predictingNYPDcrimelag17.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Testing Gradient Boosting on lag 1 and lag 7 data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import GradientBoostingRegressor\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datestop</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>SAF_COUNT</th>\n",
       "      <th>311_complaints_COUNT</th>\n",
       "      <th>passenger_count_pickup</th>\n",
       "      <th>total_amount_pickup</th>\n",
       "      <th>passenger_count_dropoff</th>\n",
       "      <th>total_amount_dropoff</th>\n",
       "      <th>NN_SAF</th>\n",
       "      <th>...</th>\n",
       "      <th>NS_TAXI_P_FARE</th>\n",
       "      <th>NS_TAXI_D_FARE</th>\n",
       "      <th>SAF_lag</th>\n",
       "      <th>311_lag</th>\n",
       "      <th>NYPD_lag</th>\n",
       "      <th>TAXI_P_PASSENGER_lag</th>\n",
       "      <th>TAXI_D_PASSENGER_lag</th>\n",
       "      <th>TAXI_P_FARE_lag</th>\n",
       "      <th>TAXI_D_FARE_lag</th>\n",
       "      <th>nypd_complaints_COUNT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>22</td>\n",
       "      <td>14</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>312.93</td>\n",
       "      <td>26.0</td>\n",
       "      <td>579.03</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52105.0</td>\n",
       "      <td>374601.91</td>\n",
       "      <td>54970.0</td>\n",
       "      <td>382690.74</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>64307.0</td>\n",
       "      <td>61028.0</td>\n",
       "      <td>459089.86</td>\n",
       "      <td>408074.78</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>15</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>809.62</td>\n",
       "      <td>461.0</td>\n",
       "      <td>9323.87</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>390.0</td>\n",
       "      <td>660.83</td>\n",
       "      <td>8008.99</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>757.78</td>\n",
       "      <td>923.0</td>\n",
       "      <td>15469.21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>56.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>116.0</td>\n",
       "      <td>896.0</td>\n",
       "      <td>1312.94</td>\n",
       "      <td>15219.87</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>9.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>585.0</td>\n",
       "      <td>5544.04</td>\n",
       "      <td>2652.0</td>\n",
       "      <td>35447.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>659.0</td>\n",
       "      <td>2587.0</td>\n",
       "      <td>5611.30</td>\n",
       "      <td>33921.34</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 45 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     datestop   x   y  SAF_COUNT  311_complaints_COUNT  \\\n",
       "0  1372543200  22  14        0.0                   0.0   \n",
       "1  1372543200  12   7        0.0                   0.0   \n",
       "2  1372543200  12  15        2.0                  12.0   \n",
       "3  1372543200  12  14        9.0                  10.0   \n",
       "4  1372543200  12  13        9.0                  30.0   \n",
       "\n",
       "   passenger_count_pickup  total_amount_pickup  passenger_count_dropoff  \\\n",
       "0                    16.0               312.93                     26.0   \n",
       "1                 52105.0            374601.91                  54970.0   \n",
       "2                    66.0               809.62                    461.0   \n",
       "3                    86.0               757.78                    923.0   \n",
       "4                   585.0              5544.04                   2652.0   \n",
       "\n",
       "   total_amount_dropoff  NN_SAF  ...  NS_TAXI_P_FARE  NS_TAXI_D_FARE  SAF_lag  \\\n",
       "0                579.03     0.0  ...             0.0             0.0      0.0   \n",
       "1             382690.74     0.0  ...             0.0             0.0      3.0   \n",
       "2               9323.87     0.0  ...             0.0             0.0      3.0   \n",
       "3              15469.21     0.0  ...             0.0             0.0     11.0   \n",
       "4              35447.31     0.0  ...             0.0             0.0     10.0   \n",
       "\n",
       "   311_lag  NYPD_lag  TAXI_P_PASSENGER_lag  TAXI_D_PASSENGER_lag  \\\n",
       "0      0.0       0.0                   0.0                   0.0   \n",
       "1     26.0      21.0               64307.0               61028.0   \n",
       "2     25.0      11.0                  57.0                 390.0   \n",
       "3     56.0      23.0                 116.0                 896.0   \n",
       "4     52.0      18.0                 659.0                2587.0   \n",
       "\n",
       "   TAXI_P_FARE_lag  TAXI_D_FARE_lag  nypd_complaints_COUNT  \n",
       "0             0.00             0.00                      1  \n",
       "1        459089.86        408074.78                      9  \n",
       "2           660.83          8008.99                      7  \n",
