Machine_Learning_Project-3-RF.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Machine Learning Project.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOl5homFaj8yXr3rgNHKvtp",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/OsandaDeemantha/Fare-Classification/blob/master/Machine_Learning_Project-3-RF.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AERv1M0io4kv",
        "colab_type": "text"
      },
      "source": [
        "# **Import Libraries**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pLQWs4-rfIGs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import itertools\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from matplotlib.ticker import NullFormatter\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.ticker as ticker\n",
        "from sklearn import preprocessing\n",
        "%matplotlib inline\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\", category=DeprecationWarning)\n",
        "warnings.simplefilter(action='ignore', category=FutureWarning)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r2X-Mm8MphDu",
        "colab_type": "text"
      },
      "source": [
        "# **Load Data**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uUOLUUWMcR57",
        "colab_type": "code",
        "outputId": "056a96a2-c200-4ed7-b062-5cbe46d939e1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "df = pd.read_csv('/content/drive/My Drive/Colab Notebooks/Fare-Classification/Data/train.csv')\n",
        "df['label'].replace(to_replace=['correct','incorrect'], value=[1,0],inplace=True)"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ukrvZeuY7Pdu",
        "colab_type": "text"
      },
      "source": [
        "# **Explore Data**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JQX7NAtQ7XHr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        },
        "outputId": "fe0d5524-3905-44c9-f50a-6c2edd82e5bd"
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 17176 entries, 0 to 17175\n",
            "Data columns (total 14 columns):\n",
            " #   Column                     Non-Null Count  Dtype  \n",
            "---  ------                     --------------  -----  \n",
            " 0   tripid                     17176 non-null  int64  \n",
            " 1   additional_fare            16974 non-null  float64\n",
            " 2   duration                   16974 non-null  float64\n",
            " 3   meter_waiting              16974 non-null  float64\n",
            " 4   meter_waiting_fare         16974 non-null  float64\n",
            " 5   meter_waiting_till_pickup  16974 non-null  float64\n",
            " 6   pickup_time                17176 non-null  object \n",
            " 7   drop_time                  17176 non-null  object \n",
            " 8   pick_lat                   17176 non-null  float64\n",
            " 9   pick_lon                   17176 non-null  float64\n",
            " 10  drop_lat                   17176 non-null  float64\n",
            " 11  drop_lon                   17176 non-null  float64\n",
            " 12  fare                       17039 non-null  float64\n",
            " 13  label                      17176 non-null  int64  \n",
            "dtypes: float64(10), int64(2), object(2)\n",
            "memory usage: 1.8+ MB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_kbBa3od5jkS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 309
        },
        "outputId": "a56d119b-d751-4900-c6ec-3d2a3eed4593"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 54,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>tripid</th>\n",
              "      <th>additional_fare</th>\n",
              "      <th>duration</th>\n",
              "      <th>meter_waiting</th>\n",
              "      <th>meter_waiting_fare</th>\n",
              "      <th>meter_waiting_till_pickup</th>\n",
              "      <th>pickup_time</th>\n",
              "      <th>drop_time</th>\n",
              "      <th>pick_lat</th>\n",
              "      <th>pick_lon</th>\n",
              "      <th>drop_lat</th>\n",
              "      <th>drop_lon</th>\n",
              "      <th>fare</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>189123628</td>\n",
              "      <td>10.5</td>\n",
              "      <td>834.0</td>\n",
              "      <td>56.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>64.0</td>\n",
              "      <td>11/1/2019 0:20</td>\n",
              "      <td>11/1/2019 0:34</td>\n",
              "      <td>6.86252</td>\n",
              "      <td>79.8993</td>\n",
              "      <td>6.90330</td>\n",
              "      <td>79.8783</td>\n",
              "      <td>270.32</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>189125358</td>\n",
              "      <td>10.5</td>\n",
              "      <td>791.0</td>\n",
              "      <td>47.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>134.0</td>\n",
              "      <td>11/1/2019 0:56</td>\n",
              "      <td>11/1/2019 1:09</td>\n",
              "      <td>6.88589</td>\n",
              "      <td>79.8984</td>\n",
              "      <td>6.91373</td>\n",
              "      <td>79.8923</td>\n",
              "      <td>197.85</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>189125719</td>\n",
              "      <td>10.5</td>\n",
              "      <td>1087.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>61.0</td>\n",
              "      <td>11/1/2019 1:08</td>\n",
              "      <td>11/1/2019 1:26</td>\n",
              "      <td>6.90839</td>\n",
              "      <td>79.8651</td>\n",
              "      <td>6.93669</td>\n",
              "      <td>79.9146</td>\n",
              "      <td>301.64</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>189127273</td>\n",
              "      <td>10.5</td>\n",
              "      <td>598.0</td>\n",
              "      <td>271.0</td>\n",
              "      <td>15.6638</td>\n",
              "      <td>68.0</td>\n",
              "      <td>11/1/2019 2:27</td>\n",
              "      <td>11/1/2019 2:37</td>\n",
              "      <td>6.92570</td>\n",
              "      <td>79.8895</td>\n",
              "      <td>6.92748</td>\n",
              "      <td>79.8971</td>\n",
              "      <td>82.30</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>189128020</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>11/1/2019 3:34</td>\n",
              "      <td>11/1/2019 3:51</td>\n",
