News_sentiment_Glove+BiLSTM.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "News sentiment Glove+BiLSTM.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "FEQ285jNg_RK",
        "outputId": "d6a8d6a6-77dc-4ccf-89aa-1de2be3669be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive',force_remount=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fCSZglOfOYyg",
        "outputId": "5fc68629-1bb8-46d1-ab15-4bd1728d6d5b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "!ls drive/My\\ Drive/ML-work/NLP/predict-news-sentiment/"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "data  glove.6B.50d.txt\tglove.6B.50d.txt.zip\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h1UgWh2AOZOl",
        "outputId": "2ea7f7f3-e98f-40a6-e67b-38e0658cb706",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "%cd drive/My\\ Drive/ML-work/NLP/predict-news-sentiment/"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/ML-work/NLP/predict-news-sentiment\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6qTFqp7Fg2ah"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report, precision_recall_curve\n",
        "\n",
        "import re\n",
        "import nltk\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.stem import WordNetLemmatizer #word stemmer class\n",
        "lemma = WordNetLemmatizer()\n",
        "\n",
        "from keras.preprocessing.text import Tokenizer\n",
        "from keras.preprocessing.sequence import pad_sequences\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Bidirectional, LSTM, Dropout, BatchNormalization\n",
        "from keras.layers.embeddings import Embedding"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cBT-I6dZltSv",
        "outputId": "a40f99da-7d01-44c5-fb65-94061f3c929a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 67
        }
      },
      "source": [
        "nltk.download('stopwords')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J4pIFy51g2fQ"
      },
      "source": [
        "!unzip -q \"./glove.6B.50d.txt.zip\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9i-UmNf7g2kv"
      },
      "source": [
        "embeddings_index = dict()\n",
        "f = open('./glove.6B.50d.txt')\n",
        "for line in f:\n",
        "    values = line.split()\n",
        "    word = values[0]\n",
        "    coefs = np.asarray(values[1:], dtype='float32')\n",
        "    embeddings_index[word] = coefs\n",
        "f.close()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ckHTeGNbg2pS"
      },
      "source": [
        "train = pd.read_csv('./data/train.csv')\n",
        "test = pd.read_csv('./data/test.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xKiiQQqNg2v1",
        "outputId": "f24323e6-ca27-4186-9025-6b9231e9e7bf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 296
        }
      },
      "source": [
        "train.head()"
      ],
      "execution_count": null,
      "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>IDLink</th>\n",
              "      <th>Title</th>\n",
              "      <th>Headline</th>\n",
              "      <th>Source</th>\n",
              "      <th>Topic</th>\n",
              "      <th>PublishDate</th>\n",
              "      <th>Facebook</th>\n",
              "      <th>GooglePlus</th>\n",
              "      <th>LinkedIn</th>\n",
              "      <th>SentimentTitle</th>\n",
              "      <th>SentimentHeadline</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Tr3CMgRv1N</td>\n",
              "      <td>Obama Lays Wreath at Arlington National Cemetery</td>\n",
              "      <td>Obama Lays Wreath at Arlington National Cemete...</td>\n",
              "      <td>USA TODAY</td>\n",
              "      <td>obama</td>\n",
              "      <td>2002-04-02 00:00:00</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>-0.053300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Wc81vGp8qZ</td>\n",
              "      <td>A Look at the Health of the Chinese Economy</td>\n",
              "      <td>Tim Haywood, investment director business-unit...</td>\n",
              "      <td>Bloomberg</td>\n",
              "      <td>economy</td>\n",
              "      <td>2008-09-20 00:00:00</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>0.208333</td>\n",
              "      <td>-0.156386</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>zNGH03CrZH</td>\n",
              "      <td>Nouriel Roubini: Global Economy Not Back to 2008</td>\n",
              "      <td>Nouriel Roubini, NYU professor and chairman at...</td>\n",
              "      <td>Bloomberg</td>\n",
              "      <td>economy</td>\n",
              "      <td>2012-01-28 00:00:00</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-0.425210</td>\n",
              "      <td>0.139754</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3sM1H0W8ts</td>\n",
              "      <td>Finland GDP Expands In Q4</td>\n",
              "      <td>Finland's economy expanded marginally in the t...</td>\n",
              "      <td>RTT News</td>\n",
              "      <td>economy</td>\n",
              "      <td>2015-03-01 00:06:00</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.026064</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>wUbnxgvqaZ</td>\n",
              "      <td>Tourism, govt spending buoys Thai economy in J...</td>\n",
              "      <td>Tourism and public spending continued to boost...</td>\n",
              "      <td>The Nation - Thailand&amp;#39;s English news</td>\n",
              "      <td>economy</td>\n",
              "      <td>2015-03-01 00:11:00</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>-1</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.141084</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       IDLink  ... SentimentHeadline\n",
              "0  Tr3CMgRv1N  ...         -0.053300\n",
              "1  Wc81vGp8qZ  ...         -0.156386\n",
              "2  zNGH03CrZH  ...          0.139754\n",
              "3  3sM1H0W8ts  ...          0.026064\n",
              "4  wUbnxgvqaZ  ...          0.141084\n",
              "\n",