       "3          1312.94         15219.87                     15  \n",
       "4          5611.30         33921.34                      5  \n",
       "\n",
       "[5 rows x 45 columns]"
      ]
     },
     "execution_count": 116,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_1=pd.read_csv('data_lag1.csv',index_col=0)\n",
    "data_1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>datestop</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>SAF_COUNT</th>\n",
       "      <th>311_complaints_COUNT</th>\n",
       "      <th>passenger_count_pickup</th>\n",
       "      <th>total_amount_pickup</th>\n",
       "      <th>passenger_count_dropoff</th>\n",
       "      <th>total_amount_dropoff</th>\n",
       "      <th>NN_SAF</th>\n",
       "      <th>...</th>\n",
       "      <th>NS_TAXI_P_FARE</th>\n",
       "      <th>NS_TAXI_D_FARE</th>\n",
       "      <th>SAF_lag</th>\n",
       "      <th>311_lag</th>\n",
       "      <th>NYPD_lag</th>\n",
       "      <th>TAXI_P_PASSENGER_lag</th>\n",
       "      <th>TAXI_D_PASSENGER_lag</th>\n",
       "      <th>TAXI_P_FARE_lag</th>\n",
       "      <th>TAXI_D_FARE_lag</th>\n",
       "      <th>nypd_complaints_COUNT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>22</td>\n",
       "      <td>14</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>312.93</td>\n",
       "      <td>26.0</td>\n",
       "      <td>579.03</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>52105.0</td>\n",
       "      <td>374601.91</td>\n",
       "      <td>54970.0</td>\n",
       "      <td>382690.74</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>53490.0</td>\n",
       "      <td>60145.0</td>\n",
       "      <td>371204.39</td>\n",
       "      <td>424270.61</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>15</td>\n",
       "      <td>2.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>809.62</td>\n",
       "      <td>461.0</td>\n",
       "      <td>9323.87</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>42.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>405.0</td>\n",
       "      <td>471.89</td>\n",
       "      <td>8388.73</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>14</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>757.78</td>\n",
       "      <td>923.0</td>\n",
       "      <td>15469.21</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>47.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>158.0</td>\n",
       "      <td>939.0</td>\n",
       "      <td>1755.31</td>\n",
       "      <td>15156.58</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1372543200</td>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>9.0</td>\n",
       "      <td>30.0</td>\n",
       "      <td>585.0</td>\n",
       "      <td>5544.04</td>\n",
       "      <td>2652.0</td>\n",
       "      <td>35447.31</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>41.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>730.0</td>\n",
       "      <td>2552.0</td>\n",
       "      <td>7227.51</td>\n",
       "      <td>36292.56</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 45 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     datestop   x   y  SAF_COUNT  311_complaints_COUNT  \\\n",
       "0  1372543200  22  14        0.0                   0.0   \n",
       "1  1372543200  12   7        0.0                   0.0   \n",
       "2  1372543200  12  15        2.0                  12.0   \n",
       "3  1372543200  12  14        9.0                  10.0   \n",
       "4  1372543200  12  13        9.0                  30.0   \n",
       "\n",
       "   passenger_count_pickup  total_amount_pickup  passenger_count_dropoff  \\\n",
       "0                    16.0               312.93                     26.0   \n",
       "1                 52105.0            374601.91                  54970.0   \n",
       "2                    66.0               809.62                    461.0   \n",
       "3                    86.0               757.78                    923.0   \n",
       "4                   585.0              5544.04                   2652.0   \n",
       "\n",
       "   total_amount_dropoff  NN_SAF  ...  NS_TAXI_P_FARE  NS_TAXI_D_FARE  SAF_lag  \\\n",
       "0                579.03     0.0  ...             0.0             0.0      0.0   \n",
       "1             382690.74     0.0  ...             0.0             0.0      4.0   \n",
       "2               9323.87     0.0  ...             0.0             0.0      0.0   \n",
       "3              15469.21     0.0  ...             0.0             0.0     19.0   \n",
       "4              35447.31     0.0  ...             0.0             0.0      5.0   \n",
       "\n",
       "   311_lag  NYPD_lag  TAXI_P_PASSENGER_lag  TAXI_D_PASSENGER_lag  \\\n",
       "0      0.0       0.0                   0.0                   0.0   \n",
       "1     38.0       9.0               53490.0               60145.0   \n",
       "2     42.0       5.0                  38.0                 405.0   \n",
       "3     47.0      14.0                 158.0                 939.0   \n",