              "      <td>6.87441</td>\n",
              "      <td>79.8615</td>\n",
              "      <td>6.84478</td>\n",
              "      <td>79.9290</td>\n",
              "      <td>358.39</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      tripid  additional_fare  duration  ...  drop_lon    fare  label\n",
              "0  189123628             10.5     834.0  ...   79.8783  270.32      1\n",
              "1  189125358             10.5     791.0  ...   79.8923  197.85      1\n",
              "2  189125719             10.5    1087.0  ...   79.9146  301.64      1\n",
              "3  189127273             10.5     598.0  ...   79.8971   82.30      1\n",
              "4  189128020              NaN       NaN  ...   79.9290  358.39      1\n",
              "\n",
              "[5 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 54
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nkG78k3T7-LD",
        "colab_type": "text"
      },
      "source": [
        "**Label**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0t1iN8aO6gtH",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 382
        },
        "outputId": "ce1a7303-70b6-4de7-a37b-265f8f1122dd"
      },
      "source": [
        "print(df['label'].value_counts(normalize=True))\n",
        "C, I = df['label'].value_counts()\n",
        "print('Correct: ',C)\n",
        "print('Incorrect : ',I)\n",
        "sns.countplot(df['label'])"
      ],
      "execution_count": 55,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1    0.902131\n",
            "0    0.097869\n",
            "Name: label, dtype: float64\n",
            "Correct:  15495\n",
            "Incorrect :  1681\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f813e797978>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 55
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b8bMkfK2qnRH",
        "colab_type": "text"
      },
      "source": [
        "**Plot Heat Map**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VZlRXCoUqGN5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 712
        },
        "outputId": "357c6aa2-aaa8-4383-899f-890708125bd2"
      },
      "source": [
        "corrmat = df.corr()\n",
        "top_corr_features = corrmat.index\n",
        "plt.figure(figsize=(10,10))\n",
        "#plot heat map\n",
        "g=sns.heatmap(df[top_corr_features].corr(),annot=True,cmap=\"RdYlGn\")"
      ],
      "execution_count": 56,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TqB59-Fj8SIX",
        "colab_type": "text"
      },
      "source": [
        "# **Feature Engineering**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jiZlAWALbQE0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def calc_prevalence(y):\n",
        " return (sum(y)/len(y))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XGUd-bLqyJaO",
        "colab_type": "text"
      },
      "source": [
        "**Split Date Time**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NSxiWINMks1n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import datetime\n",
        "from datetime import timedelta\n",
        "\n",
        "def extractDataFromDateTimeObj(time):\n",
        "  datetimeFormat = '%m/%d/%Y %H:%M'\n",
        "  dateTimeObj = datetime.datetime.strptime(time, datetimeFormat)\n",
        "  return [dateTimeObj.year, dateTimeObj.month, dateTimeObj.day, dateTimeObj.hour, dateTimeObj.minute, dateTimeObj.weekday(), dateTimeObj.weekday()]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WIlPE5aMk0w_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def generateDropTimeData(df):\n",
        "  df_drop_year = []\n",
        "  df_drop_month = []\n",
        "  df_drop_day = []\n",
        "  df_drop_hour = []\n",
        "  df_drop_minute = []\n",
        "  df_drop_weekday = []\n",
        "  df_drop_weekofyear = []\n",
        "\n",
        "  for time in df['drop_time']:\n",
        "    data = extractDataFromDateTimeObj(time)\n",
        "    df_drop_year.append(data[0])\n",
        "    df_drop_month.append(data[1])\n",
        "    df_drop_day.append(data[2])\n",
        "    df_drop_hour.append(data[3])\n",
        "    df_drop_minute.append(data[4])\n",
        "    df_drop_weekday.append(data[5])\n",
        "    df_drop_weekofyear.append(data[6])\n",
        "\n",
        "  df['drop_year'] = df_drop_year\n",
        "  df['drop_month'] = df_drop_month\n",
        "  df['drop_day'] = df_drop_day\n",
        "  df['drop_hour'] = df_drop_hour\n",
        "  df['drop_minute'] = df_drop_minute\n",
        "  df['drop_weekday'] = df_drop_weekday\n",
        "  df['drop_weekofyear'] = df_drop_weekofyear"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UvOilL6Vk5Rt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def generatePickupTimeData(df):\n",
        "  df_pickup_year = []\n",
        "  df_pickup_month = []\n",
        "  df_pickup_day = []\n",
        "  df_pickup_hour = []\n",
        "  df_pickup_minute = []\n",
        "  df_pickup_weekday = []\n",
        "  df_pickup_weekofyear = []\n",
        "\n",
        "  for time in df['pickup_time']:\n",
        "    data = extractDataFromDateTimeObj(time)\n",
        "    df_pickup_year.append(data[0])\n",
        "    df_pickup_month.append(data[1])\n",
        "    df_pickup_day.append(data[2])\n",
        "    df_pickup_hour.append(data[3])\n",
        "    df_pickup_minute.append(data[4])\n",
        "    df_pickup_weekday.append(data[5])\n",
        "    df_pickup_weekofyear.append(data[6])\n",
        "\n",
        "  df['pickup_year'] = df_pickup_year\n",
        "  df['pickup_month'] = df_pickup_month\n",
        "  df['pickup_day'] = df_pickup_day\n",
        "  df['pickup_hour'] = df_pickup_hour\n",
        "  df['pickup_minute'] = df_pickup_minute\n",
        "  df['pickup_weekday'] = df_pickup_weekday\n",
        "  df['pickup_weekofyear'] = df_pickup_weekofyear"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DYxERgMcjUIk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def generatePickupTimestamp(df):\n",
        "  time = pd.to_datetime(df['pickup_time'])\n",
        "  df['pickup_timestamp'] = time.values.astype(np.int64) // 10 ** 9\n",
        "\n",
        "def generateDropTimestamp(df):\n",
        "  time = pd.to_datetime(df['drop_time'])\n",
        "  df['drop_timestamp'] = time.values.astype(np.int64) // 10 ** 9"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cZRklygEifqe",
        "colab_type": "text"