              "[5 rows x 11 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1Z5XxliIg200",
        "outputId": "89af39bb-9c93-4fbc-af7c-f0838fab3206",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 373
        }
      },
      "source": [
        "missing_val = pd.DataFrame(train.isnull().sum())\n",
        "\n",
        "missing_val = missing_val.reset_index()\n",
        "missing_val"
      ],
      "execution_count": null,
      "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>index</th>\n",
              "      <th>0</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>IDLink</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Title</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Headline</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Source</td>\n",
              "      <td>175</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Topic</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>PublishDate</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>Facebook</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>GooglePlus</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>LinkedIn</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>SentimentTitle</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>SentimentHeadline</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                index    0\n",
              "0              IDLink    0\n",
              "1               Title    0\n",
              "2            Headline    0\n",
              "3              Source  175\n",
              "4               Topic    0\n",
              "5         PublishDate    0\n",
              "6            Facebook    0\n",
              "7          GooglePlus    0\n",
              "8            LinkedIn    0\n",
              "9      SentimentTitle    0\n",
              "10  SentimentHeadline    0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ORqK4w25g24_"
      },
      "source": [
        "train.dropna(inplace=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tMB68BDpOcnd",
        "outputId": "d0d38c4b-27df-4ef2-fca4-a398fbe9a753",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 185
        }
      },
      "source": [
        "test.isnull().sum()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "IDLink           0\n",
              "Title            0\n",
              "Headline         0\n",
              "Source         101\n",
              "Topic            0\n",
              "PublishDate      0\n",
              "Facebook         0\n",
              "GooglePlus       0\n",
              "LinkedIn         0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3mmEfgKXPGOJ",
        "outputId": "ddf881d4-bfaa-4393-de5f-fd31cd20577b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 303
        }
      },
      "source": [
        "test['Source'].fillna(value=test['Source'].mean(), inplace=True)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "TypeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-10-ffd812159a08>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mtest\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Source'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfillna\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalue\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtest\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Source'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minplace\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/pandas/core/generic.py\u001b[0m in \u001b[0;36mstat_func\u001b[0;34m(self, axis, skipna, level, numeric_only, **kwargs)\u001b[0m\n\u001b[1;32m  11467\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_agg_by_level\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlevel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mskipna\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mskipna\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m  11468\u001b[0m         return self._reduce(\n\u001b[0;32m> 11469\u001b[0;31m             \u001b[0mfunc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mskipna\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mskipna\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnumeric_only\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnumeric_only\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m  11470\u001b[0m         )\n\u001b[1;32m  11471\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/pandas/core/series.py\u001b[0m in \u001b[0;36m_reduce\u001b[0;34m(self, op, name, axis, skipna, numeric_only, filter_type, **kwds)\u001b[0m\n\u001b[1;32m   4234\u001b[0m                 )\n\u001b[1;32m   4235\u001b[0m             \u001b[0;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mall\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"ignore\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4236\u001b[0;31m                 \u001b[0;32mreturn\u001b[0m \u001b[0mop\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdelegate\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mskipna\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mskipna\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\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   4237\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4238\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_reindex_indexer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnew_index\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindexer\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/pandas/core/nanops.py\u001b[0m in \u001b[0;36m_f\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m     69\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     70\u001b[0m                 \u001b[0;32mwith\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0merrstate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minvalid\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"ignore\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 71\u001b[0;31m                     \u001b[0;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\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     