RR - Finalized Model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.linear_model import LinearRegression, Lasso, Ridge, RidgeCV\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV, KFold, cross_val_score, validation_curve\n",
    "from math import sqrt\n",
    "% matplotlib inline \n",
    "\n",
    "pd.options.display.max_columns = 999\n",
    "df_s = pd.read_csv(\"energy_data_variables_final_normalized.csv\", usecols = ['DOLLAREL', 'DRYRUSEM', 'USEWASHM', 'TOTROOMS', 'LGTOUTNUM', 'BEDROOMS', 'DRYRFUEL', 'TOTSQFT_EN', 'NUMCFAN', 'TVCOLOR', 'WINDOWS', 'NCOMBATH', 'LGTINNUM', 'ELWATER', 'FUELH2O', 'OTHROOMS', 'WHEATAGE', 'NUMFRIG', 'NHSLDMEM', 'AGERFRI2', 'LOCRFRI2', 'FUELHEAT', 'THERMAINAC', 'NUMFREEZ', 'KWH_M'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5686, 25)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check Shape of DataFrame\n",
    "df_s.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e0db940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Check distribution skewness for the target variable (KWH_M)\n",
    "ax = sns.distplot(df_s.KWH_M)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e0db828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Remove outliers that are more than 3 standard deviations from the mean \n",
    "df_s = df_s[np.abs(df_s[\"KWH_M\"]-df_s[\"KWH_M\"].mean())<=(3*df_s[\"KWH_M\"].std())]\n",
    "\n",
    "# Re-check distribution skewness for the target variable (KWH_M)\n",
    "ax = sns.distplot(df_s.KWH_M)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set X, Y Variables\n",
    "\n",
    "X = df_s.values[:, 0:24]\n",
    "Y = df_s.values[:, 24]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 6 candidates, totalling 60 fits\n",
      "   mean_fit_time  mean_score_time  mean_test_score  mean_train_score  \\\n",
      "0       0.010797         0.000399         0.822874          0.826645   \n",
      "1       0.003092         0.000366         0.815861          0.819208   \n",
      "2       0.003737         0.000413         0.801684          0.804936   \n",
      "3       0.003315         0.000414         0.786112          0.789373   \n",
      "4       0.003182         0.000343         0.771053          0.774353   \n",
      "5       0.003142         0.000367         0.646053          0.649522   \n",
      "\n",
      "  param_alpha          params  rank_test_score  split0_test_score  \\\n",
      "0         0.1  {'alpha': 0.1}                1           0.831925   \n",
      "1           5    {'alpha': 5}                2           0.823400   \n",
      "2          10   {'alpha': 10}                3           0.809639   \n",
      "3          15   {'alpha': 15}                4           0.795124   \n",
      "4          20   {'alpha': 20}                5           0.781272   \n",
      "5         100  {'alpha': 100}                6           0.663874   \n",
      "\n",
      "   split0_train_score  split1_test_score  split1_train_score  \\\n",
      "0            0.825526           0.846068            0.823981   \n",
      "1            0.818115           0.829302            0.816669   \n",
      "2            0.803837           0.808534            0.802662   \n",
      "3            0.788211           0.788358            0.787400   \n",
      "4            0.773090           0.769933            0.772679   \n",
      "5            0.646533           0.631600            0.650193   \n",
      "\n",
      "   split2_test_score  split2_train_score  split3_test_score  \\\n",
      "0           0.822588            0.826625           0.827470   \n",
      "1           0.803826            0.819484           0.812725   \n",
      "2           0.782602            0.805717           0.793829   \n",
      "3           0.762590            0.790640           0.775207   \n",
      "4           0.744581            0.776044           0.758072   \n",
      "5           0.613049            0.653390           0.627825   \n",
      "\n",
      "   split3_train_score  split4_test_score  split4_train_score  \\\n",
      "0            0.826216           0.805244            0.828470   \n",
      "1            0.818977           0.802263            0.821017   \n",
      "2            0.805042           0.790776            0.806703   \n",
      "3            0.789808           0.777084            0.791083   \n",
      "4            0.775079           0.763418            0.776003   \n",
      "5            0.651798           0.645679            0.650524   \n",
      "\n",
      "   split5_test_score  split5_train_score  split6_test_score  \\\n",
      "0           0.811765            0.827849           0.814144   \n",
      "1           0.817444            0.820185           0.814262   \n",
      "2           0.811557            0.805482           0.803812   \n",
