\n",
       "      <td>289.138889</td>\n",
       "      <td>4.117238</td>\n",
       "      <td>100165.410000</td>\n",
       "      <td>43.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>302.150000</td>\n",
       "      <td>292.138889</td>\n",
       "      <td>5.659526</td>\n",
       "      <td>101383.640000</td>\n",
       "      <td>87.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>315.150000</td>\n",
       "      <td>297.222222</td>\n",
       "      <td>15.945917</td>\n",
       "      <td>102974.100000</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       temperature_k   dewpoint_k  wind_speed_mps   sfc_press_pa  cloud_cover\n",
       "count    1416.000000  1415.000000     1412.000000    1434.000000   587.000000\n",
       "mean      297.524364   288.580628        4.212921   97498.629470    64.650767\n",
       "std         5.904731     4.258666        2.264981    3989.612292    23.862909\n",
       "min       284.150000   277.111111        0.000000   92622.940000    43.750000\n",
       "25%       292.150000   285.250000        2.570480   93528.560000    43.750000\n",
       "50%       297.150000   289.138889        4.117238  100165.410000    43.750000\n",
       "75%       302.150000   292.138889        5.659526  101383.640000    87.500000\n",
       "max       315.150000   297.222222       15.945917  102974.100000   100.000000"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# obs descr. statistics\n",
    "obs.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "contained-cargo",
   "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>fcst_dt</th>\n",
       "      <th>sfc_air_temperature</th>\n",
       "      <th>sfc_dewpoint</th>\n",
       "      <th>v_wind_10m</th>\n",
       "      <th>u_wind_10m</th>\n",
       "      <th>sfc_pressure</th>\n",
       "      <th>sea_level_pressure</th>\n",
       "      <th>cloud_cover</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "      <td>5750.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>24.931826</td>\n",
       "      <td>22.764148</td>\n",
       "      <td>14.934558</td>\n",
       "      <td>1.406369</td>\n",
       "      <td>0.316416</td>\n",
       "      <td>1012.643201</td>\n",
       "      <td>974.494766</td>\n",
       "      <td>53.615117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>13.608325</td>\n",
       "      <td>5.947356</td>\n",
       "      <td>3.863736</td>\n",
       "      <td>3.111285</td>\n",
       "      <td>2.824572</td>\n",
       "      <td>6.387029</td>\n",
       "      <td>40.771710</td>\n",
       "      <td>45.095790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>9.150000</td>\n",
       "      <td>-0.230000</td>\n",
       "      <td>-13.160000</td>\n",
       "      <td>-11.260000</td>\n",
       "      <td>997.119995</td>\n",
       "      <td>923.440002</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>13.000000</td>\n",
       "      <td>18.120001</td>\n",
       "      <td>12.300000</td>\n",
       "      <td>-0.250000</td>\n",
       "      <td>-1.927500</td>\n",
       "      <td>1008.099976</td>\n",
       "      <td>934.549988</td>\n",
       "      <td>1.880000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>25.000000</td>\n",
       "      <td>21.790001</td>\n",
       "      <td>15.370000</td>\n",
       "      <td>1.640000</td>\n",
       "      <td>0.690000</td>\n",
       "      <td>1012.419983</td>\n",
       "      <td>972.414978</td>\n",
       "      <td>65.925003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>37.000000</td>\n",
       "      <td>26.830000</td>\n",
       "      <td>18.120001</td>\n",
       "      <td>3.450000</td>\n",
       "      <td>2.550000</td>\n",
       "      <td>1016.687500</td>\n",
       "      <td>1014.119995</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>48.000000</td>\n",
       "      <td>42.340000</td>\n",
       "      <td>24.330000</td>\n",
       "      <td>11.330000</td>\n",
       "      <td>13.510000</td>\n",
       "      <td>1031.099976</td>\n",
       "      <td>1030.949951</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           fcst_dt  sfc_air_temperature  sfc_dewpoint   v_wind_10m  \\\n",
       "count  5750.000000          5750.000000   5750.000000  5750.000000   \n",
       "mean     24.931826            22.764148     14.934558     1.406369   \n",
       "std      13.608325             5.947356      3.863736     3.111285   \n",
       "min       0.000000             9.150000     -0.230000   -13.160000   \n",
       "25%      13.000000            18.120001     12.300000    -0.250000   \n",
       "50%      25.000000            21.790001     15.370000     1.640000   \n",
       "75%      37.000000            26.830000     18.120001     3.450000   \n",
       "max      48.000000            42.340000     24.330000    11.330000   \n",
       "\n",
       "        u_wind_10m  sfc_pressure  sea_level_pressure  cloud_cover  \n",
       "count  5750.000000   5750.000000         5750.000000  5750.000000  \n",
       "mean      0.316416   1012.643201          974.494766    53.615117  \n",
       "std       2.824572      6.387029           40.771710    45.095790  \n",
       "min     -11.260000    997.119995          923.440002     0.000000  \n",
       "25%      -1.927500   1008.099976          934.549988     1.880000  \n",
       "50%       0.690000   1012.419983          972.414978    65.925003  \n",
       "75%       2.550000   1016.687500         1014.119995   100.000000  \n",