72\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mValueError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     73\u001b[0m                 \u001b[0;31m# we want to transform an object array\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/pandas/core/nanops.py\u001b[0m in \u001b[0;36mf\u001b[0;34m(values, axis, skipna, **kwds)\u001b[0m\n\u001b[1;32m    127\u001b[0m                     \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0malt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mskipna\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mskipna\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    128\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 129\u001b[0;31m                 \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0malt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mskipna\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mskipna\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\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    130\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    131\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/pandas/core/nanops.py\u001b[0m in \u001b[0;36mnanmean\u001b[0;34m(values, axis, skipna, mask)\u001b[0m\n\u001b[1;32m    561\u001b[0m         \u001b[0mdtype_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    562\u001b[0m     \u001b[0mcount\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_get_counts\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmask\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype_count\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 563\u001b[0;31m     \u001b[0mthe_sum\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_ensure_numeric\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype_sum\u001b[0m\u001b[0;34m)\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    564\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    565\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0maxis\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mthe_sum\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"ndim\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/numpy/core/_methods.py\u001b[0m in \u001b[0;36m_sum\u001b[0;34m(a, axis, dtype, out, keepdims, initial, where)\u001b[0m\n\u001b[1;32m     36\u001b[0m def _sum(a, axis=None, dtype=None, out=None, keepdims=False,\n\u001b[1;32m     37\u001b[0m          initial=_NoValue, where=True):\n\u001b[0;32m---> 38\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mumr_sum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkeepdims\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minitial\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwhere\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     39\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     40\u001b[0m def _prod(a, axis=None, dtype=None, out=None, keepdims=False,\n",
            "\u001b[0;31mTypeError\u001b[0m: must be str, not int"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AL92_ajxPBSg",
        "outputId": "b84ec5d4-906b-4760-f237-87364f6ec8d2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284
        }
      },
      "source": [
        "test.describe()"
      ],
      "execution_count": null,
      "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>Facebook</th>\n",
              "      <th>GooglePlus</th>\n",
              "      <th>LinkedIn</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>37187.000000</td>\n",
              "      <td>37187.000000</td>\n",
              "      <td>37187.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>85.063033</td>\n",
              "      <td>2.903837</td>\n",
              "      <td>19.982843</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>420.610253</td>\n",
              "      <td>13.556345</td>\n",
              "      <td>225.746903</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-1.000000</td>\n",
              "      <td>-1.000000</td>\n",
              "      <td>-1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>4.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>28.000000</td>\n",
              "      <td>2.000000</td>\n",
              "      <td>4.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>16598.000000</td>\n",
              "      <td>1016.000000</td>\n",
              "      <td>20341.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           Facebook    GooglePlus      LinkedIn\n",
              "count  37187.000000  37187.000000  37187.000000\n",
              "mean      85.063033      2.903837     19.982843\n",
              "std      420.610253     13.556345    225.746903\n",
              "min       -1.000000     -1.000000     -1.000000\n",
              "25%        0.000000      0.000000      0.000000\n",
              "50%        4.000000      0.000000      0.000000\n",
              "75%       28.000000      2.000000      4.000000\n",
              "max    16598.000000   1016.000000  20341.000000"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IsdROfSeujSr"
      },
      "source": [
        "X_train_title = train.loc[:,'Title'].values\n",
        "y_train_title = train.loc[:,['SentimentTitle']].values\n",
        "\n",
        "X_train_headline = train.loc[:,'Headline'].values\n",
        "y_train_headline = train.loc[:,['SentimentHeadline']].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MaKdcaDXEj8F"
      },
      "source": [
        "X_test_title = test.loc[:,'Title'].values\n",
        "X_test_headline = test.loc[:,'Headline'].values"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IoFqfxXska_V"
      },
      "source": [
        "title_df=pd.DataFrame()\n",
        "title_df['X_train_title']=X_train_title\n",
        "title_df['y_train_title']=y_train_title\n",
        "\n",
        "headline_df=pd.DataFrame()\n",
        "headline_df['X_train_headline']=X_train_headline\n",
        "headline_df['y_train_headline']=y_train_headline\n",
        "\n",
        "test_df=pd.DataFrame()\n",
        "test_df['X_test_title']=X_test_title\n",
        "test_df['X_test_headline']=X_test_headline"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9JBhpcHmg2-6"