      "3           0.801589            0.789464           0.790272   \n",
      "4           0.790505            0.774027           0.776367   \n",
      "5           0.679937            0.646657           0.651498   \n",
      "\n",
      "   split6_train_score  split7_test_score  split7_train_score  \\\n",
      "0            0.827791           0.818789            0.827155   \n",
      "1            0.820119           0.810185            0.819826   \n",
      "2            0.805480           0.794033            0.805774   \n",
      "3            0.789589           0.776640            0.790448   \n",
      "4            0.774310           0.760007            0.775652   \n",
      "5            0.648691           0.626953            0.652364   \n",
      "\n",
      "   split8_test_score  split8_train_score  split9_test_score  \\\n",
      "0           0.816610            0.827298           0.834143   \n",
      "1           0.811624            0.819821           0.833605   \n",
      "2           0.798717            0.805465           0.823376   \n",
      "3           0.783937            0.789810           0.810356   \n",
      "4           0.769377            0.774708           0.797040   \n",
      "5           0.643976            0.649529           0.676192   \n",
      "\n",
      "   split9_train_score  std_fit_time  std_score_time  std_test_score  \\\n",
      "0            0.825534      0.023768        0.000081        0.011553   \n",
      "1            0.817862      0.000200        0.000040        0.009731   \n",
      "2            0.803204      0.000782        0.000075        0.011378   \n",
      "3            0.787274      0.000521        0.000096        0.013336   \n",
      "4            0.771941      0.000104        0.000016        0.014965   \n",
      "5            0.645541      0.000160        0.000042        0.020984   \n",
      "\n",
      "   std_train_score  \n",
      "0         0.001286  \n",
      "1         0.001241  \n",
      "2         0.001212  \n",
      "3         0.001256  \n",
      "4         0.001351  \n",
      "5         0.002517  \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  60 out of  60 | elapsed:    0.4s finished\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('mean_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split0_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split1_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split2_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split3_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split4_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split5_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split6_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split7_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split8_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split9_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('std_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n"
     ]
    }
   ],
   "source": [
    "parameters={'alpha': [0.1, 5,10,15,20,100]}\n",
    "rdg_reg = Ridge()\n",
    "grid_obj = GridSearchCV(rdg_reg,parameters,cv=10, scoring = 'r2', verbose = 1)\n",
    "grid_obj.fit(X, Y)\n",
    "results = pd.DataFrame(grid_obj.cv_results_)\n",
    "print (results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1a1b2a6eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot a Visual of the Optimal Alpha \n",
    "%matplotlib inline\n",
    "plt.style.use('seaborn')\n",
    "plt.plot(results['param_alpha'], results['mean_test_score'], 'g-', alpha=0.6)\n",
    "plt.text(0,0.7, ' Best-score = {}'.format(round(float(grid_obj.best_score_), 4)))\n",
    "plt.text(0,0.666, ' Optimal alpha = {}'.format(grid_obj.best_params_))\n",
    "plt.xlabel(\"Alpha\")\n",
    "plt.ylabel(\"Mean test score\")\n",
    "plt.title('GridsearchCV RidgeRegressor (CV=10)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 6 candidates, totalling 60 fits\n",
      "   mean_fit_time  mean_score_time  mean_test_score  mean_train_score  \\\n",
      "0       0.004000         0.000427        -0.001753         -0.001723   \n",
      "1       0.002986         0.000299        -0.001823         -0.001797   \n",
      "2       0.006980         0.000354        -0.001964         -0.001939   \n",
      "3       0.003963         0.000377        -0.002118         -0.002094   \n",
      "4       0.003458         0.000342        -0.002268         -0.002243   \n",
      "5       0.007373         0.001173        -0.003507         -0.003484   \n",
      "\n",
      "  param_alpha          params  rank_test_score  split0_test_score  \\\n",
      "0         0.1  {'alpha': 0.1}                1          -0.001789   \n",
      "1           5    {'alpha': 5}                2          -0.001880   \n",
      "2          10   {'alpha': 10}                3          -0.002027   \n",