       "max      13.510000   1031.099976         1030.949951   100.000000  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# hrrr desc. stats\n",
    "hrrr.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "continued-tutorial",
   "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>fcst_dt</th>\n",
       "      <th>sfc_air_temperature</th>\n",
       "      <th>sfc_dewpoint</th>\n",
       "      <th>v_wind_10m</th>\n",
       "      <th>u_wind_10m</th>\n",
       "      <th>sfc_pressure</th>\n",
       "      <th>sea_level_pressure</th>\n",
       "      <th>cloud_cover</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "      <td>25250.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>142.852277</td>\n",
       "      <td>21.201510</td>\n",
       "      <td>14.250137</td>\n",
       "      <td>0.098859</td>\n",
       "      <td>-0.594920</td>\n",
       "      <td>1015.650401</td>\n",
       "      <td>966.594631</td>\n",
       "      <td>37.610523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>110.472276</td>\n",
       "      <td>3.610148</td>\n",
       "      <td>5.078368</td>\n",
       "      <td>2.810436</td>\n",
       "      <td>2.509961</td>\n",
       "      <td>3.386230</td>\n",
       "      <td>48.938484</td>\n",
       "      <td>39.643010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>7.010000</td>\n",
       "      <td>-3.240000</td>\n",
       "      <td>-12.660000</td>\n",
       "      <td>-12.430000</td>\n",
       "      <td>1002.460022</td>\n",
       "      <td>905.940002</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>53.000000</td>\n",
       "      <td>19.180000</td>\n",
       "      <td>10.050000</td>\n",
       "      <td>-1.950000</td>\n",
       "      <td>-1.900000</td>\n",
       "      <td>1013.140015</td>\n",
       "      <td>918.099976</td>\n",
       "      <td>0.800000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>105.000000</td>\n",
       "      <td>21.850000</td>\n",
       "      <td>14.650000</td>\n",
       "      <td>-0.215000</td>\n",
       "      <td>-0.320000</td>\n",
       "      <td>1015.849976</td>\n",
       "      <td>963.760010</td>\n",
       "      <td>18.600000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>231.000000</td>\n",
       "      <td>23.450001</td>\n",
       "      <td>18.770000</td>\n",
       "      <td>2.010000</td>\n",
       "      <td>0.810000</td>\n",
       "      <td>1018.219971</td>\n",
       "      <td>1015.750000</td>\n",
       "      <td>78.300003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>384.000000</td>\n",
       "      <td>37.360001</td>\n",
       "      <td>23.250000</td>\n",
       "      <td>11.770000</td>\n",
       "      <td>13.170000</td>\n",
       "      <td>1025.280029</td>\n",
       "      <td>1023.880005</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            fcst_dt  sfc_air_temperature  sfc_dewpoint    v_wind_10m  \\\n",
       "count  25250.000000         25250.000000  25250.000000  25250.000000   \n",
       "mean     142.852277            21.201510     14.250137      0.098859   \n",
       "std      110.472276             3.610148      5.078368      2.810436   \n",
       "min        2.000000             7.010000     -3.240000    -12.660000   \n",
       "25%       53.000000            19.180000     10.050000     -1.950000   \n",
       "50%      105.000000            21.850000     14.650000     -0.215000   \n",
       "75%      231.000000            23.450001     18.770000      2.010000   \n",
       "max      384.000000            37.360001     23.250000     11.770000   \n",
       "\n",
       "         u_wind_10m  sfc_pressure  sea_level_pressure   cloud_cover  \n",
       "count  25250.000000  25250.000000        25250.000000  25250.000000  \n",
       "mean      -0.594920   1015.650401          966.594631     37.610523  \n",
       "std        2.509961      3.386230           48.938484     39.643010  \n",
       "min      -12.430000   1002.460022          905.940002      0.000000  \n",
       "25%       -1.900000   1013.140015          918.099976      0.800000  \n",
       "50%       -0.320000   1015.849976          963.760010     18.600000  \n",
       "75%        0.810000   1018.219971         1015.750000     78.300003  \n",
       "max       13.170000   1025.280029         1023.880005    100.000000  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# gfs stats\n",
    "gfs.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "experienced-sight",
   "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>fcst_dt</th>\n",
       "      <th>sfc_air_temperature</th>\n",
       "      <th>sfc_dewpoint</th>\n",
       "      <th>v_wind_10m</th>\n",
       "      <th>u_wind_10m</th>\n",
       "      <th>sfc_pressure</th>\n",
       "      <th>sea_level_pressure</th>\n",
       "      <th>cloud_cover</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "      <td>15122.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>75.389499</td>\n",
       "      <td>22.747916</td>\n",
       "      <td>15.249486</td>\n",
       "      <td>1.714359</td>\n",
       "      <td>0.456183</td>\n",
       "      <td>1012.594456</td>\n",
       "      <td>973.219649</td>\n",
       "      <td>47.865215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>58.219098</td>\n",
       "      <td>5.500115</td>\n",
       "      <td>3.588779</td>\n",
       "      <td>3.208443</td>\n",
       "      <td>2.699768</td>\n",
       "      <td>5.677388</td>\n",
