XGBoost_Regression_Model.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "v0tQEpW4DcfK",
        "outputId": "aa4fa55e-9a61-4375-e298-2747ddac5740"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Requirement already satisfied: ipython-autotime in /usr/local/lib/python3.7/dist-packages (0.3.1)\n",
            "Requirement already satisfied: ipython in /usr/local/lib/python3.7/dist-packages (from ipython-autotime) (7.9.0)\n",
            "Requirement already satisfied: pexpect in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (4.8.0)\n",
            "Requirement already satisfied: pygments in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (2.6.1)\n",
            "Requirement already satisfied: prompt-toolkit<2.1.0,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (2.0.10)\n",
            "Requirement already satisfied: backcall in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (0.2.0)\n",
            "Requirement already satisfied: pickleshare in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (0.7.5)\n",
            "Requirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (5.1.1)\n",
            "Requirement already satisfied: setuptools>=18.5 in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (57.4.0)\n",
            "Requirement already satisfied: jedi>=0.10 in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (0.18.2)\n",
            "Requirement already satisfied: decorator in /usr/local/lib/python3.7/dist-packages (from ipython->ipython-autotime) (4.4.2)\n",
            "Requirement already satisfied: parso<0.9.0,>=0.8.0 in /usr/local/lib/python3.7/dist-packages (from jedi>=0.10->ipython->ipython-autotime) (0.8.3)\n",
            "Requirement already satisfied: wcwidth in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.1.0,>=2.0.0->ipython->ipython-autotime) (0.2.5)\n",
            "Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.1.0,>=2.0.0->ipython->ipython-autotime) (1.15.0)\n",
            "Requirement already satisfied: ptyprocess>=0.5 in /usr/local/lib/python3.7/dist-packages (from pexpect->ipython->ipython-autotime) (0.7.0)\n",
            "The autotime extension is already loaded. To reload it, use:\n",
            "  %reload_ext autotime\n",
            "Mounted at /content/drive\n",
            "time: 5.02 s (started: 2022-11-23 10:18:25 +00:00)\n"
          ]
        }
      ],
      "source": [
        "!pip install ipython-autotime\n",
        "%load_ext autotime\n",
        "import pandas as pd\n",
        "from google.colab import drive\n",
        "import pathlib\n",
        "import numpy as np\n",
        "import gc\n",
        "import matplotlib.pyplot as plt\n",
        "from xgboost import XGBRegressor\n",
        "import pyarrow.parquet as pq\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import mean_squared_error\n",
        "from sklearn.metrics import mean_absolute_error\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from xgboost import plot_importance\n",
        "\n",
        "\n",
        "drive.mount('/content/drive',force_remount=True)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "random_state=17"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IGgAeOxfY4lh",
        "outputId": "49d2e06f-b35a-4a9c-aeba-0bfa0722c3fe"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 729 µs (started: 2022-11-23 06:56:44 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k9XRTmWtOVsC",
        "outputId": "da65a3f6-a6ab-43bf-c4e2-01d18ae61156"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 892 µs (started: 2022-11-23 06:56:44 +00:00)\n"
          ]
        }
      ],
      "source": [
        "XGBRegressorModel = XGBRegressor(max_depth=6, learning_rate=0.09, iteration=250, random_state=random_state)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "taxiData = pd.DataFrame()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3d_anN31U2tJ",
        "outputId": "d657d49a-0692-4dc1-9c12-6e8d7d3f3a4c"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 2.49 ms (started: 2022-11-23 06:56:44 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "7f6byBVKPoFG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b5c9e25f-887d-467f-9ef6-16f10c1e15c3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 174 ms (started: 2022-11-23 06:56:44 +00:00)\n"
          ]
        }
      ],
      "source": [
        "data_dir = \"/content/drive/MyDrive/yellow_tripdata_2020_data.parquet\"\n",
        "data_dir = pathlib.Path(data_dir)\n",
        "yellowTaxiData2020 = pq.ParquetFile(data_dir)\n",
        "del data_dir"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "mOhmoFQpZR7Z",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7b5a50d3-cb07-42ef-ea0e-99f25dd3bdb3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 2.92 s (started: 2022-11-23 06:56:45 +00:00)\n"
          ]
        }
      ],
      "source": [
        "for taxiDataBatch in yellowTaxiData2020.iter_batches(batch_size=2000000):\n",
        "    taxiDataBatchDF = taxiDataBatch.to_pandas()\n",