      },
      "source": [
        "**Calculate Distance**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zLUb3CmCidKe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from geopy.distance import geodesic\n",
        "\n",
        "def calculateDistance(pick_lat,pick_lon, drop_lat, drop_lon):\n",
        "\n",
        "  coords_1 = (pick_lat, pick_lon)\n",
        "  coords_2 = (drop_lat, drop_lon)\n",
        "  #pick_lat = radians(float(pick_lat))\n",
        "  #pick_lon = radians(float(pick_lon))\n",
        "  #drop_lat = radians(float(drop_lat))\n",
        "  #drop_lon = radians(float(drop_lon))\n",
        "\n",
        "  if (pick_lon == drop_lon and pick_lat == drop_lat):\n",
        "    distance = 0.0\n",
        "    return round(distance, 2)\n",
        "  else:\n",
        "    distance = geodesic(coords_1, coords_2).km\n",
        "    return round(distance, 2)\n",
        "\n",
        "def calculateDistanceDf(pick_lat_frame,pick_lon_frame, drop_lat_frame, drop_lon_frame):\n",
        "  distanceList = []\n",
        "  for i in range(len(drop_lon_frame)):\n",
        "    pick_lat = pick_lat_frame[i]\n",
        "    pick_lon = pick_lon_frame[i]\n",
        "    drop_lat = drop_lat_frame[i]\n",
        "    drop_lon = drop_lon_frame[i]\n",
        "\n",
        "    try:\n",
        "      distance = calculateDistance(pick_lat,pick_lon, drop_lat, drop_lon)\n",
        "    except:\n",
        "      print(\"Error occured! , at\", i)\n",
        "\n",
        "    distanceList.append(distance)\n",
        "  return distanceList"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EaEn4Yka0lxa",
        "colab_type": "text"
      },
      "source": [
        "**Imputation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3g3ZzYCurg6g",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.impute import SimpleImputer\n",
        "\n",
        "def naImpute(features, column_names):\n",
        "  imp=SimpleImputer(missing_values=np.NaN)\n",
        "  idf=pd.DataFrame(imp.fit_transform(features))\n",
        "  idf.columns=features.columns\n",
        "  idf.index=features.index\n",
        "\n",
        "  for column_name in column_names:\n",
        "    idf[column_name].isna().sum()\n",
        "\n",
        "  \n",
        "  return idf"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fm0YHE4CJy56",
        "colab_type": "text"
      },
      "source": [
        "**Split Training/ Test Data**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jyPpDz3xsdCX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# from sklearn.model_selection import train_test_split\n",
        "# X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n",
        "\n",
        "# shuffle the samples\n",
        "# df = df.sample(n = len(df), random_state = 42)\n",
        "# df = df.reset_index(drop = True)\n",
        "# valid_df = df.sample(frac = 0.2, random_state = 42)\n",
        "# train_df = df.drop(valid_df.index)\n",
        "valid_df = df\n",
        "train_df = df"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZR03kJU_ayIg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "0b4c20e3-4951-4357-9fa7-5e8ba4afd419"
      },
      "source": [
        "print('Valid prevalence(n = %d):%.3f'%(len(valid_df),calc_prevalence(valid_df.label.values)))\n",
        "print('Train prevalence(n = %d):%.3f'%(len(train_df), calc_prevalence(train_df.label.values)))"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Valid prevalence(n = 17176):0.902\n",
            "Train prevalence(n = 17176):0.902\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cyMldPKVM-u_",
        "colab_type": "text"
      },
      "source": [
        "**Pre Process**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5xbmwnOLpBTe",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def preProcess(df, features):\n",
        "  generatePickupTimeData(df)\n",
        "  generateDropTimeData(df)\n",
        "  generatePickupTimestamp(df)\n",
        "  generateDropTimestamp(df)\n",
        "\n",
        "  df['distance'] = calculateDistanceDf(df['pick_lat'], df['pick_lon'], df['drop_lat'], df['drop_lon'])\n",
        "\n",
        "  features += ['pickup_year', 'pickup_month', 'pickup_day', 'pickup_hour', 'pickup_minute', 'pickup_weekday', 'pickup_weekofyear', 'drop_year', 'drop_month', 'drop_day', 'drop_hour', 'drop_minute', 'drop_weekday', 'drop_weekofyear', 'distance','pickup_timestamp','drop_timestamp']\n",
        "  Features = df[features]\n",
        "  Features = naImpute(Features,features)\n",
        "  return Features"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PAx8KviOqbfH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def getX(Features):\n",
        "    X= preprocessing.StandardScaler().fit(Features).transform(Features)\n",
        "    return X"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xZ3U83W3ZZjm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 309
        },
        "outputId": "c0eb3668-2a04-48b5-c090-2d6c099d5f84"
      },
      "source": [
        "train_y = train_df.label\n",
        "list = ['tripid','label']\n",
        "train_x = train_df.drop(list,axis = 1 )\n",
        "train_x.head()"
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>additional_fare</th>\n",
              "      <th>duration</th>\n",
              "      <th>meter_waiting</th>\n",
              "      <th>meter_waiting_fare</th>\n",
              "      <th>meter_waiting_till_pickup</th>\n",
              "      <th>pickup_time</th>\n",
              "      <th>drop_time</th>\n",
              "      <th>pick_lat</th>\n",
              "      <th>pick_lon</th>\n",
              "      <th>drop_lat</th>\n",
              "      <th>drop_lon</th>\n",
              "      <th>fare</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10.5</td>\n",
              "      <td>834.0</td>\n",
              "      <td>56.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>64.0</td>\n",
              "      <td>11/1/2019 0:20</td>\n",
              "      <td>11/1/2019 0:34</td>\n",
              "      <td>6.86252</td>\n",
              "      <td>79.8993</td>\n",
              "      <td>6.90330</td>\n",
              "      <td>79.8783</td>\n",
              "      <td>270.32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.5</td>\n",
              "      <td>791.0</td>\n",
              "      <td>47.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>134.0</td>\n",