       "      <td>39.544192</td>\n",
       "      <td>40.082115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>10.600000</td>\n",
       "      <td>-0.910000</td>\n",
       "      <td>-14.740000</td>\n",
       "      <td>-10.860000</td>\n",
       "      <td>992.820007</td>\n",
       "      <td>918.190002</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>31.000000</td>\n",
       "      <td>18.559999</td>\n",
       "      <td>12.652500</td>\n",
       "      <td>-0.170000</td>\n",
       "      <td>-1.490000</td>\n",
       "      <td>1008.530029</td>\n",
       "      <td>934.500000</td>\n",
       "      <td>3.910000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>62.000000</td>\n",
       "      <td>21.830000</td>\n",
       "      <td>15.490000</td>\n",
       "      <td>2.070000</td>\n",
       "      <td>0.450000</td>\n",
       "      <td>1012.570007</td>\n",
       "      <td>969.650024</td>\n",
       "      <td>44.529999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>102.000000</td>\n",
       "      <td>26.330000</td>\n",
       "      <td>18.197501</td>\n",
       "      <td>3.970000</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>1016.289978</td>\n",
       "      <td>1011.799988</td>\n",
       "      <td>92.190002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>240.000000</td>\n",
       "      <td>42.419998</td>\n",
       "      <td>23.340000</td>\n",
       "      <td>10.100000</td>\n",
       "      <td>11.540000</td>\n",
       "      <td>1030.719971</td>\n",
       "      <td>1029.160034</td>\n",
       "      <td>100.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            fcst_dt  sfc_air_temperature  sfc_dewpoint    v_wind_10m  \\\n",
       "count  15122.000000         15122.000000  15122.000000  15122.000000   \n",
       "mean      75.389499            22.747916     15.249486      1.714359   \n",
       "std       58.219098             5.500115      3.588779      3.208443   \n",
       "min        1.000000            10.600000     -0.910000    -14.740000   \n",
       "25%       31.000000            18.559999     12.652500     -0.170000   \n",
       "50%       62.000000            21.830000     15.490000      2.070000   \n",
       "75%      102.000000            26.330000     18.197501      3.970000   \n",
       "max      240.000000            42.419998     23.340000     10.100000   \n",
       "\n",
       "         u_wind_10m  sfc_pressure  sea_level_pressure   cloud_cover  \n",
       "count  15122.000000  15122.000000        15122.000000  15122.000000  \n",
       "mean       0.456183   1012.594456          973.219649     47.865215  \n",
       "std        2.699768      5.677388           39.544192     40.082115  \n",
       "min      -10.860000    992.820007          918.190002      0.000000  \n",
       "25%       -1.490000   1008.530029          934.500000      3.910000  \n",
       "50%        0.450000   1012.570007          969.650024     44.529999  \n",
       "75%        2.650000   1016.289978         1011.799988     92.190002  \n",
       "max       11.540000   1030.719971         1029.160034    100.000000  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ecmwf stats\n",
    "ecmwf.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "about-shelf",
   "metadata": {},
   "source": [
    "###### Observations & Tasks \n",
    " - Data is hourly collected  \n",
    " - Time Range: 00z June 1 2021 - 00z July 1 2021\n",
    " - GFS & ECMWF are missing the 1st hourly observation\n",
    " - Observations need to be preprocessed (i.e. convert all data to the same units, calculate wind speeds, etc)\n",
    " - Missing vals in the observation set\n",
    " - Only interested in temperature, dewpoint, and wind speed\n",
    "\n",
    "Max forecast hour dt is 48 (HRRR), which means in order to make a consistent analysis across all 3 models, only the forecast hours of 0-48z can be considered in order to preserve the consistency.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "weighted-allergy",
   "metadata": {},
   "source": [
    "### 2. Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "proud-melbourne",
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert observational data to correct units\n",
    "obs_processed = obs\n",
    "\n",
    "# convert Kelvin to Celsius | mph to m/s | pascal to mb\n",
    "obs_processed[\"temperature_k\"] = obs_processed[\"temperature_k\"] - 273.15\n",
    "obs_processed[\"dewpoint_k\"] = obs_processed[\"dewpoint_k\"] - 273.15\n",
    "obs_processed[\"wind_speed_mps\"] = obs_processed[\"wind_speed_mps\"] * 0.44704\n",
    "\n",
    "# rename columns\n",
    "dict = {\"temperature_k\":\"temperature_c\",\"dewpoint_k\":\"dewpoint_c\",\"wind_speed_mps\":\"wind_speed_ms\"}\n",
    "obs_processed = obs_processed.rename(columns=dict)\n",
    "\n",
    "# remove cloud cover/sfcpress\n",
    "obs_processed = obs_processed.drop([\"sfc_press_pa\",\"cloud_cover\"],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "numeric-superintendent",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate wind speed for model data\n",
    "gfs_processed = Preprocessing.calc_windspeed(gfs)\n",
    "hrrr_processed = Preprocessing.calc_windspeed(hrrr)\n",
    "ecmwf_processed = Preprocessing.calc_windspeed(ecmwf)\n",
    "\n",
    "# only select variables of interest\n",
    "gfs_processed = gfs_processed.drop([\"sfc_pressure\",\"sea_level_pressure\",\"cloud_cover\"],axis=1)\n",
    "hrrr_processed = hrrr_processed.drop([\"sfc_pressure\",\"sea_level_pressure\",\"cloud_cover\"],axis=1)\n",