      },
      "source": [
        "def preprocess_text(texts):\n",
        "    texts = texts.lower() \n",
        "    texts = re.sub(r'[^\\x00-\\x7F]+',' ', texts) \n",
        "    words = texts.split()\n",
        "    words = filter(lambda x: x[0]!= '@' , texts.split()) \n",
        "    words = [word for word in words if word not in set(stopwords.words('english'))] \n",
        "    texts = \" \".join(words)\n",
        "    return texts"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RmUCM9Jkg3DZ",
        "outputId": "8356d87b-3815-4262-da6f-89be3a55c3ad",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 195
        }
      },
      "source": [
        "title_df['X_train_title'] = title_df.X_train_title.apply(preprocess_text)\n",
        "display(title_df.head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "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>X_train_title</th>\n",
              "      <th>y_train_title</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>obama lays wreath arlington national cemetery</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>look health chinese economy</td>\n",
              "      <td>0.208333</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>nouriel roubini: global economy back 2008</td>\n",
              "      <td>-0.425210</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>finland gdp expands q4</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>tourism, govt spending buoys thai economy january</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                       X_train_title  y_train_title\n",
              "0      obama lays wreath arlington national cemetery       0.000000\n",
              "1                        look health chinese economy       0.208333\n",
              "2          nouriel roubini: global economy back 2008      -0.425210\n",
              "3                             finland gdp expands q4       0.000000\n",
              "4  tourism, govt spending buoys thai economy january       0.000000"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ykn9lB8mg3If",
        "outputId": "cba81463-8493-4389-aa2b-68373cf8b7c8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 195
        }
      },
      "source": [
        "headline_df['X_train_headline'] = headline_df.X_train_headline.apply(preprocess_text)\n",
        "display(headline_df.head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "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>X_train_headline</th>\n",
              "      <th>y_train_headline</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>obama lays wreath arlington national cemetery....</td>\n",
              "      <td>-0.053300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>tim haywood, investment director business-unit...</td>\n",
              "      <td>-0.156386</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>nouriel roubini, nyu professor chairman roubin...</td>\n",
              "      <td>0.139754</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>finland's economy expanded marginally three mo...</td>\n",
              "      <td>0.026064</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>tourism public spending continued boost econom...</td>\n",
              "      <td>0.141084</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                    X_train_headline  y_train_headline\n",
              "0  obama lays wreath arlington national cemetery....         -0.053300\n",
              "1  tim haywood, investment director business-unit...         -0.156386\n",
              "2  nouriel roubini, nyu professor chairman roubin...          0.139754\n",
              "3  finland's economy expanded marginally three mo...          0.026064\n",
              "4  tourism public spending continued boost econom...          0.141084"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qNFmIG-bg3OG",
        "outputId": "4ba31216-2093-4ffa-951a-cd93082ceb2d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 195
        }
      },
      "source": [
        "test_df['X_test_title'] = test_df.X_test_title.apply(preprocess_text)\n",
        "test_df['X_test_headline'] = test_df.X_test_headline.apply(preprocess_text)\n",
        "display(test_df.head())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "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>X_test_title</th>\n",
              "      <th>X_test_headline</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>sliding economy: fg fights back n3trn tsa funds</td>\n",
              "      <td>2016 budget passed national assembly n3trillio...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>microsoft shows hololens bring distant family ...</td>\n",
              "      <td>recent microsoft research video shows $3000 au...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>microsoft twitter robot praises hitler, trump ...</td>\n",
              "      <td>* microsoft teamed bing create taytweets, acco...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>flood central bank moves can't get world econo...</td>\n",
              "      <td>central bankers managed steer world economy cl...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>usd/jpy: bears lining mixed u.s. economy outlook</td>\n",
              "      <td>however, streak seven-day gains might end mark...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                        X_test_title                                    X_test_headline\n",
              "0    sliding economy: fg fights back n3trn tsa funds  2016 budget passed national assembly n3trillio...\n",
              "1  microsoft shows hololens bring distant family ...  recent microsoft research video shows $3000 au...\n",
              "2  microsoft twitter robot praises hitler, trump ...  * microsoft teamed bing create taytweets, acco...\n",
              "3  flood central bank moves can't get world econo...  central bankers managed steer world economy cl...\n",
              "4   usd/jpy: bears lining mixed u.s. economy outlook  however, streak seven-day gains might end mark..."