              "      <td>11/1/2019 0:56</td>\n",
              "      <td>11/1/2019 1:09</td>\n",
              "      <td>6.88589</td>\n",
              "      <td>79.8984</td>\n",
              "      <td>6.91373</td>\n",
              "      <td>79.8923</td>\n",
              "      <td>197.85</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>10.5</td>\n",
              "      <td>1087.0</td>\n",
              "      <td>80.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>61.0</td>\n",
              "      <td>11/1/2019 1:08</td>\n",
              "      <td>11/1/2019 1:26</td>\n",
              "      <td>6.90839</td>\n",
              "      <td>79.8651</td>\n",
              "      <td>6.93669</td>\n",
              "      <td>79.9146</td>\n",
              "      <td>301.64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>10.5</td>\n",
              "      <td>598.0</td>\n",
              "      <td>271.0</td>\n",
              "      <td>15.6638</td>\n",
              "      <td>68.0</td>\n",
              "      <td>11/1/2019 2:27</td>\n",
              "      <td>11/1/2019 2:37</td>\n",
              "      <td>6.92570</td>\n",
              "      <td>79.8895</td>\n",
              "      <td>6.92748</td>\n",
              "      <td>79.8971</td>\n",
              "      <td>82.30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>11/1/2019 3:34</td>\n",
              "      <td>11/1/2019 3:51</td>\n",
              "      <td>6.87441</td>\n",
              "      <td>79.8615</td>\n",
              "      <td>6.84478</td>\n",
              "      <td>79.9290</td>\n",
              "      <td>358.39</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   additional_fare  duration  meter_waiting  ...  drop_lat  drop_lon    fare\n",
              "0             10.5     834.0           56.0  ...   6.90330   79.8783  270.32\n",
              "1             10.5     791.0           47.0  ...   6.91373   79.8923  197.85\n",
              "2             10.5    1087.0           80.0  ...   6.93669   79.9146  301.64\n",
              "3             10.5     598.0          271.0  ...   6.92748   79.8971   82.30\n",
              "4              NaN       NaN            NaN  ...   6.84478   79.9290  358.39\n",
              "\n",
              "[5 rows x 12 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZDG1bYgZqxzr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        },
        "outputId": "af542255-ba07-4513-ba52-1591a9cd773a"
      },
      "source": [
        "features = ['additional_fare','duration', 'meter_waiting', 'meter_waiting_fare', 'meter_waiting_till_pickup', 'fare']\n",
        "train_x = preProcess(train_df, features)\n",
        "train_x.head()"
      ],
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>additional_fare</th>\n",
              "      <th>duration</th>\n",
              "      <th>meter_waiting</th>\n",
              "      <th>meter_waiting_fare</th>\n",
              "      <th>meter_waiting_till_pickup</th>\n",
              "      <th>fare</th>\n",
              "      <th>pickup_year</th>\n",
              "      <th>pickup_month</th>\n",
              "      <th>pickup_day</th>\n",
              "      <th>pickup_hour</th>\n",
              "      <th>pickup_minute</th>\n",
              "      <th>pickup_weekday</th>\n",
              "      <th>pickup_weekofyear</th>\n",
              "      <th>drop_year</th>\n",
              "      <th>drop_month</th>\n",
              "      <th>drop_day</th>\n",
              "      <th>drop_hour</th>\n",
              "      <th>drop_minute</th>\n",
              "      <th>drop_weekday</th>\n",
              "      <th>drop_weekofyear</th>\n",
              "      <th>distance</th>\n",
              "      <th>pickup_timestamp</th>\n",
              "      <th>drop_timestamp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10.500000</td>\n",
              "      <td>834.000000</td>\n",
              "      <td>56.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>64.000000</td>\n",
              "      <td>270.32</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>5.07</td>\n",
              "      <td>1.572568e+09</td>\n",
              "      <td>1.572568e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.500000</td>\n",
              "      <td>791.000000</td>\n",
              "      <td>47.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>134.000000</td>\n",
              "      <td>197.85</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>56.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>9.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>3.15</td>\n",
              "      <td>1.572570e+09</td>\n",
              "      <td>1.572571e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>10.500000</td>\n",
              "      <td>1087.000000</td>\n",
              "      <td>80.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>61.000000</td>\n",
              "      <td>301.64</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>8.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>6.30</td>\n",
              "      <td>1.572570e+09</td>\n",
              "      <td>1.572572e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>10.500000</td>\n",
              "      <td>598.000000</td>\n",
              "      <td>271.000000</td>\n",
              "      <td>15.663800</td>\n",
              "      <td>68.000000</td>\n",
              "      <td>82.30</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>27.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>37.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>0.86</td>\n",
              "      <td>1.572575e+09</td>\n",
              "      <td>1.572576e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>13.719651</td>\n",
              "      <td>1702.858077</td>\n",
              "      <td>629.074231</td>\n",
              "      <td>32.057666</td>\n",
              "      <td>112.466832</td>\n",
              "      <td>358.39</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>11.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>51.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>4.0</td>\n",
              "      <td>8.15</td>\n",
              "      <td>1.572579e+09</td>\n",
              "      <td>1.572580e+09</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   additional_fare     duration  ...  pickup_timestamp  drop_timestamp\n",