       "4     41.0       8.0                 730.0                2552.0   \n",
       "\n",
       "   TAXI_P_FARE_lag  TAXI_D_FARE_lag  nypd_complaints_COUNT  \n",
       "0             0.00             0.00                      1  \n",
       "1        371204.39        424270.61                      9  \n",
       "2           471.89          8388.73                      7  \n",
       "3          1755.31         15156.58                     15  \n",
       "4          7227.51         36292.56                      5  \n",
       "\n",
       "[5 rows x 45 columns]"
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_7=pd.read_csv('data_lag7.csv',index_col=0)\n",
    "data_7.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_1 = data_1[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'SAF_lag', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "y_1 = data_1[['nypd_complaints_COUNT']].to_numpy()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_7 = data_7[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'SAF_lag', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "y_7 = data_7[['nypd_complaints_COUNT']].to_numpy()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train1, X_test1, y_train1, y_test1 = train_test_split(X_1, y_1, test_size=.2, random_state=42)\n",
    "X_train7, X_test7, y_train7, y_test7 = train_test_split(X_7, y_7, test_size=.2, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import mean_squared_error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import svm\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.ensemble import GradientBoostingRegressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "need to scale for svm to work better!!\n",
    "### Support Vector machines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 2.177582588456455\n"
     ]
    }
   ],
   "source": [
    "regr1 = make_pipeline(StandardScaler(), svm.SVR())\n",
    "regr1.fit(X_train1, y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 1.5634106890945219\n"
     ]
    }
   ],
   "source": [
    "regr7 = make_pipeline(StandardScaler(), svm.SVR())\n",
    "regr7.fit(X_train7, y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr7.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn import linear_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 1.0547627202353789\n"
     ]
    }
   ],
   "source": [
    "regr1 = make_pipeline(StandardScaler(),linear_model.Lasso())\n",
    "regr1.fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.9827734606233235\n"
     ]
    }
   ],
   "source": [
    "regr7 = make_pipeline(StandardScaler(),linear_model.Lasso())\n",
    "regr7.fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr7.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.006061551621541203\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),RandomForestRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.00285839007414132\n"
     ]
    }
   ],
   "source": [
    "regr7=make_pipeline(StandardScaler(),RandomForestRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr1.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.0017773979677851042\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.00041419405719613265\n"
     ]
    }
   ],
   "source": [
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr1.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "+ trying some nearest neighbour if possible!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 3.8080243975007435\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),KNeighborsRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 3.3285398698744135\n"
     ]
    }
   ],
   "source": [
    "regr7=make_pipeline(StandardScaler(),KNeighborsRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr1.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import RadiusNeighborsRegressor\n",
    "#will not be using becasue it performs terrible here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 2.422271832366395e+37\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),RadiusNeighborsRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 2.0202495645423396e+37\n"
     ]
    }
   ],
   "source": [
    "regr7=make_pipeline(StandardScaler(),RadiusNeighborsRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr1.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Guess we're not using that one ahhahaha, trying again with SGDRegressor and MLPregressor with the standardscaler to see if it helps us with getting better values?! no not for SGDRegressor, sure it gave better result but not nearly good enought to be considered for the task."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neural_network import MLPRegressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.018537281269497174\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),MLPRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.0036999347099115187\n"
     ]
    }
   ],
   "source": [
