gluonts-forecasting-notebooks-1/Gluonts_twitter_volume_forecasting.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Gluonts twitter volume forecasting.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPnECKy9/9x3uizOlZK+rEl",
      "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/mmaithani/data-science/blob/main/Gluonts_twitter_volume_forecasting.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "E3_PpfKsEYMS",
        "outputId": "7c7222bc-100e-44bf-fc6a-774afa77d2e6"
      },
      "source": [
        "pip install gluonts"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting gluonts\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/60/b4/4893bddfebbab963ab6406e2c33a241359463b25549a8a9937f87aa7aba9/gluonts-0.6.4-py3-none-any.whl (569kB)\n",
            "\u001b[K     |████████████████████████████████| 573kB 5.7MB/s \n",
            "\u001b[?25hRequirement already satisfied: matplotlib~=3.0 in /usr/local/lib/python3.6/dist-packages (from gluonts) (3.2.2)\n",
            "Requirement already satisfied: numpy~=1.16 in /usr/local/lib/python3.6/dist-packages (from gluonts) (1.19.4)\n",
            "Collecting ujson~=1.35\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/16/c4/79f3409bc710559015464e5f49b9879430d8f87498ecdc335899732e5377/ujson-1.35.tar.gz (192kB)\n",
            "\u001b[K     |████████████████████████████████| 194kB 19.0MB/s \n",
            "\u001b[?25hCollecting pydantic<1.7,~=1.1\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/aa/5f/855412ad12817ae87f1c77d3af2fc384eaed3adfb8f3994816d75483fa20/pydantic-1.6.1-cp36-cp36m-manylinux2014_x86_64.whl (8.7MB)\n",
            "\u001b[K     |████████████████████████████████| 8.7MB 16.2MB/s \n",
            "\u001b[?25hRequirement already satisfied: toolz~=0.10 in /usr/local/lib/python3.6/dist-packages (from gluonts) (0.11.1)\n",
            "Requirement already satisfied: tqdm~=4.23 in /usr/local/lib/python3.6/dist-packages (from gluonts) (4.41.1)\n",
            "Requirement already satisfied: holidays>=0.9 in /usr/local/lib/python3.6/dist-packages (from gluonts) (0.10.4)\n",
            "Requirement already satisfied: pandas>=1.0 in /usr/local/lib/python3.6/dist-packages (from gluonts) (1.1.5)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.6/dist-packages (from matplotlib~=3.0->gluonts) (0.10.0)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib~=3.0->gluonts) (2.4.7)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib~=3.0->gluonts) (2.8.1)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.6/dist-packages (from matplotlib~=3.0->gluonts) (1.3.1)\n",
            "Requirement already satisfied: dataclasses>=0.6; python_version < \"3.7\" in /usr/local/lib/python3.6/dist-packages (from pydantic<1.7,~=1.1->gluonts) (0.8)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.6/dist-packages (from holidays>=0.9->gluonts) (1.15.0)\n",
            "Requirement already satisfied: convertdate<=2.2.0 in /usr/local/lib/python3.6/dist-packages (from holidays>=0.9->gluonts) (2.2.0)\n",
            "Requirement already satisfied: korean-lunar-calendar in /usr/local/lib/python3.6/dist-packages (from holidays>=0.9->gluonts) (0.2.1)\n",
            "Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.6/dist-packages (from pandas>=1.0->gluonts) (2018.9)\n",
            "Requirement already satisfied: pymeeus<=1,>=0.3.6 in /usr/local/lib/python3.6/dist-packages (from convertdate<=2.2.0->holidays>=0.9->gluonts) (0.3.7)\n",
            "Building wheels for collected packages: ujson\n",
            "  Building wheel for ujson (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for ujson: filename=ujson-1.35-cp36-cp36m-linux_x86_64.whl size=68034 sha256=170e662944b3fefb82cb255a0c01d6cf53ce10bcceece9f3319f2213e06fe7e9\n",
            "  Stored in directory: /root/.cache/pip/wheels/28/77/e4/0311145b9c2e2f01470e744855131f9e34d6919687550f87d1\n",
            "Successfully built ujson\n",
            "Installing collected packages: ujson, pydantic, gluonts\n",
            "Successfully installed gluonts-0.6.4 pydantic-1.6.1 ujson-1.35\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9IZeKRpXEndu",
        "outputId": "f0912c8e-bd4a-4d25-c620-4e4890c6c870"
      },
      "source": [
        "pip install mxnet"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting mxnet\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/29/bb/54cbabe428351c06d10903c658878d29ee7026efbe45133fd133598d6eb6/mxnet-1.7.0.post1-py2.py3-none-manylinux2014_x86_64.whl (55.0MB)\n",
            "\u001b[K     |████████████████████████████████| 55.0MB 76kB/s \n",
            "\u001b[?25hCollecting graphviz<0.9.0,>=0.8.1\n",
            "  Downloading https://files.pythonhosted.org/packages/53/39/4ab213673844e0c004bed8a0781a0721a3f6bb23eb8854ee75c236428892/graphviz-0.8.4-py2.py3-none-any.whl\n",
            "Requirement already satisfied: requests<3,>=2.20.0 in /usr/local/lib/python3.6/dist-packages (from mxnet) (2.23.0)\n",
            "Requirement already satisfied: numpy<2.0.0,>1.16.0 in /usr/local/lib/python3.6/dist-packages (from mxnet) (1.19.4)\n",
            "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.20.0->mxnet) (1.24.3)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.20.0->mxnet) (2020.12.5)\n",