              "0        10.500000   834.000000  ...      1.572568e+09    1.572568e+09\n",
              "1        10.500000   791.000000  ...      1.572570e+09    1.572571e+09\n",
              "2        10.500000  1087.000000  ...      1.572570e+09    1.572572e+09\n",
              "3        10.500000   598.000000  ...      1.572575e+09    1.572576e+09\n",
              "4        13.719651  1702.858077  ...      1.572579e+09    1.572580e+09\n",
              "\n",
              "[5 rows x 23 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "80zQDjiYc6AU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "outputId": "00a5d83a-de52-4807-c441-6ce5d220b5fc"
      },
      "source": [
        "ax = sns.countplot(train_x.pickup_weekday,label=\"pickup_weekday\")\n",
        "A, B, C, D, E, F, G = train_x.pickup_weekday.value_counts()\n",
        "print(A,B,C,D,E,F,G)"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2754 2572 2523 2501 2450 2436 1940\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mPq78IXkdB5m",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "outputId": "1d16138e-e74e-4911-a7ed-93ab929207f7"
      },
      "source": [
        "ax = sns.countplot(train_x.drop_month,label=\"drop_month\")\n",
        "A, B, C = train_x.drop_month.value_counts()\n",
        "print('1=',A,',11=',B,',12=',C)"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1= 5952 ,11= 5827 ,12= 5397\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IBjlsJCrN4K-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "0b967d6f-8992-46cf-e3dc-c55fc0ce5017"
      },
      "source": [
        "y = train_y.values\n",
        "y[0:5]"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, 1, 1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "p4pmLqpHLXPI",
        "colab_type": "text"
      },
      "source": [
        "**Normalize Data**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YZG0mLAnrxbd",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 697
        },
        "outputId": "29a670dc-b959-4f81-93bb-6b6e5112f7ad"
      },
      "source": [
        "X= getX(train_x)\n",
        "X[0:5]"
      ],
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[-3.27238746e-02, -1.07638075e-01, -7.14960240e-02,\n",
              "        -6.92802181e-02, -1.10293576e-01, -7.02239052e-02,\n",
              "        -7.16359299e-01,  6.09098301e-01, -1.68354365e+00,\n",
              "        -2.86436081e+00, -5.47836750e-01,  5.61094962e-01,\n",
              "         5.61094962e-01, -7.16545398e-01,  6.09275746e-01,\n",
              "        -1.68364402e+00, -2.88449913e+00,  2.63958702e-01,\n",
              "         5.60211082e-01,  5.60211082e-01,  5.22623770e-03,\n",
              "        -1.75706685e+00, -1.75755182e+00],\n",
              "       [-3.27238746e-02, -1.12965110e-01, -7.26188527e-02,\n",
              "        -6.92802181e-02,  4.90019681e-02, -1.90433433e-01,\n",
              "        -7.16359299e-01,  6.09098301e-01, -1.68354365e+00,\n",
              "        -2.86436081e+00,  1.53456998e+00,  5.61094962e-01,\n",
              "         5.61094962e-01, -7.16545398e-01,  6.09275746e-01,\n",
              "        -1.68364402e+00, -2.67986929e+00, -1.17456128e+00,\n",
              "         5.60211082e-01,  5.60211082e-01, -3.94734695e-02,\n",
              "        -1.75612359e+00, -1.75663471e+00],\n",
              "       [-3.27238746e-02, -7.62952888e-02, -6.85018140e-02,\n",
              "        -6.92802181e-02, -1.17120528e-01, -1.82718918e-02,\n",
              "        -7.16359299e-01,  6.09098301e-01, -1.68354365e+00,\n",
              "        -2.65706185e+00, -1.24197233e+00,  5.61094962e-01,\n",
              "         5.61094962e-01, -7.16545398e-01,  6.09275746e-01,\n",
              "        -1.68364402e+00, -2.67986929e+00, -1.96367693e-01,\n",
              "         5.60211082e-01,  5.60211082e-01,  3.38619876e-02,\n",
              "        -1.75580917e+00, -1.75618926e+00],\n",
              "       [-3.27238746e-02, -1.36874824e-01, -4.46728930e-02,\n",
              "        -3.54289860e-02, -1.01190973e-01, -3.82101860e-01,\n",
              "        -7.16359299e-01,  6.09098301e-01, -1.68354365e+00,\n",
              "        -2.44976290e+00, -1.42924331e-01,  5.61094962e-01,\n",
              "         5.61094962e-01, -7.16545398e-01,  6.09275746e-01,\n",
              "        -1.68364402e+00, -2.47523946e+00,  4.36581100e-01,\n",
              "         5.60211082e-01,  5.60211082e-01, -9.27871827e-02,\n",
              "        -1.75373925e+00, -1.75432884e+00],\n",
              "       [ 1.80545298e-17,  0.00000000e+00, -1.41834276e-17,\n",
              "         0.00000000e+00,  3.23389405e-17,  7.58621043e-02,\n",
              "        -7.16359299e-01,  6.09098301e-01, -1.68354365e+00,\n",
              "        -2.24246394e+00,  2.61988088e-01,  5.61094962e-01,\n",
              "         5.61094962e-01, -7.16545398e-01,  6.09275746e-01,\n",
              "        -1.68364402e+00, -2.27060962e+00,  1.24215229e+00,\n",
              "         5.60211082e-01,  5.60211082e-01,  7.69320179e-02,\n",
              "        -1.75198374e+00, -1.75238980e+00]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t4SiYWA3YWBq",
        "colab_type": "text"
      },
      "source": [
        "# **Random Forest**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m2wCi2dL-oEx",
        "colab_type": "text"
      },
      "source": [
        "# **Hyperparameter Tuning**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uKOkmWkO-q5C",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 232
        },
        "outputId": "235d9faf-bb86-40b7-bdb2-07d3eaf93e41"
      },
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "rf = RandomForestClassifier(max_features='auto', oob_score=True, random_state=1, n_jobs=-1)\n",
        "\n",
        "param_grid = { \"criterion\" : [\"gini\", \"entropy\"], \"min_samples_leaf\" : [1, 5, 10], \"min_samples_split\" : [2, 4, 10, 12, 16], \"n_estimators\": [50, 100, 400, 700, 1000]}\n",
        "\n",
        "gs = GridSearchCV(estimator=rf, param_grid=param_grid, scoring='accuracy', cv=3, n_jobs=-1)\n",
        "\n",
        "gs = gs.fit(train_x, train_y)"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "error",
