Predict based on months and rolling or shifted data.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# PREDICT RESIDUALS AFTER MONTHLY TREND REMOVED on ROLLING FEATURE DATA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "plt.style.use('bmh')\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LinearRegression, LogisticRegression\n",
    "from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "\n",
    "from sklearn.metrics import mean_absolute_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# read in pickled features\n",
    "df_sj = pd.read_pickle('./datasets/all_sj.pkl')\n",
    "df_iq = pd.read_pickle('./datasets/all_iq.pkl')\n",
    "\n",
    "total_cases = pd.read_csv('./datasets/dengue_labels_train.csv')\n",
    "cases_sj = total_cases[total_cases['city'] == 'sj']['total_cases']\n",
    "cases_iq = total_cases[total_cases['city'] == 'iq']['total_cases']\n",
    "\n",
    "# create month variable\n",
    "df_sj['month'] = df_sj.index.month\n",
    "df_iq['month'] = df_iq.index.month\n",
    "\n",
    "# create mean ndvi\n",
    "df_sj['ndvi_mean'] = (df_sj['ndvi_ne'] + df_sj['ndvi_nw'] + df_sj['ndvi_se'] + df_sj['ndvi_sw']) / 4.0\n",
    "df_iq['ndvi_mean'] = (df_iq['ndvi_ne'] + df_iq['ndvi_nw'] + df_iq['ndvi_se'] + df_iq['ndvi_sw']) / 4.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x11617ae90>"
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJ0AAAHsCAYAAAB49tyEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1168ba050>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,8))\n",
    "plt.plot(df_sj[:936].index, cases_sj)\n",
    "plt.xlabel('Date', fontsize = 24)\n",
    "plt.ylabel('Number of Cases', fontsize = 24)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train-Validation-Test Split\n",
    "Because this is time-series data and the features set is dependent on the past, the train test split is very delicate. The validation and test feature set will actually draw from data in the past (rolling mean, std, etc.). Therefore, when creating the model, the entire dataset must be used. But I must be careful not to train the model on data from the 'future'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "675 675 675\n",
      "261 261 261\n",
      "260 260\n"
     ]
    }
   ],
   "source": [
    "# train test split\n",
    "#sj\n",
    "\n",
    "# choose split dates\n",
    "sj_valid_split = '2003-4-20'\n",
    "sj_test_split = '2008-4-27' # this will split between pre and post submission dates\n",
    "\n",
    "# split into train, valid, test (no y)\n",
    "sj_train = df_sj.loc[:sj_valid_split]\n",
    "sj_Xtrain = sj_train\n",
    "sj_ytrain = cases_sj[:len(sj_train)]\n",
    "\n",
    "sj_valid = df_sj.loc[sj_valid_split : sj_test_split]\n",
    "sj_Xvalid = sj_valid\n",
    "sj_yvalid = cases_sj[len(sj_train):]\n",
    "\n",
    "sj_test = df_sj.loc[sj_test_split:]\n",
    "sj_Xtest = sj_test\n",
    "\n",
    "print len(sj_train), len(sj_Xtrain), len(sj_ytrain)\n",
    "print len(sj_valid), len(sj_Xvalid), len(sj_yvalid)\n",
    "print len(sj_test), len(sj_Xtest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "364 364 364\n",
      "156 156 156\n",
      "156 156\n"
     ]
    }
   ],
   "source": [
    "#iq\n",
    "\n",
    "# choose split dates\n",
    "iq_valid_split = '2007-7-01'\n",
    "iq_test_split = '2010-7-01' # this will split between pre and post submission dates\n",
    "\n",
    "# split\n",
    "iq_train = df_iq.loc[:iq_valid_split]\n",
    "iq_Xtrain = iq_train\n",
    "iq_ytrain = cases_iq[:len(iq_Xtrain)]\n",
    "\n",
    "iq_valid = df_iq.loc[iq_valid_split : iq_test_split]\n",
    "iq_Xvalid = iq_valid\n",
    "iq_yvalid = cases_iq[len(iq_train):]\n",
    "\n",
    "iq_test = df_iq.loc[iq_test_split:]\n",
    "iq_Xtest = iq_test\n",
    "\n",
    "# check the lengths\n",
    "print len(iq_train), len(iq_Xtrain), len(iq_ytrain)\n",
    "print len(iq_valid), len(iq_Xvalid), len(iq_yvalid)\n",
    "print len(iq_test), len(iq_Xtest)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SJ"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "675 675\n",
      "261 261\n"
     ]
    }
   ],
   "source": [
    "print len(pd.get_dummies(sj_Xtrain['month'], prefix='month')), len(sj_ytrain.values)\n",
    "print len(pd.get_dummies(sj_Xvalid['month'], prefix='month')), len(sj_yvalid.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD3CAYAAADi8sSvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqSsoKCgMIv4/Q6cuSKJcppsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112ab4a50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "24.5775862069\n"
     ]
    }
   ],
   "source": [
    "# sj monthly trend\n",
    "\n",
    "lr_sj = LinearRegression()\n",
    "X = pd.get_dummies(sj_Xtrain['month'], prefix='month')\n",
    "y = sj_ytrain.values\n",
    "\n",
    "lr_sj.fit(X, y)\n",
    "sj_residuals_train = y - lr_sj.predict(X)\n",
    "\n",
    "\n",
    "# on validation data\n",
    "# note: monthly trend does not need previous weeks data, so this can use the validation set\n",
    "Xtest = pd.get_dummies(sj_Xvalid['month'], prefix='month')\n",
    "ytest = sj_yvalid.values\n",
    "sj_residuals_test = ytest - lr_sj.predict(Xtest)\n",
    "\n",
    "sj_predicts = lr_sj.predict(Xtest)\n",
    "plt.plot(sj_predicts)\n",
    "plt.plot(ytest)\n",
    "plt.show()\n",
    "print mean_absolute_error(sj_predicts, ytest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-16.0833333333\n",
      "675 261\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6MAAAEICAYAAACqIiIYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113d62190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Find the residuals of the monthly trend\n",
    "\n",
    "plt.figure(figsize=(16, 4))\n",
    "plt.plot(sj_residuals_test, label = 'residuals', linewidth = 3)\n",
    "plt.plot(sj_residuals_test + lr_sj.predict(Xtest), 'g-*', alpha = .4, label = 'true values')\n",
    "plt.plot(lr_sj.predict(Xtest), alpha = .4, label = 'monthly trend')\n",
    "plt.title('True Values, Monthly Trend, and Residuals of Dengue Cases in San Juan, Puerto Rico')\n",
    "plt.legend()\n",
    "print np.mean(sj_residuals_test)\n",
    "\n",
    "print len(sj_residuals_train), len(sj_residuals_test)\n",
    "\n",
    "## PREDICT THE BLUE!!!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Explore the of correlations on rolling mean/std windows and shifted days back"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([u'weekofyear', u'ndvi_ne', u'ndvi_nw', u'ndvi_se', u'ndvi_sw',\n",
       "       u'precipitation_amt_mm', u'reanalysis_air_temp_k',\n",
       "       u'reanalysis_avg_temp_k', u'reanalysis_dew_point_temp_k',\n",
       "       u'reanalysis_max_air_temp_k', u'reanalysis_min_air_temp_k',\n",
       "       u'reanalysis_precip_amt_kg_per_m2',\n",
       "       u'reanalysis_relative_humidity_percent',\n",
       "       u'reanalysis_sat_precip_amt_mm',\n",
       "       u'reanalysis_specific_humidity_g_per_kg', u'reanalysis_tdtr_k',\n",
       "       u'station_avg_temp_c', u'station_diur_temp_rng_c',\n",
       "       u'station_max_temp_c', u'station_min_temp_c', u'station_precip_mm',\n",
       "       u'month', u'ndvi_mean'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sj_train.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rolling Mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1125caf90>]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD3CAYAAAANMK+RAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niURSJsiCLpFIJGWCLOgSiURSJvx/CchmlyRRIeYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f03b790>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between rolling mean of all features and residuals\n",
    "mean_score = []\n",
    "for i in range (2, 100):\n",
    "    df = sj_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.rolling(i).mean()[i-1:]\n",
    "    df['resid'] = sj_residuals_train[i-1:]\n",
    "\n",
    "    mean_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(mean_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.260470</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.256694</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.237641</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.236774</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.182867</td>\n",