    "ecmwf_processed = ecmwf_processed.drop([\"sfc_pressure\",\"sea_level_pressure\",\"cloud_cover\"],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "frozen-alexander",
   "metadata": {},
   "outputs": [],
   "source": [
    "# separate into location dataframes in order to merge together with observations\n",
    "BOS_obs_processed = obs_processed[obs_processed.icao==\"KBOS\"]\n",
    "HLC_obs_processed = obs_processed[obs_processed.icao==\"KHLC\"]\n",
    "\n",
    "# gfs\n",
    "BOS_gfs_processed = gfs_processed[gfs_processed.pt==\"KBOS\"]\n",
    "HLC_gfs_processed = gfs_processed[gfs_processed.pt==\"KHLC\"]\n",
    "\n",
    "#hrrr\n",
    "BOS_hrrr_processed = hrrr_processed[hrrr_processed.pt==\"KBOS\"]\n",
    "HLC_hrrr_processed = hrrr_processed[hrrr_processed.pt==\"KHLC\"]\n",
    "\n",
    "# ecmwf\n",
    "BOS_ecmwf_processed = ecmwf_processed[ecmwf_processed.pt==\"KBOS\"]\n",
    "HLC_ecmwf_processed = ecmwf_processed[ecmwf_processed.pt==\"KHLC\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "associate-season",
   "metadata": {},
   "source": [
    "### 3. Analysis of Best Forecast"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "developmental-restaurant",
   "metadata": {},
   "outputs": [],
   "source": [
    "# combine forecast data for each location with observational data to prepare for analysis\n",
    "gfs_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_gfs_processed,HLC_gfs_processed],[BOS_obs_processed,HLC_obs_processed],48)\n",
    "hrrr_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_hrrr_processed,HLC_hrrr_processed],[BOS_obs_processed,HLC_obs_processed],48)\n",
    "ecmwf_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_ecmwf_processed,HLC_ecmwf_processed],[BOS_obs_processed,HLC_obs_processed],48)\n",
    "\n",
    "# Remove missing values from the dataframes \n",
    "gfs_analysis_df = gfs_analysis_df.dropna()\n",
    "hrrr_analysis_df = hrrr_analysis_df.dropna()\n",
    "ecmwf_analysis_df = ecmwf_analysis_df.dropna()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "needed-knowing",
   "metadata": {},
   "source": [
    "### Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "rocky-blend",
   "metadata": {},
   "outputs": [],
   "source": [
    "KBOS_ecmwf_rmse = mean_squared_error(ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KBOS\"].sfc_air_temperature,ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KBOS\"].temperature_c,squared=False)\n",
    "KHLC_ecmwf_rmse = mean_squared_error(ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KHLC\"].sfc_air_temperature,ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KHLC\"].temperature_c,squared=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "variable-botswana",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Location RMSE scores by model\n",
    "hrrr_perf_loc = Preprocessing.calc_rmse_by_station(hrrr_analysis_df,\"HRRR\")\n",
    "gfs_perf_loc = Preprocessing.calc_rmse_by_station(gfs_analysis_df,\"GFS\")\n",
    "ecmwf_perf_loc = Preprocessing.calc_rmse_by_station(ecmwf_analysis_df,\"ECMWF\")\n",
    "\n",
    "# Variable RMSE scores by model\n",
    "hrrr_perf_variables = Preprocessing.calc_rmse_by_variable(hrrr_analysis_df,\"HRRR\")\n",
    "gfs_perf_variables = Preprocessing.calc_rmse_by_variable(gfs_analysis_df,\"GFS\")\n",
    "ecmwf_perf_variables = Preprocessing.calc_rmse_by_variable(ecmwf_analysis_df,\"ECMWF\")\n",
    "\n",
    "\n",
    "# create dataframe of all model performances\n",
    "model_perf_loc= pd.concat([hrrr_perf_loc,gfs_perf_loc,ecmwf_perf_loc])\n",
    "model_perf_variables= pd.concat([hrrr_perf_variables,gfs_perf_variables,ecmwf_perf_variables])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "major-constitution",
   "metadata": {},
   "outputs": [],
   "source": [
    "# if figures folder doesn't exist create it\n",
    "directory = \"figures/\"\n",
    "if not os.path.exists(directory):\n",
    "    os.makedirs(directory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "creative-compatibility",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "sns.set(rc={'figure.figsize':(8,5)})\n",
    "fig1 = sns.barplot(x=\"model\",y=\"RMSE\",hue=\"location\",data=model_perf_loc,ci=None,palette=[\"deepskyblue\",\"lightgrey\"])\n",
    "fig1.set_title(\"Overall Model Performance Error for KBOS & KHLC\")\n",
    "fig1 = fig1.get_figure()\n",
    "fig1.savefig(\"figures/overall_model_perf_by_station.png\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "adult-denial",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "sns.set(rc={'figure.figsize':(8,5)})\n",
    "fig2 = sns.barplot(x=\"model\",y=\"RMSE\",hue=\"variable\",data=model_perf_variables,ci=None,palette=[\"cyan\",\"silver\",\"navy\"])\n",
    "fig2.set_title(\"Overall Model Performance Error by Variable for KBOS & KHLC\")\n",
    "fig2 = fig2.get_figure()\n",
    "fig2.savefig(\"figures/overall_model_perf_by_variable.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "alpine-madrid",
   "metadata": {},
   "source": [
    "By variable we see that temperature prediction has the highest error. Dewpoint is the best in both the Euro & HRRR, however wind speed is the best performer for the GFS. Surprisingly, wind prediction with the GFS is on par with the HRRR & Euro, despite the very poor temp & dewpoint performance."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "weighted-meditation",
   "metadata": {},
   "source": [
    "### 4. Bias Correction of Forecasts\n",