    "regr7=make_pipeline(StandardScaler(),MLPRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr1.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## New attempt with K-fold and standardscaler in the pipeline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import ShuffleSplit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "ss = ShuffleSplit(n_splits=5, test_size=0.20,random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.03783170955537885 for shufflesplit validation lag: 1 day, model: MLPRegressor\n"
     ]
    }
   ],
   "source": [
    "aRMSE1=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_1):\n",
    "    X_train1ss, X_test1ss, y_train1ss, y_test1ss = X_1[train], X_1[test], y_1[train], y_1[test]\n",
    "    regr1=make_pipeline(StandardScaler(),MLPRegressor()).fit(X_train1ss,y_train1ss.ravel())\n",
    "    aRMSE1.append(mean_squared_error(y_test1ss.ravel(), regr1.predict(X_test1ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE1[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE1), 'for shufflesplit validation lag: 1 day, model: MLPRegressor')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.032997691503888496 for shufflesplit validation lag: 7 day, model: MLPRegressor\n"
     ]
    }
   ],
   "source": [
    "aRMSE7=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_7):\n",
    "    X_train7ss, X_test7ss, y_train7ss, y_test7ss = X_7[train], X_7[test], y_7[train], y_7[test]\n",
    "    regr7=make_pipeline(StandardScaler(),MLPRegressor()).fit(X_train7ss,y_train7ss.ravel())\n",
    "    aRMSE7.append(mean_squared_error(y_test7ss.ravel(), regr7.predict(X_test7ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE7),'for shufflesplit validation lag: 7 day, model: MLPRegressor')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.003962086548645787 for shufflesplit validation lag: 7 day, model: RandomForestRegressor\n"
     ]
    }
   ],
   "source": [
    "aRMSE7=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_7):\n",
    "    X_train7ss, X_test7ss, y_train7ss, y_test7ss = X_7[train], X_7[test], y_7[train], y_7[test]\n",
    "    regr7=make_pipeline(StandardScaler(),RandomForestRegressor()).fit(X_train7ss,y_train7ss.ravel())\n",
    "    aRMSE7.append(mean_squared_error(y_test7ss.ravel(), regr7.predict(X_test7ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE7),'for shufflesplit validation lag: 7 day, model: RandomForestRegressor')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.001198805037930691 for shufflesplit validation lag: 1 day, model: GradientBoostingRegressor\n"
     ]
    }
   ],
   "source": [
    "aRMSE1=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_1):\n",
    "    X_train1ss, X_test1ss, y_train1ss, y_test1ss = X_1[train], X_1[test], y_1[train], y_1[test]\n",
    "    regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1ss,y_train1ss.ravel())\n",
    "    aRMSE1.append(mean_squared_error(y_test1ss.ravel(), regr1.predict(X_test1ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE1),'for shufflesplit validation lag: 1 day, model: GradientBoostingRegressor')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " testing other scalers, such as normalizer aswell."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0009258135304499173\n"
     ]
    }
   ],
   "source": [
    "\n",
    "aRMSE7=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_7):\n",
    "    X_train7ss, X_test7ss, y_train7ss, y_test7ss = X_7[train], X_7[test], y_7[train], y_7[test]\n",
    "    regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7ss,y_train7ss.ravel())\n",
    "    aRMSE7.append(mean_squared_error(y_test7ss.ravel(), regr7.predict(X_test7ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import Normalizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "43.64854058049916\n"
     ]
    }
   ],
   "source": [
    "aRMSE7=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_7):\n",
    "    X_train7ss, X_test7ss, y_train7ss, y_test7ss = X_7[train], X_7[test], y_7[train], y_7[test]\n",
    "    regr7=make_pipeline(Normalizer(),GradientBoostingRegressor()).fit(X_train7ss,y_train7ss.ravel())\n",
    "    aRMSE7.append(mean_squared_error(y_test7ss.ravel(), regr7.predict(X_test7ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30.782760117946186\n"
     ]
    }
   ],
   "source": [
    "aRMSE7=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_7):\n",
    "    X_train7ss, X_test7ss, y_train7ss, y_test7ss = X_7[train], X_7[test], y_7[train], y_7[test]\n",
    "    regr7=make_pipeline(Normalizer(),RandomForestRegressor()).fit(X_train7ss,y_train7ss.ravel())\n",
    "    aRMSE7.append(mean_squared_error(y_test7ss.ravel(), regr7.predict(X_test7ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE7[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "67.68220243550618\n"
     ]
    }
   ],
   "source": [
    "aRMSE1=[]\n",
    "i=0\n",
    "for train, test in ss.split(X_1):\n",
    "    X_train1ss, X_test1ss, y_train1ss, y_test1ss = X_1[train], X_1[test], y_1[train], y_1[test]\n",
    "    regr1=make_pipeline(Normalizer(),linear_model.Lasso()).fit(X_train1ss,y_train1ss.ravel())\n",
    "    aRMSE1.append(mean_squared_error(y_test1ss.ravel(), regr1.predict(X_test1ss), multioutput='uniform_average'))\n",
    "    #print('aRMSE',aRMSE1[i], 'for one version of KFold', i)\n",