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "twl2J8mig3UB"
      },
      "source": [
        "max_len_title = title_df.X_train_title.apply(lambda x: len(x.split())).max()\n",
        "\n",
        "tok_title = Tokenizer()\n",
        "tok_title.fit_on_texts(title_df.X_train_title)\n",
        "vocab_size_title = len(tok_title.word_index) + 1\n",
        "encoded_title = tok_title.texts_to_sequences(title_df.X_train_title)\n",
        "padded_title = pad_sequences(encoded_title, maxlen=max_len_title, padding='post')\n",
        "\n",
        "vocab_size_title = len(tok_title.word_index) + 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3wvyGUHCrGme"
      },
      "source": [
        "title_embedding_matrix = np.zeros((vocab_size_title, 50))\n",
        "for word, i in tok_title.word_index.items():\n",
        "    t_embedding_vector = embeddings_index.get(word)\n",
        "    if t_embedding_vector is not None:\n",
        "        title_embedding_matrix[i] = t_embedding_vector"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gTYtIhu0g3Yo"
      },
      "source": [
        "max_len_headline = headline_df.X_train_headline.apply(lambda x: len(x.split())).max()\n",
        "\n",
        "tok_headline = Tokenizer()\n",
        "tok_headline.fit_on_texts(headline_df.X_train_headline)\n",
        "vocab_size_headline = len(tok_headline.word_index) + 1\n",
        "encoded_headline = tok_headline.texts_to_sequences(headline_df.X_train_headline)\n",
        "padded_headline = pad_sequences(encoded_headline, maxlen=max_len_headline, padding='post')\n",
        "\n",
        "vocab_size_headline = len(tok_headline.word_index) + 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6JwhxtvqrHhS"
      },
      "source": [
        "headline_embedding_matrix = np.zeros((vocab_size_headline, 50))\n",
        "for word, i in tok_headline.word_index.items():\n",
        "    h_embedding_vector = embeddings_index.get(word)\n",
        "    if h_embedding_vector is not None:\n",
        "        headline_embedding_matrix[i] = h_embedding_vector"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "S7m9Ac5GqRh_"
      },
      "source": [
        "test_max_len_title = test_df.X_test_title.apply(lambda x: len(x.split())).max()\n",
        "\n",
        "test_tok_title = Tokenizer()\n",
        "test_tok_title.fit_on_texts(test_df.X_test_title)\n",
        "test_vocab_size_title = len(test_tok_title.word_index) + 1\n",
        "test_encoded_title = test_tok_title.texts_to_sequences(test_df.X_test_title)\n",
        "test_padded_title = pad_sequences(test_encoded_title, maxlen=test_max_len_title, padding='post')\n",
        "\n",
        "test_vocab_size_title = len(test_tok_title.word_index) + 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EA1Hg-5-qRiD"
      },
      "source": [
        "test_max_len_headline = test_df.X_test_headline.apply(lambda x: len(x.split())).max()\n",
        "\n",
        "test_tok_headline = Tokenizer()\n",
        "test_tok_headline.fit_on_texts(test_df.X_test_headline)\n",
        "test_vocab_size_headline = len(test_tok_headline.word_index) + 1\n",
        "test_encoded_headline = test_tok_headline.texts_to_sequences(test_df.X_test_headline)\n",
        "test_padded_headline = pad_sequences(test_encoded_headline, maxlen=test_max_len_headline, padding='post')\n",
        "\n",
        "test_vocab_size_headline = len(test_tok_headline.word_index) + 1"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zr6fulsEg3iq"
      },
      "source": [
        "x_train_title, x_valid_title, Y_train_title, y_valid_title = train_test_split(padded_title, y_train_title, shuffle = True, test_size = 0.1)\n",
        "x_train_headline, x_valid_headline, Y_train_headline, y_valid_headline = train_test_split(padded_headline, y_train_headline, shuffle = True, test_size = 0.1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fDk1mwx10rZq"
      },
      "source": [
        "import math\n",
        "from math import exp\n",
        "from keras import backend as K"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ka5CW9Geg3uS"
      },
      "source": [
        "def mod_tanh(x):\n",
        "    return K.tanh(0.6*x)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DwsB-3EA1dZs"
      },
      "source": [
        "### Title model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "23st4YDag33M"
      },
      "source": [
        "title_model = Sequential()\n",
        "title_model.add(Embedding(vocab_size_title, 50, input_length=max_len_title, weights=[title_embedding_matrix], trainable=True))\n",
        "title_model.add(Bidirectional(LSTM(20, return_sequences=True)))\n",
        "title_model.add(Dropout(0.3))\n",
        "title_model.add(BatchNormalization())\n",
        "title_model.add(Bidirectional(LSTM(20, return_sequences=True)))\n",
        "title_model.add(Dropout(0.3))\n",
        "title_model.add(BatchNormalization())\n",