    "\n",
    "In order to bias-correct the forecast data, a machine learning algorithm will be implemented. The reason that a machine learning approach will be taken is in my experience of using a simple mean diurnal cycle bias correction (subtracting the mean forecast by the mean observed between hours 0-23), this method is very difficult to scale up. If data is collected on seasonal/yearly basis, then a seasonal bias correction will need to occur, and even perhaps a yearly bias correction. ML algorithms are flexible, easier to scale up, and more often than not highly effective. \n",
    "\n",
    "For temperature and wind speed bias correction, I have found that Gradient Boosted Trees or XGBoost perform often better than regular regression techniques. Therefore GBT from Sci-Kit Learn was selected.\n",
    "\n",
    "###### Experiment\n",
    "\n",
    "Two variables will be used as inputs, fcst_dt and the forecast variable. The METAR observation variables will be used as the targets. Fcst_dt will represent the temporal variation in the data, particularly how bias increases/decreases as the forecast increases in time. These two variables will be used to train the GBT model. In addition, a bias-correction model will be trained for each variable for each location, and this is to account for the typical differences in bias between two locations (KBOS & KHLC).\n",
    "\n",
    "My hypothesis is that if GBT is used for bias correction, then temperature/dewpoint/windspeed will all significantly improve in respect to the observational data. However, wind speed will improve the least.\n",
    "\n",
    "I will first train/test HRRR temperature forecasts for KBOS, then dew point, and windspeed and visualize the results. Then I will perform the training/testing for each model for each location and collect the results.\n",
    "\n",
    "\n",
    "\n",
    "##### Temperature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "known-driving",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Focus on KBOS first\n",
    "KBOS_hrrr_analysis_df = hrrr_analysis_df[hrrr_analysis_df.pt==\"KBOS\"].dropna()\n",
    "#KBOS_hrrr_analysis_df = KBOS_gfs_analysis_df\n",
    "\n",
    "# use fcst_dt and model forecasted temperatures as inputs to the model, and the obs data as the target\n",
    "X_KBOS_hrrr = KBOS_hrrr_analysis_df[[\"fcst_dt\",\"sfc_air_temperature\"]]\n",
    "Y_KBOS_hrrr = KBOS_hrrr_analysis_df[\"temperature_c\"]\n",
    "\n",
    "# split into training/testing use 80/20 split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_KBOS_hrrr, Y_KBOS_hrrr, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "square-signature",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HRRR Forecast R2 Score: 0.46490406592092026\n",
      "HRRR Forecast RMSE: 3.917417295405916\n",
      "Bias Corrected HRRR R2 Score: 0.8726296447379525\n",
      "Bias Corrected HRRR RMSE: 1.9112530478085994\n"
     ]
    }
   ],
   "source": [
    "# train a Gradient Boosted Tree algorithm and print the results\n",
    "gbr = GradientBoostingRegressor(n_estimators=40)\n",
    "gbr.fit(X_train,y_train)\n",
    "pred = gbr.predict(X_test)\n",
    "\n",
    "# print results\n",
    "print(f\"HRRR Forecast R2 Score: {r2_score(y_test,X_test.sfc_air_temperature)}\")\n",
    "print(f\"HRRR Forecast RMSE: {mean_squared_error(y_test,X_test.sfc_air_temperature,squared=False)}\")\n",
    "\n",
    "print(f\"Bias Corrected HRRR R2 Score: {r2_score(y_test,pred)}\")\n",
    "print(f\"Bias Corrected HRRR RMSE: {mean_squared_error(y_test,pred,squared=False)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "legal-binding",
   "metadata": {},
   "outputs": [],
   "source": [
    "verify_temp = pd.concat([X_test,y_test],axis=1)\n",
    "verify_temp[\"pred\"] = pred\n",
    "verify_temp_mean = verify_temp.groupby(\"fcst_dt\").mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "anticipated-canyon",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"] = (10,5)\n",
    "plt.plot(verify_temp_mean.index,verify_temp_mean.sfc_air_temperature,label=\"HRRR Forecast\",c=\"green\")\n",
    "plt.plot(verify_temp_mean.index,verify_temp_mean.pred,label=\"Bias-Corrected Forecast\",c=\"blue\")\n",
    "plt.plot(verify_temp_mean.index,verify_temp_mean.temperature_c,'--',label=\"Observed\",c=\"red\")\n",
    "plt.legend()\n",
    "plt.title(\"KBOS HRRR Bias-Corrected Temperature Average Performance\")\n",
    "plt.xlabel(\"Forecast Hour (dt)\")\n",
    "plt.ylabel(\"Temperature (C)\")\n",
    "plt.xlim([0,48])\n",
    "plt.savefig(\"figures/KBOS_HRRR_bias_corrected_temp.png\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "increasing-suggestion",
   "metadata": {},
   "source": [
    "There is a clear improvement on the forecast as can be seen in the figure above. \n",
    "\n",
    " \n",
    "\n",
    "Overall excellent results."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "synthetic-intent",
   "metadata": {},
   "source": [
    "###### Dewpoint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "restricted-enforcement",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Focus on KBOS first\n",
    "KBOS_hrrr_analysis_df = hrrr_analysis_df[hrrr_analysis_df.pt==\"KBOS\"].dropna()\n",
    "\n",
    "# use fcst_dt and model forecasted dewpoints as inputs to the model, and the obs data as the target\n",
    "X_KBOS_hrrr = KBOS_hrrr_analysis_df[[\"fcst_dt\",\"sfc_dewpoint\"]]\n",
    "Y_KBOS_hrrr = KBOS_hrrr_analysis_df[\"dewpoint_c\"]\n",
    "\n",