       "      <td>month</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.162529</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.095999</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.066561</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.064774</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.064010</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.053969</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.046420</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.045675</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.043662</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.043217</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.029118</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.019521</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>-0.000061</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-0.008503</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-0.020693</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.020693</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.041680</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.208552</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "16  0.260470                     station_avg_temp_c      53\n",
       "18  0.256694                     station_max_temp_c      51\n",
       "19  0.237641                     station_min_temp_c      52\n",
       "17  0.236774                station_diur_temp_rng_c      59\n",
       "21  0.182867                                  month      52\n",
       "10  0.162529              reanalysis_min_air_temp_k      53\n",
       "1   0.095999                                ndvi_ne      59\n",
       "9   0.066561              reanalysis_max_air_temp_k      48\n",
       "7   0.064774                  reanalysis_avg_temp_k      50\n",
       "4   0.064010                                ndvi_sw       2\n",
       "8   0.053969            reanalysis_dew_point_temp_k      56\n",
       "12  0.046420   reanalysis_relative_humidity_percent       2\n",
       "11  0.045675        reanalysis_precip_amt_kg_per_m2       4\n",
       "0   0.043662                             weekofyear      53\n",
       "6   0.043217                  reanalysis_air_temp_k      50\n",
       "14  0.029118  reanalysis_specific_humidity_g_per_kg      58\n",
       "2   0.019521                                ndvi_nw      59\n",
       "20 -0.000061                      station_precip_mm       2\n",
       "22 -0.008503                              ndvi_mean      59\n",
       "13 -0.020693           reanalysis_sat_precip_amt_mm       2\n",
       "5  -0.020693                   precipitation_amt_mm       2\n",
       "15 -0.041680                      reanalysis_tdtr_k       2\n",
       "3  -0.208552                                ndvi_se      59"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in sj_train.columns:    \n",
    "    df = pd.DataFrame({'feat': sj_train[i]})\n",
    "    for j in range(2, 60):\n",
    "        df_roll = df.rolling(j).mean()[j-1:]\n",
    "        df_roll['resid'] = sj_residuals_train[j-1:]\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in sj_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rolling Standard Deviations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x112624650>]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWwAAAD3CAYAAAA5SW6NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f198810>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between rolling std of all features and residuals\n",
    "std_score = []\n",
    "for i in range (2, 100):\n",
    "    df = sj_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.rolling(i).std()[i-1:]\n",
    "    df['resid'] = sj_residuals_train[i-1:]\n",
    "\n",
    "    std_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(std_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.182867</td>\n",
       "      <td>month</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.100527</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.089871</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.081150</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.076210</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.073469</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.066716</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.045984</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.029485</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.020505</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.009682</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.009590</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.004884</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.004628</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.004628</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.000445</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-0.001440</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.003987</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.012309</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-0.036354</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.053079</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-0.065705</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.085159</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "21  0.182867                                  month      52\n",
       "1   0.100527                                ndvi_ne      56\n",
       "12  0.089871   reanalysis_relative_humidity_percent       3\n",
       "9   0.081150              reanalysis_max_air_temp_k       4\n",
       "7   0.076210                  reanalysis_avg_temp_k       2\n",
       "0   0.073469                             weekofyear      99\n",
       "6   0.066716                  reanalysis_air_temp_k       2\n",
       "18  0.045984                     station_max_temp_c       5\n",
       "11  0.029485        reanalysis_precip_amt_kg_per_m2       2\n",
       "2   0.020505                                ndvi_nw      46\n",
       "19  0.009682                     station_min_temp_c       2\n",
       "16  0.009590                     station_avg_temp_c       2\n",
       "20  0.004884                      station_precip_mm       2\n",
       "5   0.004628                   precipitation_amt_mm       2\n",
       "13  0.004628           reanalysis_sat_precip_amt_mm       2\n",
       "10  0.000445              reanalysis_min_air_temp_k       2\n",
       "14 -0.001440  reanalysis_specific_humidity_g_per_kg       2\n",
       "8  -0.003987            reanalysis_dew_point_temp_k       2\n",
       "15 -0.012309                      reanalysis_tdtr_k       2\n",
       "17 -0.036354                station_diur_temp_rng_c      89\n",
       "4  -0.053079                                ndvi_sw      94\n",
       "22 -0.065705                              ndvi_mean      51\n",
       "3  -0.085159                                ndvi_se      45"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in sj_train.columns:    \n",
    "    df = pd.DataFrame({'feat': sj_train[i]})\n",
    "    for j in range(2, 100):\n",
    "        df_roll = df.rolling(j).std()\n",
    "        df_roll['resid'] = sj_residuals_train\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in sj_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shifted Weeks Back"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1129b4190>]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAD3CAYAAAD8O/QcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n9o7jOAOAJ3vHcZwB4H/1lI+6GWnFBgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125353d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between shifted weeks back of all features and residuals\n",
    "shift_score = []\n",
    "for i in range (2, 20):\n",
    "    df = sj_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.shift(i)[i-1:]\n",
    "    df['resid'] = sj_residuals_train[i-1:]\n",
    "\n",