    "    #i+=1\n",
    "print(np.mean(aRMSE1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "so, standardscaler is SIGNIFICANTLY better than normalizer here."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# now correlation matrixes & feature analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(data_7.corr(),cmap=\"Blues\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Too hard to see anything, aswell as I dont think all features can be considered. Will only check the ones for today, aswell as the NYPD-values for all the neighbours and the lag."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(data_1[[ 'SAF_COUNT', '311_complaints_COUNT',\n",
    "       'passenger_count_pickup', 'total_amount_pickup',\n",
    "       'passenger_count_dropoff', 'total_amount_dropoff','NN_NYPD','NW_NYPD','NS_NYPD','NE_NYPD','NYPD_lag','nypd_complaints_COUNT']].corr(),cmap=\"Blues\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.heatmap(data_7[[ 'SAF_COUNT', '311_complaints_COUNT',\n",
    "       'passenger_count_pickup', 'total_amount_pickup',\n",
    "       'passenger_count_dropoff', 'total_amount_dropoff','NN_NYPD','NW_NYPD','NS_NYPD','NE_NYPD','NYPD_lag','nypd_complaints_COUNT']].corr(),cmap=\"Blues\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## gradientBoostingRegressor but minus 1 feature and its values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7,y_train7.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7.ravel(), regr7.predict(X_test7), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1,y_train1.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1.ravel(), regr1.predict(X_test1), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_1_saf = data_1[['datestop', 'x', 'y', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "y_1 = data_1[['nypd_complaints_COUNT']].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_1_311 = data_1[['datestop', 'x', 'y', 'SAF_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'SAF_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_1_taxi = data_1[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NS_SAF', 'NS_311', 'NS_NYPD', 'SAF_lag', '311_lag', 'NYPD_lag', 'nypd_complaints_COUNT']].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_1_neigh = data_1[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'SAF_lag', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "\n",
    "X_1_lag = data_1[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'nypd_complaints_COUNT']].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_7_saf = data_7[['datestop', 'x', 'y', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "y_7 = data_7[['nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "X_7_311 = data_7[['datestop', 'x', 'y', 'SAF_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'SAF_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "X_7_taxi = data_7[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NS_SAF', 'NS_311', 'NS_NYPD', 'SAF_lag', '311_lag', 'NYPD_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "\n",
    "X_7_neigh = data_7[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'SAF_lag', '311_lag', 'NYPD_lag',\n",
    "\n",
    "'TAXI_P_PASSENGER_lag', 'TAXI_D_PASSENGER_lag', 'TAXI_P_FARE_lag',\n",
    "\n",
    "'TAXI_D_FARE_lag', 'nypd_complaints_COUNT']].to_numpy()\n",
    "\n",
    "\n",
    "X_7_lag = data_7[['datestop', 'x', 'y', 'SAF_COUNT', '311_complaints_COUNT',\n",
    "\n",
    "'passenger_count_pickup', 'total_amount_pickup',\n",
    "\n",
    "'passenger_count_dropoff', 'total_amount_dropoff', 'NN_SAF', 'NN_311', 'NN_NYPD',\n",
    "\n",
    "'NN_TAXI_P_PASSENGER', 'NN_TAXI_D_PASSENGER', 'NN_TAXI_P_FARE',\n",
    "\n",
    "'NN_TAXI_D_FARE', 'NW_SAF', 'NW_311', 'NW_NYPD', 'NW_TAXI_P_PASSENGER',\n",
    "\n",
    "'NW_TAXI_D_PASSENGER', 'NW_TAXI_P_FARE', 'NW_TAXI_D_FARE', 'NE_SAF',\n",
    "\n",
    "'NE_311', 'NE_NYPD', 'NE_TAXI_P_PASSENGER', 'NE_TAXI_D_PASSENGER',\n",
    "\n",
    "'NE_TAXI_P_FARE', 'NE_TAXI_D_FARE', 'NS_SAF', 'NS_311', 'NS_NYPD',\n",
    "\n",
    "'NS_TAXI_P_PASSENGER', 'NS_TAXI_D_PASSENGER', 'NS_TAXI_P_FARE',\n",
    "\n",
    "'NS_TAXI_D_FARE', 'nypd_complaints_COUNT']].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train1_311, X_test1_311, y_train1_311, y_test1_311 = train_test_split(X_1_311, y_1, test_size=.2, random_state=42)\n",
    "X_train7_311, X_test7_311, y_train7_311, y_test7_311 = train_test_split(X_7_311, y_7, test_size=.2, random_state=42)\n",
    "\n",
    "X_train1_saf, X_test1_saf, y_train1_saf, y_test1_saf = train_test_split(X_1_saf, y_1, test_size=.2, random_state=42)\n",
    "X_train7_saf, X_test7_saf, y_train7_saf, y_test7_saf = train_test_split(X_7_saf, y_7, test_size=.2, random_state=42)\n",
    "\n",
    "X_train1_taxi, X_test1_taxi, y_train1_taxi, y_test1_taxi = train_test_split(X_1_taxi, y_1, test_size=.2, random_state=42)\n",
    "X_train7_taxi, X_test7_taxi, y_train7_taxi, y_test7_taxi = train_test_split(X_7_taxi, y_7, test_size=.2, random_state=42)\n",
    "\n",
    "X_train1_neigh, X_test1_neigh, y_train1_neigh, y_test1_neigh = train_test_split(X_1_neigh, y_1, test_size=.2, random_state=42)\n",