        "title_model.add(Bidirectional(LSTM(20)))\n",
        "title_model.add(Dropout(0.3))\n",
        "title_model.add(BatchNormalization())\n",
        "title_model.add(Dense(64, activation='relu'))\n",
        "title_model.add(Dense(64, activation='relu'))\n",
        "title_model.add(Dense(1, activation=mod_tanh))\n",
        "title_model.compile(loss='mse', optimizer='adam', metrics=['mse', 'mae'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vJXzc4j22ETl"
      },
      "source": [
        "### Headline model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QPWIxGkmg38W"
      },
      "source": [
        "headline_model = Sequential()\n",
        "headline_model.add(Embedding(vocab_size_headline, 50, input_length=max_len_headline, weights=[headline_embedding_matrix], trainable=True))\n",
        "headline_model.add(Bidirectional(LSTM(20, return_sequences=True)))\n",
        "headline_model.add(Dropout(0.3))\n",
        "headline_model.add(BatchNormalization())\n",
        "headline_model.add(Bidirectional(LSTM(20, return_sequences=True)))\n",
        "headline_model.add(Dropout(0.3))\n",
        "headline_model.add(BatchNormalization())\n",
        "headline_model.add(Bidirectional(LSTM(20)))\n",
        "headline_model.add(Dropout(0.3))\n",
        "headline_model.add(BatchNormalization())\n",
        "headline_model.add(Dense(64, activation='relu'))\n",
        "headline_model.add(Dense(64, activation='relu'))\n",
        "headline_model.add(Dense(1, activation=mod_tanh))\n",
        "headline_model.compile(loss='mse', optimizer='adam', metrics=['mse', 'mae'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2CJtaiNpUL_E"
      },
      "source": [
        "import tensorflow"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nwwOvwLrg4Ay",
        "outputId": "111d9555-d06e-4011-80a4-ea1a04f0a7cd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        }
      },
      "source": [
        "with tensorflow.device('/device:GPU:0'):\n",
        "    title_model.fit(x_train_title, Y_train_title, epochs = 10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "1569/1569 [==============================] - 33s 21ms/step - loss: 0.0204 - mse: 0.0204 - mae: 0.1060\n",
            "Epoch 2/10\n",
            "1569/1569 [==============================] - 33s 21ms/step - loss: 0.0149 - mse: 0.0149 - mae: 0.0899\n",
            "Epoch 3/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0098 - mse: 0.0098 - mae: 0.0704\n",
            "Epoch 4/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0075 - mse: 0.0075 - mae: 0.0605\n",
            "Epoch 5/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0064 - mse: 0.0064 - mae: 0.0554\n",
            "Epoch 6/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0057 - mse: 0.0057 - mae: 0.0520\n",
            "Epoch 7/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0053 - mse: 0.0053 - mae: 0.0495\n",
            "Epoch 8/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0048 - mse: 0.0048 - mae: 0.0473\n",
            "Epoch 9/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0044 - mse: 0.0044 - mae: 0.0448\n",
            "Epoch 10/10\n",
            "1569/1569 [==============================] - 32s 21ms/step - loss: 0.0041 - mse: 0.0041 - mae: 0.0429\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i_y1fdW3g4FT",
        "outputId": "9a85ef73-e15e-4742-abc2-d6df1a721bc3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 353
        }
      },
      "source": [
        "with tensorflow.device('/device:GPU:0'):\n",
        "    headline_model.fit(x_train_headline, Y_train_headline, epochs = 10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "1569/1569 [==============================] - 47s 30ms/step - loss: 0.0229 - mse: 0.0229 - mae: 0.1168\n",
            "Epoch 2/10\n",
            "1569/1569 [==============================] - 46s 30ms/step - loss: 0.0187 - mse: 0.0187 - mae: 0.1055\n",
            "Epoch 3/10\n",
            "1569/1569 [==============================] - 46s 30ms/step - loss: 0.0128 - mse: 0.0128 - mae: 0.0864\n",
            "Epoch 4/10\n",
            "1569/1569 [==============================] - 46s 30ms/step - loss: 0.0090 - mse: 0.0090 - mae: 0.0715\n",
            "Epoch 5/10\n",
            "1569/1569 [==============================] - 46s 30ms/step - loss: 0.0074 - mse: 0.0074 - mae: 0.0642\n",
            "Epoch 6/10\n",
            "1569/1569 [==============================] - 46s 30ms/step - loss: 0.0064 - mse: 0.0064 - mae: 0.0593\n",
            "Epoch 7/10\n",
            "1569/1569 [==============================] - 47s 30ms/step - loss: 0.0056 - mse: 0.0056 - mae: 0.0554\n",
            "Epoch 8/10\n",
            "1569/1569 [==============================] - 46s 29ms/step - loss: 0.0049 - mse: 0.0049 - mae: 0.0517\n",
            "Epoch 9/10\n",
            "1569/1569 [==============================] - 46s 29ms/step - loss: 0.0045 - mse: 0.0045 - mae: 0.0490\n",