    "    shift_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(shift_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.097888</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.074214</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.065684</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.051002</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.036221</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.035011</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.028586</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.018430</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.018183</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.015623</td>\n",
       "      <td>month</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.015332</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.007983</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.007939</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.006086</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.005700</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.002265</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.005252</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-0.011134</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>-0.011134</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.021299</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.023786</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>-0.044197</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.107634</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "18  0.097888                     station_max_temp_c       4\n",
       "19  0.074214                     station_min_temp_c       6\n",
       "16  0.065684                     station_avg_temp_c       5\n",
       "4   0.051002                                ndvi_sw       1\n",
       "12  0.036221   reanalysis_relative_humidity_percent       1\n",
       "11  0.035011        reanalysis_precip_amt_kg_per_m2       1\n",
       "9   0.028586              reanalysis_max_air_temp_k       1\n",
       "17  0.018430                station_diur_temp_rng_c       1\n",
       "10  0.018183              reanalysis_min_air_temp_k       7\n",
       "21  0.015623                                  month       4\n",
       "0   0.015332                             weekofyear       2\n",
       "8   0.007983            reanalysis_dew_point_temp_k       1\n",
       "7   0.007939                  reanalysis_avg_temp_k       8\n",
       "14  0.006086  reanalysis_specific_humidity_g_per_kg       1\n",
       "20  0.005700                      station_precip_mm       1\n",
       "6   0.002265                  reanalysis_air_temp_k       8\n",
       "2  -0.005252                                ndvi_nw       1\n",
       "13 -0.011134           reanalysis_sat_precip_amt_mm       1\n",
       "5  -0.011134                   precipitation_amt_mm       1\n",
       "15 -0.021299                      reanalysis_tdtr_k       6\n",
       "1  -0.023786                                ndvi_ne       9\n",
       "22 -0.044197                              ndvi_mean       9\n",
       "3  -0.107634                                ndvi_se       9"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in sj_train.columns:    \n",
    "    df = pd.DataFrame({'feat': sj_train[i]})\n",
    "    for j in range(1, 10):\n",
    "        df_roll = df.shift(j)\n",
    "        df_roll['resid'] = sj_residuals_train\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in sj_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes here:\n",
    "* rolling mean of all temp vars should be of ~50 observations. That is about 1 full year\n",
    "* shifted back should be around 30 observations\n",
    "* should i do a variation of this test for each feature? weeks rolled back should depend on the feature, not featureset as a whole"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test on valid from grid search results\n",
    "use values from 3 tables above"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "615 615\n",
      "201 201\n",
      "19.8063426145\n",
      "0.14586819852\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x112a136d0>]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD3CAYAAADi8sSvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n8P8BN+2DTyVn8m8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112a134d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# munge data\n",
    "\n",
    "# max roll back set to 59\n",
    "\n",
    "# rolling means df\n",
    "Xtrain_means1 = df_sj['station_avg_temp_c'].rolling(window = 53).mean()[60:675]\n",
    "Xtrain_means2 = df_sj['ndvi_ne'].rolling(window = 53).mean()[60:675]\n",
    "Xtrain_means3 = df_sj['ndvi_sw'].rolling(window = 2).mean()[60:675]\n",
    "Xtrain_means4 = df_sj['reanalysis_relative_humidity_percent'].rolling(window = 2).mean()[60:675]\n",
    "\n",
    "# rolling stds df\n",
    "Xtrain_std1 = df_sj['reanalysis_avg_temp_k'].rolling(window = 2).std()[60:675]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtrain_shift1 = df_sj['precipitation_amt_mm'].shift(4)[60:675]\n",
    "Xtrain_shift2 = df_sj['ndvi_sw'].shift(1)[60:675]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtrain = pd.concat([Xtrain_means1, Xtrain_means2, Xtrain_means3, Xtrain_means4,\n",
    "                    Xtrain_std1, \n",
    "                    Xtrain_shift1, Xtrain_shift2], axis = 1)\n",
    "ytrain = sj_residuals_train[60:]\n",
    "\n",
    "print len(Xtrain), len(ytrain)\n",
    "\n",
    "# rolling means df\n",
    "Xvalid_means1 = df_sj['station_avg_temp_c'].rolling(window = 53).mean()[675:936]\n",
    "Xvalid_means2 = df_sj['ndvi_ne'].rolling(window = 53).mean()[675:936]\n",
    "Xvalid_means3 = df_sj['ndvi_sw'].rolling(window = 2).mean()[675:936]\n",
    "Xvalid_means4 = df_sj['reanalysis_relative_humidity_percent'].rolling(window = 2).mean()[675:936]\n",
    "\n",
    "# rolling stds df\n",
    "Xvalid_std1 = df_sj['reanalysis_avg_temp_k'].rolling(window = 2).std()[675:936]\n",
    "\n",
    "# weeks shifted back\n",
    "Xvalid_shift1 = df_sj['precipitation_amt_mm'].shift(4)[675:936]\n",
    "Xvalid_shift2 = df_sj['ndvi_sw'].shift(1)[675:936]\n",
    "\n",
    "# combine all dfs\n",
    "Xvalid = pd.concat([Xvalid_means1, Xvalid_means2, Xvalid_means3, Xvalid_means4,\n",
    "                    Xvalid_std1, \n",
    "                    Xvalid_shift1, Xvalid_shift2], axis = 1)[60:]\n",
    "yvalid = sj_residuals_test[60:]\n",
    "\n",
    "print len(Xvalid), len(yvalid)\n",
    "\n",
    "# fit on model, predict\n",
    "lr_sj_residual = LinearRegression()\n",
    "lr_sj_residual.fit(Xtrain, ytrain)\n",
    "\n",
    "sj_valid_preds = lr_sj_residual.predict(Xvalid)\n",
    "\n",
    "print mean_absolute_error(sj_valid_preds, yvalid)\n",
    "print lr_sj_residual.score(Xvalid, yvalid)\n",
    "plt.plot(yvalid)\n",
    "plt.plot(sj_valid_preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes:\n",
    "\n",
    "* I will slowly add more features to this graph"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test grid searched rollback terms on actual cases\n",
    "Add the monthly trend and the predicted residuals to get predicted cases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds_df = pd.DataFrame({'pred_resid':sj_valid_preds,\n",
    "                         'real_resid':yvalid,\n",
    "                         'monthly_trend':sj_predicts[60:],\n",
    "                         'actual_cases':ytest[60:]})\n",
    "preds_df['pred_cases'] = preds_df['monthly_trend'] + preds_df['pred_resid']\n",
    "preds_df['pred_cases'] = preds_df['pred_cases'].apply(lambda x: int(x)).apply(lambda x: 0 if x < 0 else x)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "predicted and actual MAE:\n",
      "19.0995024876\n",
      "monthly trend and actual MAE:\n",
      "26.6082089552\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11392cb90>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzsAAADBCAYAAADozq7NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJCTOSyRnR0JCQkJCQkJCQkLivOT/zF5zaBFI5GQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1137dad10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print 'predicted and actual MAE:'\n",