        "    taxiDataBatchDF.drop(columns=['tpep_pickup_datetime', 'tpep_dropoff_datetime', 'pickup_date', 'pickup_time', 'dropoff_date', 'dropoff_time'], inplace=True)\n",
        "    taxiData = pd.concat([taxiData, taxiDataBatchDF])\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "taxiData.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1g-iQNiFVsM2",
        "outputId": "52784c9b-2d11-4c6c-96ff-afc0f048dbd0"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(4069192, 18)"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 7.93 ms (started: 2022-11-23 06:56:48 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "878B9VjddAhl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "77a11ab9-335b-4a73-b088-e7221de33bb4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 553 µs (started: 2022-11-23 06:56:48 +00:00)\n"
          ]
        }
      ],
      "source": [
        "del yellowTaxiData2020"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "oiCyjCAtu_-3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a181f7d2-e4d1-4099-f693-b6a3c1f85138"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "66"
            ]
          },
          "metadata": {},
          "execution_count": 9
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 77.1 ms (started: 2022-11-23 06:56:48 +00:00)\n"
          ]
        }
      ],
      "source": [
        "gc.collect()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "TmC2RQvAc0-H",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "9f5b87b8-11d0-42c2-c7e9-d877cde632d9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 162 ms (started: 2022-11-23 06:56:49 +00:00)\n"
          ]
        }
      ],
      "source": [
        "data_dir = \"/content/drive/MyDrive/yellow_tripdata_2021_data.parquet\"\n",
        "data_dir = pathlib.Path(data_dir)\n",
        "yellowTaxiData2021 = pq.ParquetFile(data_dir)\n",
        "del data_dir"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for taxiDataBatch in yellowTaxiData2021.iter_batches(batch_size=2000000):\n",
        "    taxiDataBatchDF = taxiDataBatch.to_pandas()\n",
        "    taxiDataBatchDF.drop(columns=['tpep_pickup_datetime', 'tpep_dropoff_datetime', 'pickup_date', 'pickup_time', 'dropoff_date', 'dropoff_time'], inplace=True)\n",
        "    taxiData = pd.concat([taxiData, taxiDataBatchDF])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JrDD87TCVLPc",
        "outputId": "744b9383-f825-4afd-b9e1-6da29bba20d8"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 16.5 s (started: 2022-11-23 06:56:51 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "scaler = MinMaxScaler()\n",
        "taxiData = pd.DataFrame(scaler.fit_transform(taxiData.values), columns=taxiData.columns, index=taxiData.index)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VBl86d05cDQr",
        "outputId": "6e868c58-9415-4861-b779-de91d61eceb2"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 7.91 s (started: 2022-11-23 06:57:08 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(taxiData.loc[:, (taxiData.columns!='ride_duration')], \n",
        "                                                    taxiData.loc[:, 'ride_duration'], \n",
        "                                                    test_size=0.4, \n",
        "                                                    random_state=random_state)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6bo__xA3b3iH",
        "outputId": "b26d15a2-550c-4d1a-d4eb-949465add264"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 9.93 s (started: 2022-11-23 06:57:16 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "del yellowTaxiData2021"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dpvTdg4GX3xW",
        "outputId": "c66173ea-31af-4470-b466-4ac1fd016300"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 635 µs (started: 2022-11-23 06:57:25 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "del taxiData"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GcdrUhbac131",
        "outputId": "0f20478c-d338-4d91-d30c-7d08d7741901"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 634 µs (started: 2022-11-23 06:57:25 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "gc.collect()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_xvDBME2X5sJ",
        "outputId": "e6953a47-6b04-455f-8ae5-52468a980745"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "175"
            ]
          },
          "metadata": {},
          "execution_count": 16
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 57.2 ms (started: 2022-11-23 06:57:25 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "del taxiDataBatchDF"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9aNZe-o2Zjdj",
        "outputId": "90af8363-27fb-4b28-fac7-52ea49d14ecb"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 446 µs (started: 2022-11-23 06:57:26 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "gc.collect()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UQhQPyUYZ3hQ",
        "outputId": "82479823-7b89-4474-aa32-a70f6f7c87dc"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "44"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 54.8 ms (started: 2022-11-23 06:57:26 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "XGBRegressorModel.fit(X_train, y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8E06IzXwX8kQ",