          "ename": "AttributeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-75-5155b32a9118>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     10\u001b[0m \u001b[0mgs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrain_x\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_y\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbestscore\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     13\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbestparams\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mgs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcvresults\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mAttributeError\u001b[0m: 'GridSearchCV' object has no attribute 'bestscore'"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IWJehKMxVcok",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "e59acd59-5856-484e-8212-b4a99ed811ef"
      },
      "source": [
        "print(gs.best_score_)\n",
        "print(gs.best_params_)\n",
        "#print(gs.cv_results_)"
      ],
      "execution_count": 81,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.5046230148138093\n",
            "{'criterion': 'entropy', 'min_samples_leaf': 10, 'min_samples_split': 2, 'n_estimators': 100}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "x__olnLcFbgD",
        "colab_type": "text"
      },
      "source": [
        "# **Model Estimation and Evaluation**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YIVbb5cEYiaK",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "00b4edb3-0722-4f65-d7a4-e65667eb9901"
      },
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "# rf=RandomForestClassifier(criterion='entropy',\n",
        "#                           n_estimators=75,\n",
        "#                           min_samples_split=2,\n",
        "#                           min_samples_leaf=10,\n",
        "#                           max_features='auto',\n",
        "#                           oob_score=True,\n",
        "#                           random_state=1,\n",
        "#                           n_jobs=-1)\n",
        "rf=RandomForestClassifier(criterion='gini',\n",
        "                          n_estimators=700,\n",
        "                          min_samples_split=10,\n",
        "                          min_samples_leaf=1,\n",
        "                          max_features='auto',\n",
        "                          oob_score=True,\n",
        "                          random_state=1,\n",
        "                          n_jobs=-1)\n",
        "rf.fit(train_x, train_y)\n",
        "print(\"%.4f\" % rf.oob_score_)"
      ],
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "0.9405\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nWrw6gSAHJCZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 762
        },
        "outputId": "62159a58-2ea9-4259-9819-0a83d7e6f70e"
      },
      "source": [
        "pd.concat((pd.DataFrame(train_x.columns, columns = ['variable']), \n",
        "           pd.DataFrame(rf.feature_importances_, columns = ['importance'])), \n",
        "          axis = 1).sort_values(by='importance', ascending = False)[:23]"
      ],
      "execution_count": 158,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>variable</th>\n",
              "      <th>importance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>duration</td>\n",
              "      <td>0.208245</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>fare</td>\n",
              "      <td>0.168719</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>20</th>\n",
              "      <td>distance</td>\n",
              "      <td>0.126819</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>meter_waiting</td>\n",
              "      <td>0.086462</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>meter_waiting_fare</td>\n",
              "      <td>0.061849</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>meter_waiting_till_pickup</td>\n",
              "      <td>0.052004</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>additional_fare</td>\n",
              "      <td>0.048974</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>21</th>\n",
              "      <td>pickup_timestamp</td>\n",
              "      <td>0.040292</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>22</th>\n",
              "      <td>drop_timestamp</td>\n",
              "      <td>0.040056</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>pickup_minute</td>\n",
              "      <td>0.027807</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>drop_minute</td>\n",
              "      <td>0.027650</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>drop_hour</td>\n",
              "      <td>0.020677</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>pickup_hour</td>\n",
              "      <td>0.020522</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>pickup_day</td>\n",
              "      <td>0.016312</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>drop_day</td>\n",
              "      <td>0.016006</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>pickup_weekofyear</td>\n",
              "      <td>0.008062</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>pickup_weekday</td>\n",
              "      <td>0.007943</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>18</th>\n",
              "      <td>drop_weekday</td>\n",
              "      <td>0.007927</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>19</th>\n",
              "      <td>drop_weekofyear</td>\n",
              "      <td>0.007746</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>drop_month</td>\n",
              "      <td>0.002112</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>pickup_month</td>\n",
              "      <td>0.001960</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>drop_year</td>\n",
              "      <td>0.000940</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>pickup_year</td>\n",
              "      <td>0.000914</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                     variable  importance\n",
              "1                    duration    0.208245\n",
              "5                        fare    0.168719\n",
              "20                   distance    0.126819\n",
              "2               meter_waiting    0.086462\n",
              "3          meter_waiting_fare    0.061849\n",
              "4   meter_waiting_till_pickup    0.052004\n",
              "0             additional_fare    0.048974\n",
              "21           pickup_timestamp    0.040292\n",
              "22             drop_timestamp    0.040056\n",
              "10              pickup_minute    0.027807\n",