    "print mean_absolute_error(preds_df['pred_cases'], preds_df['actual_cases'])\n",
    "print 'monthly trend and actual MAE:'\n",
    "print mean_absolute_error(preds_df['monthly_trend'], preds_df['actual_cases'])\n",
    "preds_df[['monthly_trend', 'actual_cases', 'pred_cases']].plot(figsize=(14,3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ROLLING STATS and R-SQUARED\n",
    "\n",
    "Here, I look at which features and at how many weeks to go back can predict the residuals. I plan to slowly add the rolling stats of a feature to a model and iteratively see which new features can be added to the model.\n",
    "\n",
    "Note, San Juan and Iquitos each require this, as different weather stats can signal different things against the residuals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sj_train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "print len(df_sj[5-1:675])\n",
    "print len(sj_residuals_train[5-1:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# rolling mean of feature to predict residuals\n",
    "\n",
    "feature = []\n",
    "rsq_score = []\n",
    "window = []\n",
    "mae = []\n",
    "\n",
    "for i in ['ndvi_ne', u'ndvi_nw', u'ndvi_se', u'ndvi_sw',\n",
    "       u'precipitation_amt_mm', u'reanalysis_air_temp_k',\n",
    "       u'reanalysis_avg_temp_k', u'reanalysis_dew_point_temp_k',\n",
    "       u'reanalysis_max_air_temp_k', u'reanalysis_min_air_temp_k',\n",
    "       u'reanalysis_precip_amt_kg_per_m2',\n",
    "       u'reanalysis_relative_humidity_percent',\n",
    "       u'reanalysis_sat_precip_amt_mm',\n",
    "       u'reanalysis_specific_humidity_g_per_kg', u'reanalysis_tdtr_k',\n",
    "       u'station_diur_temp_rng_c',\n",
    "       u'station_max_temp_c', u'station_min_temp_c', u'station_precip_mm']:\n",
    "    \n",
    "    for j in range (5, 60):\n",
    "        win_len = j\n",
    "        \n",
    "        #set up training data\n",
    "        Xtrain_feat_meanroll = df_sj[i].rolling(window = win_len).mean()[:675]\n",
    "        Xtrain_feat_stdroll = df_sj[i].rolling(window = win_len).std()[:675]\n",
    "        Xtrain = pd.concat([Xtrain_feat_meanroll, Xtrain_feat_stdroll], axis = 1)[win_len-1:]\n",
    "        ytrain = sj_residuals_train[win_len-1:]\n",
    "\n",
    "#         print len(Xtrain), len(ytrain)\n",
    "\n",
    "        # set up validation data\n",
    "        Xvalid_feat_meanroll = df_sj[i].rolling(window = win_len).mean()[675:936]\n",
    "        Xvalid_feat_stdroll = df_sj[i].rolling(window = win_len).std()[675:936]\n",
    "        Xvalid = pd.concat([Xvalid_feat_meanroll, Xvalid_feat_stdroll], axis = 1)[win_len-1:]\n",
    "        yvalid = sj_residuals_test[win_len-1:]\n",
    "\n",
    "        #check lengths\n",
    "#         print len(Xvalid), len(yvalid)\n",
    "\n",
    "        # model it!\n",
    "        lr_sj_resids = LinearRegression()\n",
    "        lr_sj_resids.fit(Xtrain, ytrain)\n",
    "        sj_resid_preds = lr_sj_resids.predict(Xvalid)\n",
    "\n",
    "        # save info sj residual predictions\n",
    "        window.append(j)\n",
    "        feature.append(i)\n",
    "        rsq_score.append(lr_sj_resids.score(Xvalid, yvalid))\n",
    "        mae.append(mean_absolute_error(sj_resid_preds, yvalid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# print top features by their r-squared at predicting the residuals alone\n",
    "pd.DataFrame(zip(feature, rsq_score, window, mae)).sort_values(1, ascending = False).head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### SHIFTED\n",
    "\n",
    "This should be incorporated into the for loop above to see if shifted days back explain the variance in the residuals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# select which features to use\n",
    "\n",
    "features = ['station_avg_temp_c']\n",
    "\n",
    "# Fit shifted stats of features on residuals\n",
    "\n",
    "lr_sj_resid = LinearRegression()\n",
    "\n",
    "# for loop to see when is best window (how many weeks back)\n",
    "\n",
    "scores = []\n",
    "mae = []\n",
    "shift_num = []\n",
    "for i in range(2, 60):\n",
    "    shift = i\n",
    "    \n",
    "    # fit on training\n",
    "    ss = StandardScaler()\n",
    "    ss.fit(sj_Xtrain[features])\n",
    "    Xs = pd.DataFrame(ss.transform(sj_Xtrain[features]))\n",
    "    Xshift = Xs.shift(shift)\n",
    "    X = Xshift[shift:]\n",
    "    y = sj_residuals_train[shift:]\n",
    "\n",
    "    lr_sj_resid.fit(X, y)\n",
    "    \n",
    "    # score on test\n",
    "    Xstest = pd.DataFrame(ss.transform(sj_Xtest[features]))\n",
    "    Xshift_test = Xstest.shift(shift)\n",
    "    X_test = Xshift_test[shift:]\n",
    "    y_test = sj_residuals_test[shift:]\n",
    "    \n",
    "    shift_num.append(i)\n",
    "    predicts = lr_sj_resid.predict(X_test)\n",
    "    mae.append(mean_absolute_error(predicts, y_test[1:]))\n",
    "    scores.append(lr_sj_resid.score(X_test, y_test[1:]))\n",
    "    \n",
    "#     plt.plot(predicts)\n",
    "#     plt.plot(y_test)\n",
    "#     plt.title(str(i))\n",
    "#     plt.show()\n",
    "    \n",
    "    \n",
    "plt.plot(shift_num, scores)\n",
    "plt.title('r-square')\n",
    "plt.show()\n",
    "# 7 weeks!\n",
    "\n",
    "plt.plot(shift_num, mae)\n",
    "plt.title('mae')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes:\n",
    "* Right now, from my own exploration, there are no features that have a positive r-squared when predicting the residuals.\n",
    "* a thorough check similar to the rolling stats will be a better method"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## IQ\n",
    "\n",
    "Do the same as I did for SJ above, but to see which IQ features explain the variance in the residuals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "364 364\n",
      "156 156\n"
     ]
    }
   ],
   "source": [
    "print len(pd.get_dummies(iq_Xtrain['month'], prefix='month')), len(iq_ytrain.values)\n",
    "print len(pd.get_dummies(iq_Xvalid['month'], prefix='month')), len(iq_yvalid.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD3CAYAAAAALt/WAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11538df10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7.0661005661\n"
     ]
    }
   ],
   "source": [
    "# iq monthly trend\n",
    "\n",
    "lr_iq = LinearRegression()\n",
    "X = pd.get_dummies(iq_Xtrain['month'], prefix='month')\n",
    "y = iq_ytrain.values\n",
    "\n",
    "lr_iq.fit(X, y)\n",
    "iq_residuals_train = y - lr_iq.predict(X)\n",
    "\n",
    "\n",
    "# on validation data\n",
    "# note: monthly trend does not need previous weeks data, so this can use the validation set\n",
    "Xtest = pd.get_dummies(iq_Xvalid['month'], prefix='month')\n",
    "ytest = iq_yvalid.values\n",
    "iq_residuals_test = ytest - lr_iq.predict(Xtest)\n",
    "\n",
    "iq_predicts = lr_iq.predict(Xtest)\n",
    "plt.plot(iq_predicts)\n",
    "plt.plot(ytest)\n",
    "plt.show()\n",
    "print mean_absolute_error(iq_predicts, ytest)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.40677100677\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6MAAAEICAYAAACqIiIYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxvn/+Rp61RYqmREAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115dfa250>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Find the residuals of the monthly trend\n",
    "\n",
    "plt.figure(figsize=(16, 4))\n",
    "plt.plot(iq_residuals_test, label = 'residuals', linewidth = 3)\n",
    "plt.plot(iq_residuals_test + lr_iq.predict(Xtest), 'g-*', alpha = .4, label = 'true values')\n",
    "plt.plot(lr_iq.predict(Xtest), alpha = .4, label = 'monthly trend')\n",
    "plt.title('True Values, Monthly Trend, and Residuals of Dengue Cases in Iquitos, Peru')\n",
    "plt.legend()\n",
    "print np.mean(iq_residuals_test)\n",
    "\n",
    "## PREDICT THE BLUE!!!\n",