            "Epoch 10/10\n",
            "1569/1569 [==============================] - 46s 29ms/step - loss: 0.0040 - mse: 0.0040 - mae: 0.0465\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rKsdd6Jog4Re"
      },
      "source": [
        "title_valid_pred = title_model.predict(x_valid_title)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iRCijWC7g4aG"
      },
      "source": [
        "headline_valid_pred = headline_model.predict(x_valid_headline)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dctTSTWxg4fQ"
      },
      "source": [
        "from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bKEHO_ZbBRkg"
      },
      "source": [
        "mae_title=mean_absolute_error(y_valid_title,title_valid_pred)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vyd-11iFBRsl"
      },
      "source": [
        "mae_headline=mean_absolute_error(y_valid_headline,headline_valid_pred)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MSn__JgXg4jf"
      },
      "source": [
        "score=1-((0.4*mae_title)+(0.6*mae_headline))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RbvcZ_rKg4o6",
        "outputId": "98d66a24-eeec-4552-d67c-e23c37329443",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "score"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9320236916220008"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "b-7XwOfRYF0_"
      },
      "source": [
        "pred_title=title_model.predict(test_padded_title)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZedfKjY1YGFL",
        "outputId": "1c017d5d-641b-44cc-b8aa-98aa675bfddf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "source": [
        "pred_headline=headline_model.predict(test_padded_headline)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Model was constructed with shape (None, 60) for input Tensor(\"embedding_3_input:0\", shape=(None, 60), dtype=float32), but it was called on an input with incompatible shape (None, 53).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aUdcFbpFNawJ"
      },
      "source": [
        "submission=pd.DataFrame()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uP0eIlbyNa_t"
      },
      "source": [
        "submission['IDLink']=test['IDLink'].to_list()\n",
        "submission['SentimentTitle']=pred_title\n",
        "submission['SentimentHeadline']=pred_headline"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZQh7VU9UNarD",
        "outputId": "37480370-5c9c-4d45-abdc-6f118d766d0d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 195
        }
      },
      "source": [
        "submission.head()"
      ],
      "execution_count": null,
      "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>IDLink</th>\n",
              "      <th>SentimentTitle</th>\n",
              "      <th>SentimentHeadline</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>tFrqIR6Chj</td>\n",
              "      <td>0.172802</td>\n",
              "      <td>-0.000203</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>DVAaGErjlF</td>\n",
              "      <td>-0.094027</td>\n",
              "      <td>0.253928</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>OT9UIZm5M2</td>\n",
              "      <td>-0.001465</td>\n",
              "      <td>-0.084972</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>lflGp3q2Fj</td>\n",
              "      <td>-0.140512</td>\n",
              "      <td>0.170248</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>zDYG0SoovZ</td>\n",
              "      <td>-0.007781</td>\n",
              "      <td>0.014870</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       IDLink  SentimentTitle  SentimentHeadline\n",
              "0  tFrqIR6Chj        0.172802          -0.000203\n",
              "1  DVAaGErjlF       -0.094027           0.253928\n",
              "2  OT9UIZm5M2       -0.001465          -0.084972\n",
              "3  lflGp3q2Fj       -0.140512           0.170248\n",
              "4  zDYG0SoovZ       -0.007781           0.014870"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "79WUSHtDNaSj"
      },
      "source": [
        "submission.to_csv('./data/submission2.csv')"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L87reRsSg4xA"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7yF2KjCjg46J"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-_-LjA6kg42J"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3iej9U_wg4tU"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x4pMvWwGg4Wi"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WdEiK4C0g4ON"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jdO2ntQOg4Kv"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "31EcSgngg3z0"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GqIV1r6eg3p-"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zjonXT3Ag3nb"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}