              "17                drop_minute    0.027650\n",
              "16                  drop_hour    0.020677\n",
              "9                 pickup_hour    0.020522\n",
              "8                  pickup_day    0.016312\n",
              "15                   drop_day    0.016006\n",
              "12          pickup_weekofyear    0.008062\n",
              "11             pickup_weekday    0.007943\n",
              "18               drop_weekday    0.007927\n",
              "19            drop_weekofyear    0.007746\n",
              "14                 drop_month    0.002112\n",
              "7                pickup_month    0.001960\n",
              "13                  drop_year    0.000940\n",
              "6                 pickup_year    0.000914"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 158
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ivZj7ANCViqz",
        "colab_type": "text"
      },
      "source": [
        "# **Model Evaluation using Test set**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f8BewghzVnEW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import jaccard_similarity_score\n",
        "from sklearn.metrics import f1_score\n",
        "from sklearn.metrics import log_loss"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uhg3uAsSeRCz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "testScores={}"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OXGjvga5XsBt",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "3e0bbbd5-5e68-4fba-a905-44b2d7c4f71e"
      },
      "source": [
        "testScores"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EbZvVJEXYvyA",
        "colab_type": "text"
      },
      "source": [
        "# **Prediction**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jY7oGS2cuSmQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 309
        },
        "outputId": "49d175ce-142c-449e-9733-b255e1e27241"
      },
      "source": [
        "df_test = pd.read_csv('/content/drive/My Drive/Colab Notebooks/Fare-Classification/Data/test.csv')\n",
        "features_test = ['additional_fare','duration', 'meter_waiting', 'meter_waiting_fare', 'meter_waiting_till_pickup', 'fare']\n",
        "df_test.head()"
      ],
      "execution_count": 159,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>tripid</th>\n",
              "      <th>additional_fare</th>\n",
              "      <th>duration</th>\n",
              "      <th>meter_waiting</th>\n",
              "      <th>meter_waiting_fare</th>\n",
              "      <th>meter_waiting_till_pickup</th>\n",
              "      <th>pickup_time</th>\n",
              "      <th>drop_time</th>\n",
              "      <th>pick_lat</th>\n",
              "      <th>pick_lon</th>\n",
              "      <th>drop_lat</th>\n",
              "      <th>drop_lon</th>\n",
              "      <th>fare</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>213284604</td>\n",
              "      <td>10.5</td>\n",
              "      <td>924</td>\n",
              "      <td>42</td>\n",
              "      <td>2.4486</td>\n",
              "      <td>148</td>\n",
              "      <td>2/1/2020 0:38</td>\n",
              "      <td>2/1/2020 0:53</td>\n",
              "      <td>6.83454</td>\n",
              "      <td>79.8750</td>\n",
              "      <td>6.77490</td>\n",
              "      <td>79.8840</td>\n",
              "      <td>289.27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>213286352</td>\n",
              "      <td>10.5</td>\n",
              "      <td>4249</td>\n",
              "      <td>20</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>91</td>\n",
              "      <td>2/1/2020 1:02</td>\n",
              "      <td>2/1/2020 2:13</td>\n",
              "      <td>6.91168</td>\n",
              "      <td>79.8723</td>\n",
              "      <td>6.55091</td>\n",
              "      <td>79.9706</td>\n",
              "      <td>1912.70</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>213293973</td>\n",
              "      <td>10.5</td>\n",
              "      <td>1552</td>\n",
              "      <td>255</td>\n",
              "      <td>2.6588</td>\n",
              "      <td>23</td>\n",
              "      <td>2/1/2020 5:02</td>\n",
              "      <td>2/1/2020 5:28</td>\n",
              "      <td>6.92145</td>\n",
              "      <td>79.8478</td>\n",
              "      <td>6.90539</td>\n",
              "      <td>79.8989</td>\n",
              "      <td>394.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>213294622</td>\n",
              "      <td>10.5</td>\n",
              "      <td>462</td>\n",
              "      <td>16</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>198</td>\n",
              "      <td>2/1/2020 5:30</td>\n",
              "      <td>2/1/2020 5:38</td>\n",
              "      <td>6.77433</td>\n",
              "      <td>79.9416</td>\n",
              "      <td>6.80401</td>\n",
              "      <td>79.9407</td>\n",
              "      <td>154.32</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>213298687</td>\n",
              "      <td>10.5</td>\n",
              "      <td>814</td>\n",
              "      <td>392</td>\n",
              "      <td>12.3692</td>\n",
              "      <td>69</td>\n",
              "      <td>2/1/2020 7:00</td>\n",
              "      <td>2/1/2020 7:14</td>\n",
              "      <td>6.97968</td>\n",
              "      <td>79.9130</td>\n",
              "      <td>6.98875</td>\n",
              "      <td>79.8914</td>\n",
              "      <td>147.47</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      tripid  additional_fare  duration  ...  drop_lat  drop_lon     fare\n",
              "0  213284604             10.5       924  ...   6.77490   79.8840   289.27\n",
              "1  213286352             10.5      4249  ...   6.55091   79.9706  1912.70\n",
              "2  213293973             10.5      1552  ...   6.90539   79.8989   394.00\n",
              "3  213294622             10.5       462  ...   6.80401   79.9407   154.32\n",
              "4  213298687             10.5       814  ...   6.98875   79.8914   147.47\n",
              "\n",
              "[5 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 159
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hj7qFjhOuqR0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        },