    "# Note: the monthly predictions for iq are pretty bad over the validation set..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## See how far back we should roll in Iquitos weather"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rolling Mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x114e15710>]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXIAAAD5CAYAAAA6JL6mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxTECuYGBgYHi/H+77u5M7D45IAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x114595510>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between rolling mean of all features and residuals\n",
    "mean_score = []\n",
    "for i in range (2, 60):\n",
    "    df = iq_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.rolling(i).mean()[i-1:]\n",
    "    df['resid'] = iq_residuals_train[i-1:]\n",
    "\n",
    "    mean_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(mean_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.319138</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.275265</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.217108</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.215623</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.178138</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.168401</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.146191</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.144546</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.143880</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.143286</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.140877</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.140877</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.123911</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.098023</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.057107</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.038182</td>\n",
       "      <td>month</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.034601</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.030641</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-0.052786</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.053521</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-0.059396</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-0.066583</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.114554</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "2   0.319138                                ndvi_nw      58\n",
       "10  0.275265              reanalysis_min_air_temp_k      52\n",
       "8   0.217108            reanalysis_dew_point_temp_k      53\n",
       "14  0.215623  reanalysis_specific_humidity_g_per_kg      53\n",
       "12  0.178138   reanalysis_relative_humidity_percent      14\n",
       "20  0.168401                      station_precip_mm       6\n",
       "19  0.146191                     station_min_temp_c       4\n",
       "4   0.144546                                ndvi_sw      59\n",
       "6   0.143880                  reanalysis_air_temp_k      48\n",
       "7   0.143286                  reanalysis_avg_temp_k      47\n",
       "5   0.140877                   precipitation_amt_mm      56\n",
       "13  0.140877           reanalysis_sat_precip_amt_mm      56\n",
       "22  0.123911                              ndvi_mean      58\n",
       "1   0.098023                                ndvi_ne      58\n",
       "0   0.057107                             weekofyear      34\n",
       "21  0.038182                                  month      33\n",
       "16  0.034601                     station_avg_temp_c       3\n",
       "11  0.030641        reanalysis_precip_amt_kg_per_m2      59\n",
       "17 -0.052786                station_diur_temp_rng_c      59\n",
       "15 -0.053521                      reanalysis_tdtr_k      29\n",
       "18 -0.059396                     station_max_temp_c       2\n",
       "9  -0.066583              reanalysis_max_air_temp_k      59\n",
       "3  -0.114554                                ndvi_se       2"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in iq_train.columns:    \n",
    "    df = pd.DataFrame({'feat': iq_train[i]})\n",
    "    for j in range(2, 60):\n",
    "        df_roll = df.rolling(j).mean()[j-1:]\n",
    "        df_roll['resid'] = iq_residuals_train[j-1:]\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in iq_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rolling Standard Deviations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x114c87a10>]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD3CAYAAAANMK+RAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJiYmJjrBDOgmJiYmOsEM6CYmJiY64f8D9fobtPLgfsUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x114cf2890>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between rolling std of all features and residuals\n",
    "std_score = []\n",
    "for i in range (2, 100):\n",
    "    df = iq_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.rolling(i).std()[i-1:]\n",
    "    df['resid'] = iq_residuals_train[i-1:]\n",
    "\n",
    "    std_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(std_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.394412</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.278016</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.259171</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.232111</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.218749</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.194789</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.182267</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.171760</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.169137</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.161871</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.155343</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.140470</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.103288</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.095413</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.095413</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.081738</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.080751</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.037194</td>\n",
       "      <td>month</td>\n",
       "      <td>54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.031675</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.022243</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>73</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.018064</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.006153</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-0.003249</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>92</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "20  0.394412                      station_precip_mm      31\n",
       "22  0.278016                              ndvi_mean      62\n",
       "15  0.259171                      reanalysis_tdtr_k       5\n",
       "4   0.232111                                ndvi_sw      62\n",
       "3   0.218749                                ndvi_se      60\n",
       "17  0.194789                station_diur_temp_rng_c      20\n",
       "9   0.182267              reanalysis_max_air_temp_k      97\n",
       "1   0.171760                                ndvi_ne      62\n",
       "2   0.169137                                ndvi_nw      99\n",
       "16  0.161871                     station_avg_temp_c      38\n",
       "18  0.155343                     station_max_temp_c      24\n",
       "7   0.140470                  reanalysis_avg_temp_k       5\n",
       "6   0.103288                  reanalysis_air_temp_k      97\n",
       "5   0.095413                   precipitation_amt_mm      47\n",
       "13  0.095413           reanalysis_sat_precip_amt_mm      47\n",
       "10  0.081738              reanalysis_min_air_temp_k      18\n",
       "0   0.080751                             weekofyear      53\n",
       "21  0.037194                                  month      54\n",
       "19  0.031675                     station_min_temp_c      12\n",
       "14  0.022243  reanalysis_specific_humidity_g_per_kg      73\n",
       "12  0.018064   reanalysis_relative_humidity_percent      80\n",