      "3          15   {'alpha': 15}                4          -0.002181   \n",
      "4          20   {'alpha': 20}                5          -0.002329   \n",
      "5         100  {'alpha': 100}                6          -0.003578   \n",
      "\n",
      "   split0_train_score  split1_test_score  split1_train_score  \\\n",
      "0           -0.001719          -0.001607           -0.001740   \n",
      "1           -0.001792          -0.001782           -0.001812   \n",
      "2           -0.001932          -0.001999           -0.001950   \n",
      "3           -0.002086          -0.002209           -0.002101   \n",
      "4           -0.002235          -0.002402           -0.002247   \n",
      "5           -0.003482          -0.003846           -0.003457   \n",
      "\n",
      "   split2_test_score  split2_train_score  split3_test_score  \\\n",
      "0          -0.001761           -0.001723          -0.001685   \n",
      "1          -0.001947           -0.001794          -0.001828   \n",
      "2          -0.002158           -0.001931          -0.002013   \n",
      "3          -0.002357           -0.002081          -0.002195   \n",
      "4          -0.002536           -0.002226          -0.002362   \n",
      "5          -0.003841           -0.003445          -0.003634   \n",
      "\n",
      "   split3_train_score  split4_test_score  split4_train_score  \\\n",
      "0           -0.001731          -0.001825           -0.001715   \n",
      "1           -0.001803          -0.001853           -0.001790   \n",
      "2           -0.001942          -0.001961           -0.001933   \n",
      "3           -0.002093          -0.002089           -0.002089   \n",
      "4           -0.002240          -0.002217           -0.002240   \n",
      "5           -0.003468          -0.003321           -0.003494   \n",
      "\n",
      "   split5_test_score  split5_train_score  split6_test_score  \\\n",
      "0          -0.001839           -0.001714          -0.001861   \n",
      "1          -0.001783           -0.001791          -0.001860   \n",
      "2          -0.001841           -0.001937          -0.001965   \n",
      "3          -0.001938           -0.002096          -0.002101   \n",
      "4          -0.002047           -0.002250          -0.002240   \n",
      "5          -0.003127           -0.003519          -0.003490   \n",
      "\n",
      "   split6_train_score  split7_test_score  split7_train_score  \\\n",
      "0           -0.001710          -0.001606           -0.001739   \n",
      "1           -0.001787          -0.001682           -0.001813   \n",
      "2           -0.001932          -0.001825           -0.001954   \n",
      "3           -0.002090          -0.001980           -0.002108   \n",
      "4           -0.002242          -0.002127           -0.002257   \n",
      "5           -0.003489          -0.003306           -0.003497   \n",
      "\n",
      "   split8_test_score  split8_train_score  split9_test_score  \\\n",
      "0          -0.001828           -0.001716          -0.001731   \n",
      "1          -0.001878           -0.001790          -0.001737   \n",
      "2          -0.002006           -0.001933          -0.001844   \n",
      "3          -0.002154           -0.002089          -0.001979   \n",
      "4          -0.002299           -0.002239          -0.002118   \n",
      "5          -0.003549           -0.003483          -0.003380   \n",
      "\n",
      "   split9_train_score  std_fit_time  std_score_time  std_test_score  \\\n",
      "0           -0.001725      0.001336        0.000181        0.000089   \n",
      "1           -0.001800      0.000142        0.000024        0.000074   \n",
      "2           -0.001945      0.008746        0.000125        0.000098   \n",
      "3           -0.002103      0.000837        0.000091        0.000122   \n",
      "4           -0.002254      0.000801        0.000108        0.000141   \n",
      "5           -0.003504      0.008530        0.002256        0.000220   \n",
      "\n",
      "   std_train_score  \n",
      "0         0.000010  \n",
      "1         0.000009  \n",
      "2         0.000008  \n",
      "3         0.000008  \n",
      "4         0.000009  \n",
      "5         0.000021  \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  60 out of  60 | elapsed:    0.4s finished\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('mean_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split0_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split1_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split2_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split3_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split4_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split5_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split6_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split7_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split8_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('split9_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n",