    "# split into training/testing use 80/20 split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_KBOS_hrrr, Y_KBOS_hrrr, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "destroyed-antibody",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HRRR Forecast R2 Score: 0.8460873187416739\n",
      "HRRR Forecast RMSE: 1.8426903322511752\n",
      "Bias Corrected HRRR R2 Score: 0.8583843271313166\n",
      "Bias Corrected HRRR RMSE: 1.7675463679069916\n"
     ]
    }
   ],
   "source": [
    "# train a Gradient Boosted Tree algorithm and print the results\n",
    "gbr = GradientBoostingRegressor(n_estimators=80)\n",
    "gbr.fit(X_train,y_train)\n",
    "pred = gbr.predict(X_test)\n",
    "\n",
    "# print results\n",
    "print(f\"HRRR Forecast R2 Score: {r2_score(y_test,X_test.sfc_dewpoint)}\")\n",
    "print(f\"HRRR Forecast RMSE: {mean_squared_error(y_test,X_test.sfc_dewpoint,squared=False)}\")\n",
    "\n",
    "print(f\"Bias Corrected HRRR R2 Score: {r2_score(y_test,pred)}\")\n",
    "print(f\"Bias Corrected HRRR RMSE: {mean_squared_error(y_test,pred,squared=False)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "intermediate-lighting",
   "metadata": {},
   "outputs": [],
   "source": [
    "verify_dpt = pd.concat([X_test,y_test],axis=1)\n",
    "verify_dpt[\"pred\"] = pred\n",
    "verify_dpt_mean = verify_dpt.groupby(\"fcst_dt\").mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "geographic-partition",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"] = (10,5)\n",
    "plt.plot(verify_dpt_mean.index,verify_dpt_mean.sfc_dewpoint,label=\"HRRR Forecast\",c=\"green\")\n",
    "plt.plot(verify_dpt_mean.index,verify_dpt_mean.pred,label=\"Bias-Corrected Forecast\",c=\"blue\")\n",
    "plt.plot(verify_dpt_mean.index,verify_dpt_mean.dewpoint_c,'--',label=\"Observed\",c=\"red\")\n",
    "plt.legend()\n",
    "plt.title(\"KBOS HRRR Bias-Corrected Dew Point Average Performance\")\n",
    "plt.xlabel(\"Forecast Hour (dt)\")\n",
    "plt.ylabel(\"Dew Point (C)\")\n",
    "plt.xlim([0,48])\n",
    "plt.savefig(\"figures/KBOS_HRRR_bias_corrected_dpt.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "about-aside",
   "metadata": {},
   "source": [
    "Implementation of the GBT resulted in slightly improved results in comparison to the HRRR forecasted values, however it is worth noting that the HRRR already forecasted dewpoints accurately. Slight improvements were found in R2-score and RMSE.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "spare-breast",
   "metadata": {},
   "source": [
    "###### Wind Speed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "aerial-resident",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Focus on KBOS first\n",
    "KBOS_hrrr_analysis_df = hrrr_analysis_df[hrrr_analysis_df.pt==\"KBOS\"].dropna()\n",
    "\n",
    "# use fcst_dt and model forecasted dewpoints as inputs to the model, and the obs data as the target\n",
    "X_KBOS_hrrr = KBOS_hrrr_analysis_df[[\"fcst_dt\",\"wind_speed_ms\"]]\n",
    "Y_KBOS_hrrr = KBOS_hrrr_analysis_df[\"wind_10m\"]\n",
    "\n",
    "# split into training/testing use 80/20 split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_KBOS_hrrr, Y_KBOS_hrrr, test_size=0.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "modified-wellington",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HRRR Forecast R2 Score: -1.2017527399349905\n",
      "HRRR Forecast RMSE: 2.422061356807932\n",
      "Bias Corrected HRRR R2 Score: 0.38872945494886857\n",
      "Bias Corrected HRRR RMSE: 1.2761966567359373\n"
     ]
    }
   ],
   "source": [
    "# train a Gradient Boosted Tree algorithm and print the results\n",
    "gbr = GradientBoostingRegressor(n_estimators=40)\n",
    "gbr.fit(X_train,y_train)\n",
    "pred = gbr.predict(X_test)\n",
    "\n",
    "# print results\n",
    "print(f\"HRRR Forecast R2 Score: {r2_score(y_test,X_test.wind_speed_ms)}\")\n",
    "print(f\"HRRR Forecast RMSE: {mean_squared_error(y_test,X_test.wind_speed_ms,squared=False)}\")\n",
    "\n",
    "print(f\"Bias Corrected HRRR R2 Score: {r2_score(y_test,pred)}\")\n",
    "print(f\"Bias Corrected HRRR RMSE: {mean_squared_error(y_test,pred,squared=False)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "documented-utility",
   "metadata": {},
   "outputs": [],
   "source": [
    "verify_w10 = pd.concat([X_test,y_test],axis=1)\n",
    "verify_w10[\"pred\"] = pred\n",
    "verify_w10_mean = verify_w10.groupby(\"fcst_dt\").mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "ahead-death",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams[\"figure.figsize\"] = (10,5)\n",
    "plt.plot(verify_w10_mean.index,verify_w10_mean.wind_speed_ms,label=\"HRRR Forecast\",c=\"green\")\n",
    "plt.plot(verify_w10_mean.index,verify_w10_mean.pred,label=\"Bias-Corrected Forecast\",c=\"blue\")\n",
    "plt.plot(verify_w10_mean.index,verify_w10_mean.wind_10m,'--',label=\"Observed\",c=\"red\")\n",
    "plt.legend()\n",
    "plt.title(\"KBOS HRRR Bias-Corrected 10m Wind Average Performance\")\n",
    "plt.xlabel(\"Forecast Hour (dt)\")\n",
    "plt.ylabel(\"Speed (m/s)\")\n",
    "plt.xlim([0,48])\n",
    "plt.savefig(\"figures/KBOS_HRRR_bias_corrected_w10.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "selected-kazakhstan",
   "metadata": {},
   "source": [
    "Again outstanding improvements were found with wind speed, contrary to my hypothesis which thought this area would see the least improvement."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "trying-ending",
   "metadata": {},