        "outputId": "69f5ae3a-e507-43c0-ca19-22d4f31dde32"
      },
      "source": [
        "test_x = preProcess(df_test, features_test)\n",
        "X_pred = getX(features_for_test)\n",
        "test_x.head()"
      ],
      "execution_count": 160,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>additional_fare</th>\n",
              "      <th>duration</th>\n",
              "      <th>meter_waiting</th>\n",
              "      <th>meter_waiting_fare</th>\n",
              "      <th>meter_waiting_till_pickup</th>\n",
              "      <th>fare</th>\n",
              "      <th>pickup_year</th>\n",
              "      <th>pickup_month</th>\n",
              "      <th>pickup_day</th>\n",
              "      <th>pickup_hour</th>\n",
              "      <th>pickup_minute</th>\n",
              "      <th>pickup_weekday</th>\n",
              "      <th>pickup_weekofyear</th>\n",
              "      <th>drop_year</th>\n",
              "      <th>drop_month</th>\n",
              "      <th>drop_day</th>\n",
              "      <th>drop_hour</th>\n",
              "      <th>drop_minute</th>\n",
              "      <th>drop_weekday</th>\n",
              "      <th>drop_weekofyear</th>\n",
              "      <th>distance</th>\n",
              "      <th>pickup_timestamp</th>\n",
              "      <th>drop_timestamp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>10.5</td>\n",
              "      <td>924.0</td>\n",
              "      <td>42.0</td>\n",
              "      <td>2.4486</td>\n",
              "      <td>148.0</td>\n",
              "      <td>289.27</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>53.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>6.67</td>\n",
              "      <td>1.580517e+09</td>\n",
              "      <td>1.580518e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>10.5</td>\n",
              "      <td>4249.0</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>91.0</td>\n",
              "      <td>1912.70</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>13.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>41.35</td>\n",
              "      <td>1.580519e+09</td>\n",
              "      <td>1.580523e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>10.5</td>\n",
              "      <td>1552.0</td>\n",
              "      <td>255.0</td>\n",
              "      <td>2.6588</td>\n",
              "      <td>23.0</td>\n",
              "      <td>394.00</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>28.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.92</td>\n",
              "      <td>1.580533e+09</td>\n",
              "      <td>1.580535e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>10.5</td>\n",
              "      <td>462.0</td>\n",
              "      <td>16.0</td>\n",
              "      <td>0.0000</td>\n",
              "      <td>198.0</td>\n",
              "      <td>154.32</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>30.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>38.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>3.28</td>\n",
              "      <td>1.580535e+09</td>\n",
              "      <td>1.580535e+09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>10.5</td>\n",
              "      <td>814.0</td>\n",
              "      <td>392.0</td>\n",
              "      <td>12.3692</td>\n",
              "      <td>69.0</td>\n",
              "      <td>147.47</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2020.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>7.0</td>\n",
              "      <td>14.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>5.0</td>\n",
              "      <td>2.59</td>\n",
              "      <td>1.580540e+09</td>\n",
              "      <td>1.580541e+09</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   additional_fare  duration  ...  pickup_timestamp  drop_timestamp\n",
              "0             10.5     924.0  ...      1.580517e+09    1.580518e+09\n",
              "1             10.5    4249.0  ...      1.580519e+09    1.580523e+09\n",
              "2             10.5    1552.0  ...      1.580533e+09    1.580535e+09\n",
              "3             10.5     462.0  ...      1.580535e+09    1.580535e+09\n",
              "4             10.5     814.0  ...      1.580540e+09    1.580541e+09\n",
              "\n",
              "[5 rows x 23 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 160
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hRA_lJ4stP7H",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_pred_for_test = rf.predict(test_x)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Snc07kO8tWte",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "7ada0422-6847-488f-f21b-0e7f75088c30"
      },
      "source": [
        "y_pred_for_test"
      ],
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([1, 1, 1, ..., 1, 1, 1])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 104
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W0nPboEIYvsI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 102
        },
        "outputId": "feed696e-26ac-4bab-8f85-7039e3f475a6"
      },
      "source": [
        "import numpy as np\n",
        "x = np.array(y_pred_for_test)\n",
        "print(\"Original array:\")\n",
        "print(x)\n",
        "unique_elements, counts_elements = np.unique(x, return_counts=True)\n",
        "print(\"Frequency of unique values of the array:\")\n",
        "print(np.asarray((unique_elements, counts_elements)))"
      ],
      "execution_count": 162,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Original array:\n",
            "[1 1 1 ... 1 1 1]\n",
            "Frequency of unique values of the array:\n",
            "[[   0    1]\n",
            " [ 219 8357]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iKh4gVWDu6a_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import csv\n",
        "\n",
        "with open('/content/drive/My Drive/Colab Notebooks/Fare-Classification/Data/RF.csv', mode='w') as submission_file:\n",
        "    submission_writer = csv.writer(submission_file, delimiter=',', quotechar='\"', quoting=csv.QUOTE_MINIMAL)\n",
        "    submission_writer.writerow(['tripid','prediction'])\n",
        "    for i in range(len(y_pred_for_test)):\n",
        "      submission_writer.writerow([df_test['tripid'][i],y_pred_for_test[i]])"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}