    "X_train7_neigh, X_test7_neigh, y_train7_neigh, y_test7_neigh = train_test_split(X_7_neigh, y_7, test_size=.2, random_state=42)\n",
    "\n",
    "X_train1_lag, X_test1_lag, y_train1_lag, y_test1_lag = train_test_split(X_1_lag, y_1, test_size=.2, random_state=42)\n",
    "X_train7_lag, X_test7_lag, y_train7_lag, y_test7_lag = train_test_split(X_7_lag, y_7, test_size=.2, random_state=42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "311 complaints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.0017857391517231856 for GradientBoostingRegressor minus 311 complaints lag 1\n",
      "aRMSE 0.0013895284485207606 for GradientBoostingRegressor minus 311 complaints lag 7\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1_311,y_train1_311.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1_311.ravel(), regr1.predict(X_test1_311), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1,'for GradientBoostingRegressor minus 311 complaints lag 1')\n",
    "\n",
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7_311,y_train7_311.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7_311.ravel(), regr7.predict(X_test7_311), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7,'for GradientBoostingRegressor minus 311 complaints lag 7')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "saf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.001826921884520987 for GradientBoostingRegressor minus saf complaints lag 1\n",
      "aRMSE 0.0013976282502486527 for GradientBoostingRegressor minus saf complaints lag 7\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1_saf,y_train1_saf.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1_saf.ravel(), regr1.predict(X_test1_saf), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1,'for GradientBoostingRegressor minus saf complaints lag 1')\n",
    "\n",
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7_saf,y_train7_saf.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7_saf.ravel(), regr7.predict(X_test7_saf), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7,'for GradientBoostingRegressor minus saf complaints lag 7')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "taxi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.0018795194244052489 for GradientBoostingRegressor minus taxi complaints lag 1\n",
      "aRMSE 0.0010761305678599036 for GradientBoostingRegressor minus taxi complaints lag 7\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1_taxi,y_train1_taxi.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1_taxi.ravel(), regr1.predict(X_test1_taxi), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1,'for GradientBoostingRegressor minus taxi complaints lag 1')\n",
    "\n",
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7_taxi,y_train7_taxi.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7_taxi.ravel(), regr7.predict(X_test7_taxi), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7,'for GradientBoostingRegressor minus taxi complaints lag 7')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "neighbours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.002132407028508389 for GradientBoostingRegressor minus neigh lag 1\n",
      "aRMSE 0.0012533750957996242 for GradientBoostingRegressor minus neighbour info lag 7\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1_neigh,y_train1_neigh.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1_neigh.ravel(), regr1.predict(X_test1_neigh), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1,'for GradientBoostingRegressor minus neigh lag 1')\n",
    "\n",
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7_neigh,y_train7_neigh.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7_neigh.ravel(), regr7.predict(X_test7_neigh), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7,'for GradientBoostingRegressor minus neighbour info lag 7')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "days lag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "aRMSE 0.0015688863819538032 for GradientBoostingRegressor minus lag 1 \n",
      "aRMSE 0.0011756070922773608 for GradientBoostingRegressor minus lag 7\n"
     ]
    }
   ],
   "source": [
    "regr1=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train1_lag,y_train1_lag.ravel())\n",
    "aRMSE1=mean_squared_error(y_test1_lag.ravel(), regr1.predict(X_test1_lag), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE1,'for GradientBoostingRegressor minus lag 1 ')\n",
    "\n",
    "regr7=make_pipeline(StandardScaler(),GradientBoostingRegressor()).fit(X_train7_lag,y_train7_lag.ravel())\n",
    "aRMSE7=mean_squared_error(y_test7_lag.ravel(), regr7.predict(X_test7_lag), multioutput='uniform_average')\n",
    "print('aRMSE',aRMSE7,'for GradientBoostingRegressor minus lag 7')"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "4bd0aa0d2cfa261ac92b919594313a561b65a0bc91bd6de20b7d045bb52fdc0a"
  },
  "kernelspec": {
   "display_name": "Python 3.9.9 64-bit (windows store)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.9"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}