       "8   0.006153            reanalysis_dew_point_temp_k      74\n",
       "11 -0.003249        reanalysis_precip_amt_kg_per_m2      92"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in iq_train.columns:    \n",
    "    df = pd.DataFrame({'feat': iq_train[i]})\n",
    "    for j in range(2, 100):\n",
    "        df_roll = df.rolling(j).std()\n",
    "        df_roll['resid'] = iq_residuals_train\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in iq_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Shifted Weeks Back"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x11546b950>]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD3CAYAAADi8sSvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x114a32910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot sum of correlation between shifted weeks back of all features and residuals\n",
    "shift_score = []\n",
    "for i in range (2, 20):\n",
    "    df = iq_train.drop(['month', 'weekofyear'], axis = 1)\n",
    "    df = df.shift(i)[i-1:]\n",
    "    df['resid'] = iq_residuals_train[i-1:]\n",
    "\n",
    "    shift_score.append(abs(df.corr()['resid']).sum()-1)\n",
    "plt.plot(shift_score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>corr</th>\n",
       "      <th>feat</th>\n",
       "      <th>window</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.147939</td>\n",
       "      <td>reanalysis_relative_humidity_percent</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.142154</td>\n",
       "      <td>reanalysis_specific_humidity_g_per_kg</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.139581</td>\n",
       "      <td>reanalysis_dew_point_temp_k</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.137948</td>\n",
       "      <td>station_min_temp_c</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.135644</td>\n",
       "      <td>station_precip_mm</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.104624</td>\n",
       "      <td>reanalysis_min_air_temp_k</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.100368</td>\n",
       "      <td>precipitation_amt_mm</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.100368</td>\n",
       "      <td>reanalysis_sat_precip_amt_mm</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.049517</td>\n",
       "      <td>ndvi_ne</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.048408</td>\n",
       "      <td>reanalysis_precip_amt_kg_per_m2</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.039431</td>\n",
       "      <td>station_avg_temp_c</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.026109</td>\n",
       "      <td>weekofyear</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.019310</td>\n",
       "      <td>ndvi_nw</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.016351</td>\n",
       "      <td>ndvi_sw</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.013749</td>\n",
       "      <td>ndvi_mean</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.001572</td>\n",
       "      <td>station_max_temp_c</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.000294</td>\n",
       "      <td>reanalysis_air_temp_k</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.000218</td>\n",
       "      <td>month</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.004173</td>\n",
       "      <td>reanalysis_avg_temp_k</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.012191</td>\n",
       "      <td>ndvi_se</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.013746</td>\n",
       "      <td>reanalysis_tdtr_k</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-0.041040</td>\n",
       "      <td>reanalysis_max_air_temp_k</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-0.051081</td>\n",
       "      <td>station_diur_temp_rng_c</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        corr                                   feat  window\n",
       "12  0.147939   reanalysis_relative_humidity_percent       8\n",
       "14  0.142154  reanalysis_specific_humidity_g_per_kg       9\n",
       "8   0.139581            reanalysis_dew_point_temp_k       9\n",
       "19  0.137948                     station_min_temp_c       2\n",
       "20  0.135644                      station_precip_mm       4\n",
       "10  0.104624              reanalysis_min_air_temp_k       9\n",
       "5   0.100368                   precipitation_amt_mm       3\n",
       "13  0.100368           reanalysis_sat_precip_amt_mm       3\n",
       "1   0.049517                                ndvi_ne       5\n",
       "11  0.048408        reanalysis_precip_amt_kg_per_m2       9\n",
       "16  0.039431                     station_avg_temp_c       7\n",
       "0   0.026109                             weekofyear       5\n",
       "2   0.019310                                ndvi_nw       4\n",
       "4   0.016351                                ndvi_sw       8\n",
       "22  0.013749                              ndvi_mean       8\n",
       "18  0.001572                     station_max_temp_c       6\n",
       "6   0.000294                  reanalysis_air_temp_k       7\n",
       "21  0.000218                                  month       6\n",
       "7  -0.004173                  reanalysis_avg_temp_k       7\n",
       "3  -0.012191                                ndvi_se       8\n",
       "15 -0.013746                      reanalysis_tdtr_k       6\n",
       "9  -0.041040              reanalysis_max_air_temp_k       6\n",
       "17 -0.051081                station_diur_temp_rng_c       9"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature = []\n",
    "roll = []\n",
    "corr = []\n",
    "\n",
    "for i in iq_train.columns:    \n",
    "    df = pd.DataFrame({'feat': iq_train[i]})\n",
    "    for j in range(1, 10):\n",
    "        df_roll = df.shift(j)\n",
    "        df_roll['resid'] = iq_residuals_train\n",
    "        corr.append(df_roll.corr()['feat'].sum()-1)\n",
    "        feature.append(i)\n",
    "        roll.append(j)\n",
    "\n",
    "df_corrs = pd.DataFrame({'feature':feature,\n",
    "                            'roll':roll,\n",
    "                            'corr':corr})\n",
    "\n",
    "# group this by feature, select the max corr value, and spit out what roll back that is\n",
    "corr_max = []\n",
    "feat_max = []\n",
    "roll_max = []\n",
    "for i in iq_train.columns:\n",
    "    corr_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][0])\n",
    "    feat_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][1])\n",
    "    roll_max.append(df_corrs[df_corrs['feature'] == i].sort_values('corr', ascending = False).iloc[0][2])\n",
    "    \n",
    "pd.DataFrame({'corr':corr_max,\n",
    "              'feat':feat_max,\n",
    "              'window':roll_max}).sort_values('corr', ascending = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes on how far back to roll in IQ:\n",
    "* rolling mean should be of ~50 observations. That is about 1 full year\n",
    "* rolling std should be 30 or 79? I don't get why\n",
    "* shifted back is unclear\n",
    "* should i do a variation of this test for each feature? weeks rolled back should depend on the feature, not featureset as a whole"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## predict the residuals based on rolling weather features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df_iq.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "252"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(Xtrain)\n",
    "len(ytrain)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "304 304\n",
      "96 96\n",
      "0.0552338557193\n",
      "7.9121775868\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD3CAYAAADi8sSvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhWOJAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11546bfd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# rolling mean of feature to predict residuals\n",
    "\n",
    "# munge data\n",
    "\n",
    "# max roll back set to 59\n",
    "\n",
    "# set up training data\n",
    "# rolling means df\n",
    "Xtrain_means1 = df_iq['ndvi_nw'].rolling(window = 53).mean()[60:364]\n",