        "outputId": "2175f069-8c51-4867-8ad9-12a97bfef8a5"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[06:57:35] WARNING: /workspace/src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n",
            "[06:57:35] WARNING: /workspace/src/learner.cc:686: Tree method is automatically selected to be 'approx' for faster speed. To use old behavior (exact greedy algorithm on single machine), set tree_method to 'exact'.\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "XGBRegressor(iteration=250, learning_rate=0.09, max_depth=6, random_state=17)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 2h 24min 12s (started: 2022-11-23 06:57:29 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions = XGBRegressorModel.predict(X_test)"
      ],
      "metadata": {
        "id": "Oudr-pRMYL7f",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b357ad5d-f61e-4e03-a1b9-bb88d330992d"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 59.7 s (started: 2022-11-23 09:21:53 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions_train = XGBRegressorModel.predict(X_train)"
      ],
      "metadata": {
        "id": "pv4PQ3kG0cD3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "61f54d64-3115-45dc-8ec9-21143792d2b1"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 1min 29s (started: 2022-11-23 09:22:58 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "rmse_train = mean_squared_error(y_train, predictions_train)**0.5\n",
        "print(f'The RMSE of training prediction is: {rmse_train}')"
      ],
      "metadata": {
        "id": "7dl_O-lD0fA3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4131b334-a6e1-4074-c14b-ce8dcc4f3d1c"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The RMSE of training prediction is: 0.10741046544906936\n",
            "time: 191 ms (started: 2022-11-23 09:25:20 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mse_train = mean_squared_error(y_train, predictions_train)\n",
        "print(f'The MSE of training prediction is: {mse_train}')"
      ],
      "metadata": {
        "id": "_1J4IXfj0sHB",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "06c03ff7-6db3-4a8e-86ef-47a9863f7a73"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The MSE of training prediction is: 0.011537008087985723\n",
            "time: 73.3 ms (started: 2022-11-23 09:25:29 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mae_train = mean_absolute_error(y_train, predictions_train)\n",
        "print(f'The MAE of training prediction is: {mae_train}')"
      ],
      "metadata": {
        "id": "EjgCEXz_0uVF",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b0c59396-d9ea-4e69-a7e6-2906eafa02de"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The MAE of training prediction is: 0.08050702934962413\n",
            "time: 165 ms (started: 2022-11-23 09:25:31 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "rmse_test = mean_squared_error(y_test, predictions)**0.5\n",
        "print(f'The RMSE of prediction is: {rmse_test}')"
      ],
      "metadata": {
        "id": "B5AhKmCg0wsp",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1f889f11-9bd9-42ff-eb47-c53711f324ec"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The RMSE of prediction is: 0.10739925367263024\n",
            "time: 49.5 ms (started: 2022-11-23 09:25:32 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mse_test = mean_squared_error(y_test, predictions)\n",
        "print(f'The MSE of prediction is: {mse_test}')"
      ],
      "metadata": {
        "id": "cWs8nX6l006M",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "31e18b9b-53a6-4d2b-9fb3-b68da8ef44aa"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The MSE of prediction is: 0.01153459968943798\n",
            "time: 50.2 ms (started: 2022-11-23 09:25:35 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "mae_test = mean_absolute_error(y_test, predictions)\n",
        "print(f'The MAE of training prediction is: {mae_test}')"
      ],
      "metadata": {
        "id": "iCoqqwOg02-d",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e339917d-c00c-48f4-d868-832181c9c9da"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The MAE of training prediction is: 0.08048598340565506\n",
            "time: 55.9 ms (started: 2022-11-23 09:25:37 +00:00)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_importance(XGBRegressorModel)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 331
        },
        "id": "bdz4mt6d_wwo",
        "outputId": "5ff15804-1ed7-4e9c-8bd3-7ccd6a43480b"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fbce825fed0>"
            ]
          },
          "metadata": {},
          "execution_count": 32
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "time: 246 ms (started: 2022-11-23 10:19:28 +00:00)\n"
          ]
        }
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "machine_shape": "hm"
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}