   "source": [
    "###### Correct remaining models and locations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "sublime-chambers",
   "metadata": {},
   "outputs": [],
   "source": [
    "# redefine the analysis dataframes do contain all forecast hours rather than 0-48z only\n",
    "gfs_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_gfs_processed,HLC_gfs_processed],[BOS_obs_processed,HLC_obs_processed],None)\n",
    "hrrr_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_hrrr_processed,HLC_hrrr_processed],[BOS_obs_processed,HLC_obs_processed],None)\n",
    "ecmwf_analysis_df = Preprocessing.merge_forecast_with_obs([BOS_ecmwf_processed,HLC_ecmwf_processed],[BOS_obs_processed,HLC_obs_processed],None)\n",
    "\n",
    "\n",
    "# Define the remaining models and stations\n",
    "KBOS_hrrr_analysis_df = hrrr_analysis_df[hrrr_analysis_df.pt==\"KBOS\"].dropna()\n",
    "KHLC_hrrr_analysis_df = hrrr_analysis_df[hrrr_analysis_df.pt==\"KHLC\"].dropna()\n",
    "KBOS_gfs_analysis_df = gfs_analysis_df[gfs_analysis_df.pt==\"KBOS\"].dropna()\n",
    "KHLC_gfs_analysis_df = gfs_analysis_df[gfs_analysis_df.pt==\"KHLC\"].dropna()\n",
    "KBOS_ecmwf_analysis_df = ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KBOS\"].dropna()\n",
    "KHLC_ecmwf_analysis_df = ecmwf_analysis_df[ecmwf_analysis_df.pt==\"KHLC\"].dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "confidential-emergency",
   "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>variable</th>\n",
       "      <th>rmse</th>\n",
       "      <th>method</th>\n",
       "      <th>model</th>\n",
       "      <th>location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>1.911253</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>1.761337</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.276197</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>3.917417</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>1.842690</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>2.422061</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>1.790336</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>1.715269</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.558331</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>1.842512</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>2.227792</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>2.714450</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>HRRR</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>5.302645</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>4.696989</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.116399</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>7.822905</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>8.518990</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.432822</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>5.838239</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>3.182331</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>2.344607</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>7.002366</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>4.477176</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>3.419336</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>GFS</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>2.574287</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>2.533256</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.431000</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>5.121763</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>2.836097</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>2.912786</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KBOS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>temp</td>\n",
       "      <td>2.738627</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>dpt</td>\n",
       "      <td>2.338470</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>w10</td>\n",
       "      <td>1.427217</td>\n",
       "      <td>bias-corrected</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>temp</td>\n",
       "      <td>2.809409</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>dpt</td>\n",
       "      <td>3.069157</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>w10</td>\n",
       "      <td>2.534300</td>\n",
       "      <td>raw forecast</td>\n",
       "      <td>ECMWF</td>\n",
       "      <td>KHLC</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  variable      rmse          method  model location\n",
       "0     temp  1.911253  bias-corrected   HRRR     KBOS\n",
       "1      dpt  1.761337  bias-corrected   HRRR     KBOS\n",
       "2      w10  1.276197  bias-corrected   HRRR     KBOS\n",
       "3     temp  3.917417    raw forecast   HRRR     KBOS\n",
       "4      dpt  1.842690    raw forecast   HRRR     KBOS\n",
       "5      w10  2.422061    raw forecast   HRRR     KBOS\n",
       "0     temp  1.790336  bias-corrected   HRRR     KHLC\n",
       "1      dpt  1.715269  bias-corrected   HRRR     KHLC\n",
       "2      w10  1.558331  bias-corrected   HRRR     KHLC\n",
       "3     temp  1.842512    raw forecast   HRRR     KHLC\n",
       "4      dpt  2.227792    raw forecast   HRRR     KHLC\n",
       "5      w10  2.714450    raw forecast   HRRR     KHLC\n",
       "0     temp  5.302645  bias-corrected    GFS     KBOS\n",
       "1      dpt  4.696989  bias-corrected    GFS     KBOS\n",
       "2      w10  1.116399  bias-corrected    GFS     KBOS\n",