      "/Users/williamadams/anaconda3/lib/python3.6/site-packages/sklearn/utils/deprecation.py:122: FutureWarning: You are accessing a training score ('std_train_score'), which will not be available by default any more in 0.21. If you need training scores, please set return_train_score=True\n",
      "  warnings.warn(*warn_args, **warn_kwargs)\n"
     ]
    }
   ],
   "source": [
    "parameters={'alpha': [0.1, 5,10,15,20,100]}\n",
    "rdg_reg = Ridge()\n",
    "grid_obj = GridSearchCV(rdg_reg,parameters,cv=10, scoring = 'neg_mean_squared_error', verbose = 1)\n",
    "grid_obj.fit(X, Y)\n",
    "results = pd.DataFrame(grid_obj.cv_results_)\n",
    "print (results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The RMSE scores are [0.04063931 0.03937707 0.04724933 0.04001346 0.04300388 0.04130134\n",
      " 0.04101916 0.04016599 0.04446854 0.04007554]\n",
      "The mean of the RMSE scores is 0.041731\n"
     ]
    }
   ],
   "source": [
    "# Calculate RMSE\n",
    "ridge = Ridge(alpha=0.1)\n",
    "shuffle = KFold(n_splits=10, shuffle=True, random_state=42)\n",
    "mse_scores = cross_val_score(ridge, X, Y, cv=shuffle, scoring = \"neg_mean_squared_error\")\n",
    "mse_scores1 = -mse_scores \n",
    "rmse_scores = np.sqrt(mse_scores1)\n",
    "\n",
    "print (\"The RMSE scores are %s\" \n",
    "       % (rmse_scores))\n",
    "print (\"The mean of the RMSE scores is %f\" \n",
    "       %(rmse_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The MAE scores are [0.02922698 0.02861325 0.03153369 0.029522   0.02972687 0.02963095\n",
      " 0.02887709 0.02862291 0.03171604 0.02816267]\n",
      "The mean of the MAE scores is 0.029563\n"
     ]
    }
   ],
   "source": [
    "# Calculate MAE  \n",
    "from sklearn.metrics import mean_absolute_error\n",
    "\n",
    "ridge = Ridge(alpha=0.1)\n",
    "shuffle = KFold(n_splits=10, shuffle=True, random_state=42)\n",
    "mae_scores = cross_val_score(ridge, X, Y, cv=shuffle, scoring = \"neg_mean_absolute_error\")\n",
    "mae_scores_ = -mae_scores\n",
    "\n",
    "print (\"The MAE scores are %s\" \n",
    "       % (mae_scores_))\n",
    "print (\"The mean of the MAE scores is %f\" \n",
    "       %(mae_scores_.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The explained variance scores are [0.82979459 0.85458789 0.78931777 0.82835584 0.80702881 0.8061706\n",
      " 0.820158   0.8431936  0.79996584 0.85610263]\n",
      "The mean of the explained variance scores is 0.823468\n"
     ]
    }
   ],
   "source": [
    "# Calculate VAR\n",
    "\n",
    "ridge = Ridge(alpha=0.1)\n",
    "shuffle = KFold(n_splits=10, shuffle=True, random_state=42)\n",
    "\n",
    "ev_scores = cross_val_score(ridge, X, Y, cv=shuffle, scoring = \"explained_variance\")\n",
    "\n",
    "print (\"The explained variance scores are %s\" \n",
    "       % (ev_scores))\n",
    "print (\"The mean of the explained variance scores is %f\" \n",
    "       %(ev_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The R2 scores are [0.82953933 0.8542019  0.78929309 0.82834635 0.80687917 0.80612117\n",
      " 0.81979234 0.84290517 0.79983851 0.85609966]\n",
      "The mean of the R2 scores is 0.823302\n"
     ]
    }
   ],
   "source": [
    "# Calculate R2 \n",
    "\n",
    "ridge = Ridge(alpha=0.1)\n",
    "shuffle = KFold(n_splits=10, shuffle=True, random_state=42)\n",
    "\n",
    "r_scores = cross_val_score(ridge, X, Y, cv=shuffle, scoring = \"r2\")\n",
    "\n",
    "print (\"The R2 scores are %s\" \n",
    "       % (r_scores))\n",
    "print (\"The mean of the R2 scores is %f\" \n",
    "       %(r_scores.mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'Ridge Regression')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10c5d5ac8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the Predict vs. Actual Values Using CrossValPredict\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "shuffle = KFold(n_splits=10, shuffle=True, random_state=42)\n",
    "predicted_ridge = cross_val_predict(ridge, X, Y, cv=shuffle)\n",
    "\n",
    "%matplotlib inline\n",
    "plt.style.use('seaborn')\n",
    "plt.scatter(Y, predicted_ridge, s = 5, color = 'magenta', alpha = 0.8)\n",
    "plt.plot([0,0.6], [0,0.6], \"b--\", color = \"blue\", alpha = 0.5)\n",
    "plt.xlabel(\"Actual Monthly KWh\")\n",
    "plt.ylabel(\"Predicted Monthly KWh\")\n",
    "plt.text(0.1,0.55, ' RMSE = {}'.format(round(sqrt(mean_squared_error(Y, predicted_ridge)), 4)))\n",
    "plt.title('Ridge Regression')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}