    "Xtrain_means2 = df_iq['reanalysis_min_air_temp_k'].rolling(window = 5).mean()[60:364]\n",
    "Xtrain_means3 = df_iq['reanalysis_relative_humidity_percent'].rolling(window = 14).mean()[60:364]\n",
    "Xtrain_means4 = df_iq['station_precip_mm'].rolling(window = 6).mean()[60:364]\n",
    "\n",
    "# rolling stds df\n",
    "Xtrain_std1 = df_iq['station_precip_mm'].rolling(window = 31).std()[60:364]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtrain_shift1 = df_iq['station_min_temp_c'].shift(2)[60:364]\n",
    "Xtrain_shift2 = df_iq['station_precip_mm'].shift(1)[60:364]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtrain = pd.concat([Xtrain_means1, Xtrain_means2], axis = 1)\n",
    "ytrain = iq_residuals_train[60:]\n",
    "\n",
    "print len(Xtrain), len(ytrain)\n",
    "\n",
    "# set up validation data\n",
    "# rolling means df\n",
    "Xvalid_means1 = df_iq['ndvi_nw'].rolling(window = 53).mean()[364:520]\n",
    "Xvalid_means2 = df_iq['reanalysis_min_air_temp_k'].rolling(window = 5).mean()[364:520]\n",
    "Xvalid_means3 = df_iq['reanalysis_relative_humidity_percent'].rolling(window = 14).mean()[364:520]\n",
    "Xvalid_means4 = df_iq['station_precip_mm'].rolling(window = 6).mean()[364:520]\n",
    "\n",
    "# rolling stds df\n",
    "Xvalid_std1 = df_iq['station_precip_mm'].rolling(window = 31).std()[364:520]\n",
    "\n",
    "# weeks shifted back\n",
    "Xvalid_shift1 = df_iq['station_min_temp_c'].shift(2)[364:520]\n",
    "Xvalid_shift2 = df_iq['station_precip_mm'].shift(1)[364:520]\n",
    "\n",
    "# combine all dfs\n",
    "Xvalid = pd.concat([Xvalid_means1, Xvalid_means2], axis = 1)[60:]\n",
    "yvalid = iq_residuals_test[60:]\n",
    "\n",
    "print len(Xvalid), len(yvalid)\n",
    "\n",
    "# model it!\n",
    "\n",
    "lr_iq_resids = LinearRegression()\n",
    "lr_iq_resids.fit(Xtrain, ytrain)\n",
    "\n",
    "iq_valid_preds = lr_iq_resids.predict(Xvalid)\n",
    "\n",
    "# plot iq residual predictions\n",
    "plt.plot(yvalid, alpha = .75)\n",
    "plt.plot(iq_valid_preds)\n",
    "print lr_iq_resids.score(Xvalid, yvalid)\n",
    "print mean_absolute_error(iq_valid_preds, yvalid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Notes:\n",
    "\n",
    "* IQ resiuduals is much more erratic and harder to predict\n",
    "* Tweak these features more"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test grid searched rollback terms on actual cases\n",
    "Add the monthly trend and the predicted residuals to get predicted cases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds_df = pd.DataFrame({'pred_resid':iq_valid_preds,\n",
    "                         'real_resid':yvalid,\n",
    "                         'monthly_trend':iq_predicts[60:],\n",
    "                         'actual_cases':ytest[60:]})\n",
    "preds_df['pred_cases'] = preds_df['monthly_trend'] + preds_df['pred_resid']\n",
    "preds_df['pred_cases'] = preds_df['pred_cases'].apply(lambda x: int(x)).apply(lambda x: 0 if x < 0 else x)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "print 'predicted and actual MAE:'\n",
    "print mean_absolute_error(preds_df['pred_cases'], preds_df['actual_cases'])\n",
    "print 'monthly trend and actual MAE:'\n",
    "print mean_absolute_error(preds_df['monthly_trend'], preds_df['actual_cases'])\n",
    "preds_df[['monthly_trend', 'actual_cases', 'pred_cases']].plot(figsize=(14,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## For loop through all features to see which predict residual individually\n",
    "\n",
    "grid search through the features and window length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# rolling mean of feature to predict residuals\n",
    "\n",
    "feature = []\n",
    "rsq_score = []\n",
    "window = []\n",
    "mae = []\n",
    "\n",
    "for i in df_iq.columns:\n",
    "    for j in range (5, 100):\n",
    "        win_len = j\n",
    "        \n",
    "        #set up training data\n",
    "        Xtrain_feat_meanroll = df_iq[:364][i].rolling(window = win_len).mean()[win_len-1:364]\n",
    "        Xtrain_feat_stdroll = df_iq[:364][i].rolling(window = win_len).std()[win_len-1:364]\n",
    "        Xtrain = pd.concat([Xtrain_feat_meanroll, Xtrain_feat_stdroll], axis = 1)\n",
    "        ytrain = iq_residuals_train[win_len-1:]\n",
    "\n",
    "    #     print len(Xtrain), len(ytrain)\n",
    "\n",
    "        # set up validation data\n",
    "        Xvalid_feat_meanroll = df_iq[:520][i].rolling(window = win_len).mean()[364:520]\n",
    "        Xvalid_feat_stdroll = df_iq[:520][i].rolling(window = win_len).std()[364:520]\n",
    "        Xvalid = pd.concat([Xvalid_feat_meanroll, Xvalid_feat_stdroll], axis = 1)\n",
    "        yvalid = iq_residuals_test\n",
    "\n",
    "        #check lengths\n",
    "    #     print len(Xvalid), len(yvalid)\n",
    "\n",
    "        # model it!\n",
    "        lr_iq_resids = LinearRegression()\n",
    "        lr_iq_resids.fit(Xtrain, ytrain)\n",
    "        iq_resid_preds = lr_iq_resids.predict(Xvalid)\n",
    "\n",
    "        # save info iq residual predictions\n",
    "        window.append(j)\n",
    "        feature.append(i)\n",
    "        rsq_score.append(lr_iq_resids.score(Xvalid, yvalid))\n",
    "        mae.append(mean_absolute_error(iq_resid_preds, yvalid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pd.DataFrame(zip(feature, rsq_score, window, mae)).sort_values(1, ascending = False).head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "len(iq_resid_preds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "len(iq_predicts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "preds = []\n",
    "for i in range(156):\n",
    "    preds.append(int(iq_resid_preds[i] + iq_predicts[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.plot(ytest)\n",
    "plt.plot(iq_resid_preds)\n",
    "plt.plot(preds)\n",
    "\n",
    "mean_absolute_error(ytest, preds)\n",
    "# using station_diur_temp_rng_c window = 18 has sliiiiiightly better than monthly MAE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## USE TEST DATA, MAKE CSV OF SUBMISSIONS\n",
    "\n",
    "Need to use full dataset (train + test) because values depend on previous weather data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# transform sj data\n",
    "\n",
    "month_dums = pd.get_dummies(df_sj['month'], prefix='month')\n",
    "temp_roll_means = pd.DataFrame(df_sj[['station_avg_temp_c']].rolling(window = 55).mean())\n",
    "# temp_roll_std = pd.DataFrame(df_sj[['station_avg_temp_c', 'precipitation_amt_mm']].rolling(window = 8).mean())\n",
    "\n",
    "# combine into test set\n",
    "Xtest = pd.concat([month_dums, temp_roll_means], axis = 1)\n",
    "\n",
    "# train model\n",
    "lr_sj_full = LinearRegression()\n",
    "lr_sj_full.fit(Xtest[60:-260], cases_sj.values[60:])\n",
    "\n",
    "# predicts\n",
    "sj_full_preds = lr_sj_full.predict(Xtest[55:])\n",
    "sj_submit_preds = sj_full_preds[-260:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# transform iq data\n",
    "month_dums = pd.get_dummies(df_iq['month'], prefix='month')\n",
    "temp_roll_means = pd.DataFrame(df_iq[['ndvi_nw']].rolling(window = 68).mean())\n",
    "# temp_roll_std = pd.DataFrame(df_iq[['station_avg_temp_c', 'precipitation_amt_mm']].rolling(window = 8).mean())\n",
    "\n",
    "# combine into test\n",
    "Xtest = pd.concat([month_dums, temp_roll_means], axis = 1)\n",
    "\n",
    "# train\n",
    "lr_iq_full = LinearRegression()\n",
    "lr_iq_full.fit(Xtest[68:-156], cases_iq.values[68:])\n",
    "\n",
    "# predicts\n",
    "iq_full_preds = lr_iq_full.predict(Xtest[68:])\n",
    "iq_submit_preds = iq_full_preds[-156:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.plot(sj_submit_preds)\n",
    "plt.show()\n",
    "plt.plot(iq_submit_preds)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "total_preds = list(sj_submit_preds) + list(iq_submit_preds)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### make csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# submission\n",
    "submission_format = pd.read_csv('./datasets/submission_format.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "submission_format['total_cases'] = total_preds\n",