       "3     temp  7.822905    raw forecast    GFS     KBOS\n",
       "4      dpt  8.518990    raw forecast    GFS     KBOS\n",
       "5      w10  1.432822    raw forecast    GFS     KBOS\n",
       "0     temp  5.838239  bias-corrected    GFS     KHLC\n",
       "1      dpt  3.182331  bias-corrected    GFS     KHLC\n",
       "2      w10  2.344607  bias-corrected    GFS     KHLC\n",
       "3     temp  7.002366    raw forecast    GFS     KHLC\n",
       "4      dpt  4.477176    raw forecast    GFS     KHLC\n",
       "5      w10  3.419336    raw forecast    GFS     KHLC\n",
       "0     temp  2.574287  bias-corrected  ECMWF     KBOS\n",
       "1      dpt  2.533256  bias-corrected  ECMWF     KBOS\n",
       "2      w10  1.431000  bias-corrected  ECMWF     KBOS\n",
       "3     temp  5.121763    raw forecast  ECMWF     KBOS\n",
       "4      dpt  2.836097    raw forecast  ECMWF     KBOS\n",
       "5      w10  2.912786    raw forecast  ECMWF     KBOS\n",
       "0     temp  2.738627  bias-corrected  ECMWF     KHLC\n",
       "1      dpt  2.338470  bias-corrected  ECMWF     KHLC\n",
       "2      w10  1.427217  bias-corrected  ECMWF     KHLC\n",
       "3     temp  2.809409    raw forecast  ECMWF     KHLC\n",
       "4      dpt  3.069157    raw forecast  ECMWF     KHLC\n",
       "5      w10  2.534300    raw forecast  ECMWF     KHLC"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# calculate bias-corrected performance from all models/stations\n",
    "KBOS_hrrr_results = Bias_Correction.gradient_boosted_trees_bias_correction(KBOS_hrrr_analysis_df,\"KBOS\",\"HRRR\")\n",
    "KHLC_hrrr_results = Bias_Correction.gradient_boosted_trees_bias_correction(KHLC_hrrr_analysis_df,\"KHLC\",\"HRRR\")\n",
    "KBOS_gfs_results = Bias_Correction.gradient_boosted_trees_bias_correction(KBOS_gfs_analysis_df,\"KBOS\",\"GFS\")\n",
    "KHLC_gfs_results = Bias_Correction.gradient_boosted_trees_bias_correction(KHLC_gfs_analysis_df,\"KHLC\",\"GFS\")\n",
    "KBOS_ecmwf_results = Bias_Correction.gradient_boosted_trees_bias_correction(KBOS_ecmwf_analysis_df,\"KBOS\",\"ECMWF\")\n",
    "KHLC_ecmwf_results = Bias_Correction.gradient_boosted_trees_bias_correction(KHLC_ecmwf_analysis_df,\"KHLC\",\"ECMWF\")\n",
    "\n",
    "# concat all results for plotting\n",
    "bias_corrected_results = pd.concat([KBOS_hrrr_results,KHLC_hrrr_results,KBOS_gfs_results,KHLC_gfs_results,KBOS_ecmwf_results,KHLC_ecmwf_results])\n",
    "bias_corrected_results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "varied-gnome",
   "metadata": {},
   "source": [
    "###### Plot of bias-corrected versus raw forecast data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "dependent-amateur",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1189.72x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot of the bias-corrected data vs the raw forecast data\n",
    "sns.set_style(\"whitegrid\")\n",
    "sns.set(rc={'figure.figsize':(10,5)})\n",
    "fig3 = sns.catplot(x='model',y=\"rmse\",hue=\"method\",col = \"variable\",kind=\"bar\",data=bias_corrected_results,palette=[\"skyblue\",\"navy\"])\n",
    "fig3.savefig(\"figures/bias_corrected_vs_raw.png\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "collective-magic",
   "metadata": {},
   "source": [
    "###### Overall Improvement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "sized-colorado",
   "metadata": {},
   "outputs": [],
   "source": [
    "# collect overall improvement by model and method\n",
    "bias_corrected_results_overall = bias_corrected_results.groupby([\"method\",\"model\"]).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "convenient-spokesman",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bias Correction of the ECMWF resulted in a 32.4% overall improved forecast\n",
      "Bias Correction of the GFS resulted in a 31.2% overall improved forecast\n",
      "Bias Correction of the HRRR resulted in a 33.1% overall improved forecast\n"
     ]
    }
   ],
   "source": [
    "# calculate the % improvement\n",
    "corrected_rmse = bias_corrected_results_overall.rmse[\"bias-corrected\"]\n",
    "raw_rmse = bias_corrected_results_overall.rmse[\"raw forecast\"]\n",
    "\n",
    "perc_improvement =  ((abs(corrected_rmse - raw_rmse)) / raw_rmse) * 100\n",
    "\n",
    "\n",
    "\n",
    "print(f\"Bias Correction of the ECMWF resulted in a {perc_improvement.ECMWF.round(1)}% overall improved forecast\")\n",
    "print(f\"Bias Correction of the GFS resulted in a {perc_improvement.GFS.round(1)}% overall improved forecast\")\n",
    "print(f\"Bias Correction of the HRRR resulted in a {perc_improvement.HRRR.round(1)}% overall improved forecast\")\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "productive-dinner",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Bias Correction resulted in a 32.0% overall improved forecast\n"
     ]
    }
   ],
   "source": [
    "# calculate overall improvement across all models\n",
    "mean_rmse_bias = bias_corrected_results_overall.rmse[\"bias-corrected\"].mean()\n",
    "mean_rmse_raw = bias_corrected_results_overall.rmse[\"raw forecast\"].mean()\n",
    "\n",
    "overall_perc_improvement = (abs(mean_rmse_bias - mean_rmse_raw) / mean_rmse_raw) * 100\n",
    "print(f\"Bias Correction resulted in a {overall_perc_improvement.round(1)}% overall improved forecast\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "reverse-queens",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "considerable-drawing",
   "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}