            "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.20.0->mxnet) (3.0.4)\n",
            "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.6/dist-packages (from requests<3,>=2.20.0->mxnet) (2.10)\n",
            "Installing collected packages: graphviz, mxnet\n",
            "  Found existing installation: graphviz 0.10.1\n",
            "    Uninstalling graphviz-0.10.1:\n",
            "      Successfully uninstalled graphviz-0.10.1\n",
            "Successfully installed graphviz-0.8.4 mxnet-1.7.0.post1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vVNMTd-BEWHG"
      },
      "source": [
        "from gluonts.dataset import common\r\n",
        "from gluonts.model import deepar\r\n",
        "from gluonts.trainer import Trainer\r\n",
        "\r\n",
        "import pandas as pd"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 235
        },
        "id": "zp2hdhzlFhlg",
        "outputId": "49cc7e56-9c58-4798-9e5a-d71c1c56b05a"
      },
      "source": [
        "url = \"https://raw.githubusercontent.com/numenta/NAB/master/data/realTweets/Twitter_volume_AMZN.csv\"\r\n",
        "df = pd.read_csv(url, header=0, index_col=0)\r\n",
        "df.head()"
      ],
      "execution_count": 7,
      "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>value</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>timestamp</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2015-02-26 21:42:53</th>\n",
              "      <td>57</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015-02-26 21:47:53</th>\n",
              "      <td>43</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015-02-26 21:52:53</th>\n",
              "      <td>55</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015-02-26 21:57:53</th>\n",
              "      <td>64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2015-02-26 22:02:53</th>\n",
              "      <td>93</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                     value\n",
              "timestamp                 \n",
              "2015-02-26 21:42:53     57\n",
              "2015-02-26 21:47:53     43\n",
              "2015-02-26 21:52:53     55\n",
              "2015-02-26 21:57:53     64\n",
              "2015-02-26 22:02:53     93"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AAMItQe-FmmX"
      },
      "source": [
        "data = common.ListDataset([{\r\n",
        "    \"start\": df.index[0],\r\n",
        "    \"target\": df.value[:\"2015-04-05 00:00:00\"]\r\n",
        "}],\r\n",
        "                          freq=\"5min\")\r\n"
      ],
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gqxl1PoDFpjh",
        "outputId": "fddc6dcf-d859-48a2-a561-cc3f546a6685"
      },
      "source": [
        "trainer = Trainer(epochs=10)\r\n",
        "estimator = deepar.DeepAREstimator(\r\n",
        "    freq=\"5min\", prediction_length=12, trainer=trainer)\r\n",
        "predictor = estimator.train(training_data=data)\r\n",
        "\r\n",
        "prediction = next(predictor.predict(data))"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\r  0%|          | 0/50 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "learning rate from ``lr_scheduler`` has been overwritten by ``learning_rate`` in optimizer.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 50/50 [00:03<00:00, 15.25it/s, epoch=1/10, avg_epoch_loss=4.49]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.75it/s, epoch=2/10, avg_epoch_loss=4.09]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.64it/s, epoch=3/10, avg_epoch_loss=4.04]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.88it/s, epoch=4/10, avg_epoch_loss=4.02]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.62it/s, epoch=5/10, avg_epoch_loss=3.98]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.42it/s, epoch=6/10, avg_epoch_loss=3.97]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.48it/s, epoch=7/10, avg_epoch_loss=3.95]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.72it/s, epoch=8/10, avg_epoch_loss=3.96]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.62it/s, epoch=9/10, avg_epoch_loss=3.97]\n",
            "100%|██████████| 50/50 [00:03<00:00, 15.68it/s, epoch=10/10, avg_epoch_loss=3.93]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[48.119385 45.482513 43.079456 40.907524 41.094902 38.321095 38.837597\n",
            " 38.26018  39.68032  40.427383 42.762894 41.39221 ]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0ig3yLnhF1l3",
        "outputId": "b593b65c-c469-4f61-e684-b31ee5a4d660"
      },
      "source": [
        "print(prediction.mean)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[48.119385 45.482513 43.079456 40.907524 41.094902 38.321095 38.837597\n",
            " 38.26018  39.68032  40.427383 42.762894 41.39221 ]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 277
        },
        "id": "E27gbMFrEh0E",
        "outputId": "4bdbba3c-31e8-4d35-c0a6-4ff0a17bd7c4"
      },
      "source": [
        "prediction.plot(output_file='graph.png')"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "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": "OBEvWiyjFRgK"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}