    "submission_format['total_cases'] = submission_format['total_cases'].apply(lambda x: int(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save to CSV, use current date\n",
    "submission_format.to_csv('submissions/dengue_submission_9_12_17v1.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make predictions v2\n",
    "Predict the residuals, then add them to the monthly trend"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### San Juan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# get monthly trend of whole test df\n",
    "lr_sj_month = LinearRegression()\n",
    "X_months = pd.get_dummies(df_sj['month'], prefix='month')[:936]\n",
    "Xtest_months = pd.get_dummies(df_sj['month'], prefix='month')[936:]\n",
    "y = cases_sj.values\n",
    "\n",
    "lr_sj_month.fit(X_months, y)\n",
    "monthly_trend = lr_sj_month.predict(X_months)\n",
    "sj_residuals_all = y - monthly_trend\n",
    "\n",
    "# create test df of rolling weather stats\n",
    "# rolling means df\n",
    "Xtrain_means1 = df_sj['station_avg_temp_c'].rolling(window = 53).mean()[60:936]\n",
    "Xtrain_means2 = df_sj['ndvi_ne'].rolling(window = 53).mean()[60:936]\n",
    "Xtrain_means3 = df_sj['ndvi_sw'].rolling(window = 2).mean()[60:936]\n",
    "Xtrain_means4 = df_sj['reanalysis_relative_humidity_percent'].rolling(window = 2).mean()[60:936]\n",
    "\n",
    "# rolling stds df\n",
    "Xtrain_std1 = df_sj['reanalysis_avg_temp_k'].rolling(window = 2).std()[60:936]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtrain_shift1 = df_sj['precipitation_amt_mm'].shift(4)[60:936]\n",
    "Xtrain_shift2 = df_sj['ndvi_sw'].shift(1)[60:936]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtrain = pd.concat([Xtrain_means1, Xtrain_means2, Xtrain_means3, Xtrain_means4,\n",
    "                    Xtrain_std1, \n",
    "                    Xtrain_shift1, Xtrain_shift2], axis = 1)\n",
    "ytrain = sj_residuals_all[60:]\n",
    "\n",
    "\n",
    "# create test df on rolling weather stats\n",
    "# rolling means df\n",
    "Xtest_means1 = df_sj['station_avg_temp_c'].rolling(window = 53).mean()[936:]\n",
    "Xtest_means2 = df_sj['ndvi_ne'].rolling(window = 53).mean()[936:]\n",
    "Xtest_means3 = df_sj['ndvi_sw'].rolling(window = 2).mean()[936:]\n",
    "Xtest_means4 = df_sj['reanalysis_relative_humidity_percent'].rolling(window = 2).mean()[936:]\n",
    "\n",
    "# rolling stds df\n",
    "Xtest_std1 = df_sj['reanalysis_avg_temp_k'].rolling(window = 2).std()[936:]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtest_shift1 = df_sj['precipitation_amt_mm'].shift(4)[936:]\n",
    "Xtest_shift2 = df_sj['ndvi_sw'].shift(1)[936:]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtest_weather = pd.concat([Xtest_means1, Xtest_means2, Xtest_means3, Xtest_means4,\n",
    "                    Xtest_std1, \n",
    "                    Xtest_shift1, Xtest_shift2], axis = 1)\n",
    "\n",
    "# fit on model\n",
    "lr_sj_resid = LinearRegression()\n",
    "lr_sj_resid.fit(Xtrain, ytrain)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# make predictions on monthly data and residual data\n",
    "sj_monthly_preds = lr_sj_month.predict(Xtest_months)\n",
    "sj_resid_preds = lr_sj_resid.predict(Xtest_weather)\n",
    "sj_cases_pred = pd.Series(sj_resid_preds + sj_monthly_preds)\n",
    "sj_cases_pred = sj_cases_pred.apply(lambda x: 0 if x < 1 else int(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.plot(sj_resid_preds)\n",
    "plt.plot(sj_monthly_preds)\n",
    "plt.plot(sj_cases_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Iquitos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "len(cases_iq)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# get monthly trend of whole test df\n",
    "lr_iq_month = LinearRegression()\n",
    "X_months = pd.get_dummies(df_iq['month'], prefix='month')[:520]\n",
    "Xtest_months = pd.get_dummies(df_iq['month'], prefix='month')[520:]\n",
    "y = cases_iq.values\n",
    "\n",
    "lr_iq_month.fit(X_months, y)\n",
    "monthly_trend = lr_iq_month.predict(X_months)\n",
    "iq_residuals_all = y - monthly_trend\n",
    "\n",
    "# create test df of rolling weather stats\n",
    "# rolling means df\n",
    "Xtrain_means1 = df_iq['ndvi_nw'].rolling(window = 53).mean()[60:520]\n",
    "Xtrain_means2 = df_iq['station_min_temp_c'].rolling(window = 3).mean()[60:520]\n",
    "Xtrain_means3 = df_iq['reanalysis_relative_humidity_percent'].rolling(window = 14).mean()[60:520]\n",
    "Xtrain_means4 = df_iq['station_precip_mm'].rolling(window = 6).mean()[60:520]\n",
    "\n",
    "# rolling stds df\n",
    "Xtrain_std1 = df_iq['station_precip_mm'].rolling(window = 31).std()[60:520]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtrain_shift1 = df_iq['station_min_temp_c'].shift(2)[60:520]\n",
    "Xtrain_shift2 = df_iq['station_precip_mm'].shift(4)[60:520]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtrain = pd.concat([Xtrain_means1, Xtrain_means2, Xtrain_means3, Xtrain_means4,\n",
    "                    Xtrain_std1, \n",
    "                    Xtrain_shift1, Xtrain_shift2], axis = 1)\n",
    "ytrain = iq_residuals_all[60:]\n",
    "\n",
    "\n",
    "# create test df on rolling weather stats\n",
    "# rolling means df\n",
    "Xtest_means1 = df_iq['ndvi_nw'].rolling(window = 53).mean()[520:]\n",
    "Xtest_means2 = df_iq['station_min_temp_c'].rolling(window = 3).mean()[520:]\n",
    "Xtest_means3 = df_iq['reanalysis_relative_humidity_percent'].rolling(window = 14).mean()[520:]\n",
    "Xtest_means4 = df_iq['station_precip_mm'].rolling(window = 6).mean()[520:]\n",
    "\n",
    "# rolling stds df\n",
    "Xtest_std1 = df_iq['station_precip_mm'].rolling(window = 31).std()[520:]\n",
    "\n",
    "# weeks shifted back\n",
    "Xtest_shift1 = df_iq['station_min_temp_c'].shift(2)[520:]\n",
    "Xtest_shift2 = df_iq['station_precip_mm'].shift(4)[520:]\n",
    "\n",
    "\n",
    "# combine all dfs\n",
    "Xtest_weather = pd.concat([Xtest_means1, Xtest_means2, Xtest_means3, Xtest_means4,\n",
    "                    Xtest_std1, \n",
    "                    Xtest_shift1, Xtest_shift2], axis = 1)\n",
    "\n",
    "# fit on model\n",
    "lr_iq_resid = LinearRegression()\n",
    "lr_iq_resid.fit(Xtrain, ytrain)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# make predictions on monthly data and residual data\n",
    "iq_monthly_preds = lr_iq_month.predict(Xtest_months)\n",
    "iq_resid_preds = lr_iq_resid.predict(Xtest_weather)\n",
    "iq_cases_pred = pd.Series(iq_resid_preds + iq_monthly_preds)\n",
    "iq_cases_pred = iq_cases_pred.apply(lambda x: 0 if x < 1 else int(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "plt.plot(iq_resid_preds)\n",
    "plt.plot(iq_monthly_preds)\n",
    "plt.plot(iq_cases_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### make csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "total_preds = list(sj_cases_pred) + list(iq_cases_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# sub`mission\n",
    "submission_format = pd.read_csv('./datasets/submission_format.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "submission_format['total_cases'] = total_preds\n",
    "submission_format['total_cases'] = submission_format['total_cases'].apply(lambda x: int(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Save to CSV, use current date\n",
    "submission_format.to_csv('submissions/dengue_submission_9_14_17v1.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Submitted MAEs!!!!\n",
    "* monthly trend and mean `station_avg_temp_c` with window = 52 was 20.7764\n",
    "* monthly trend and mean `station_avg_temp_c` with window = 52 and std window = 8 was 20.7981\n",
    "* monthly trend and mean `station_avg_temp_c` and `precipitation_amt_mm` with window = 52 and std window = 8 was 24.1274\n",
    "* monthly trend and sj used mean `station_avg_temp_c` with window = 55 and iq used 'ndvi_nw' with window = 68 was 23.5697\n",
    "* Next steps: use other rolling features (and std? and shifted? and exponentially weighted mean?)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}