Yelp_Combined_Reviews_(no_business_attributes)_DNN_RA.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyParf+J2s5Fd427vZtHB0bL",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/EAName/yelp-reviews/blob/main/Yelp_Combined_Reviews_(no_business_attributes)_DNN_RA.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jx5TpMB65nVW"
      },
      "source": [
        "## Import Packages"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "A4cNq1ya5nVW"
      },
      "outputs": [],
      "source": [
        "from packaging import version\n",
        "\n",
        "import tensorflow as tf\n",
        "import tensorflow_datasets as tfds\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras import layers\n",
        "from tensorflow.keras.layers import TextVectorization\n",
        "\n",
        "\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import re\n",
        "import string\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib as mpl\n",
        "import seaborn as sns\n",
        "import statistics\n",
        "\n",
        "%matplotlib inline\n",
        "\n",
        "from sklearn.metrics import confusion_matrix, classification_report\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.decomposition import PCA\n",
        "from sklearn.manifold import TSNE\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.metrics import mean_squared_error as MSE\n",
        "from sklearn.metrics import mean_absolute_error\n",
        "from sklearn.metrics import accuracy_score\n",
        "from sklearn.metrics import roc_auc_score, accuracy_score, confusion_matrix, silhouette_score\n",
        "from sklearn.metrics import precision_score, recall_score, f1_score, classification_report\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from datetime import datetime\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "3_tETfHF5nVZ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ed2324aa-dfe9-494f-be18-bc054e095467"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Load Business Dataset"
      ],
      "metadata": {
        "id": "8p-uBrOCSLl-"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "df_stlouis =pd.read_csv('/content/gdrive/MyDrive/Data/business_df_stlouis.csv')"
      ],
      "metadata": {
        "id": "Dd6K3Rz_m7wy"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_stlouis.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4wgCbVRom8UX",
        "outputId": "642caaed-30ac-47bb-f9bc-1eb71f33396a"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(3181, 69)"
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_stlouis.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 525
        },
        "id": "ptj_XWFlm8WQ",
        "outputId": "2e723b19-2fae-4df4-a83f-c325de7ca62e"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "              business_id                                 name  \\\n",
              "0  k0hlBqXX-Bt0vf1op7Jr1w                Tsevi's Pub And Grill   \n",
              "1  ljxNT9p0y7YMPx0fcNBGig  Tony's Restaurant & 3rd Street Cafe   \n",
              "2  1MeIwdbTnZOBFCKOrgaxuw               Ricardo's Italian Cafe   \n",
              "3  VrGxoSNZfSiT-G4OqRRGIw                     Sunset 44 Bistro   \n",
              "4  wjyUxHxkrvUvCOVqy4g_wg                         Oscar's Cafe   \n",
              "\n",
              "                  address         city state  postal_code   latitude  \\\n",
              "0       8025 Mackenzie Rd       Affton    MO      63123.0  38.565165   \n",
              "1            312 Piasa St        Alton    IL      62002.0  38.896563   \n",
              "2           1931 Park Ave  Saint Louis    MO      63104.0  38.617272   \n",
              "3         118 W Adams Ave  Saint Louis    MO      63122.0  38.583223   \n",
              "4  7520 Natural Bridge Rd     Normandy    MO      63121.0  38.704765   \n",
              "\n",
              "   longitude  stars  review_count  ...  Chicken Wings Sports Bars  \\\n",
              "0 -90.321087    3.0            19  ...              0           0   \n",
              "1 -90.186203    3.0            94  ...              0           0   \n",
              "2 -90.212784    3.5            80  ...              0           0   \n",
              "3 -90.407187    4.0            73  ...              0           0   \n",
              "4 -90.299811    4.0            11  ...              0           0   \n",
              "\n",
              "  JapaneseTotal HasTV_NA  RestaurantsPriceRange2_NA  \\\n",
              "0             0      0.0                        0.0   \n",
              "1             0      0.0                        0.0   \n",
              "2             0      0.0                        0.0   \n",
              "3             0      0.0                        0.0   \n",
              "4             0      1.0                        0.0   \n",
              "\n",
              "   RestaurantsReservations_NA  garage_NA validated_NA  lot_NA  valet_NA  \n",
              "0                         0.0        0.0          0.0     0.0       0.0  \n",
              "1                         0.0        0.0          0.0     0.0       0.0  \n",
              "2                         0.0        0.0          0.0     0.0       0.0  \n",
              "3                         0.0        0.0          0.0     0.0       0.0  \n",
              "4                         1.0        0.0          0.0     0.0       0.0  \n",
              "\n",
              "[5 rows x 69 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ed60cbb1-41c8-408e-95ef-0b1ac9c5ef7c\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>business_id</th>\n",
              "      <th>name</th>\n",
              "      <th>address</th>\n",
              "      <th>city</th>\n",
              "      <th>state</th>\n",
              "      <th>postal_code</th>\n",
              "      <th>latitude</th>\n",
              "      <th>longitude</th>\n",
              "      <th>stars</th>\n",
              "      <th>review_count</th>\n",
              "      <th>...</th>\n",
              "      <th>Chicken Wings</th>\n",
              "      <th>Sports Bars</th>\n",
              "      <th>JapaneseTotal</th>\n",
              "      <th>HasTV_NA</th>\n",
              "      <th>RestaurantsPriceRange2_NA</th>\n",
              "      <th>RestaurantsReservations_NA</th>\n",
              "      <th>garage_NA</th>\n",
              "      <th>validated_NA</th>\n",
              "      <th>lot_NA</th>\n",
              "      <th>valet_NA</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>k0hlBqXX-Bt0vf1op7Jr1w</td>\n",
              "      <td>Tsevi's Pub And Grill</td>\n",
              "      <td>8025 Mackenzie Rd</td>\n",
              "      <td>Affton</td>\n",
              "      <td>MO</td>\n",
              "      <td>63123.0</td>\n",
              "      <td>38.565165</td>\n",
              "      <td>-90.321087</td>\n",
              "      <td>3.0</td>\n",
              "      <td>19</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ljxNT9p0y7YMPx0fcNBGig</td>\n",
              "      <td>Tony's Restaurant &amp; 3rd Street Cafe</td>\n",
              "      <td>312 Piasa St</td>\n",
              "      <td>Alton</td>\n",
              "      <td>IL</td>\n",
              "      <td>62002.0</td>\n",
              "      <td>38.896563</td>\n",
              "      <td>-90.186203</td>\n",
              "      <td>3.0</td>\n",
              "      <td>94</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1MeIwdbTnZOBFCKOrgaxuw</td>\n",
              "      <td>Ricardo's Italian Cafe</td>\n",
              "      <td>1931 Park Ave</td>\n",
              "      <td>Saint Louis</td>\n",
              "      <td>MO</td>\n",
              "      <td>63104.0</td>\n",
              "      <td>38.617272</td>\n",
              "      <td>-90.212784</td>\n",
              "      <td>3.5</td>\n",
              "      <td>80</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>VrGxoSNZfSiT-G4OqRRGIw</td>\n",
              "      <td>Sunset 44 Bistro</td>\n",
              "      <td>118 W Adams Ave</td>\n",
              "      <td>Saint Louis</td>\n",
              "      <td>MO</td>\n",
              "      <td>63122.0</td>\n",
              "      <td>38.583223</td>\n",
              "      <td>-90.407187</td>\n",
              "      <td>4.0</td>\n",
              "      <td>73</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>wjyUxHxkrvUvCOVqy4g_wg</td>\n",
              "      <td>Oscar's Cafe</td>\n",
              "      <td>7520 Natural Bridge Rd</td>\n",
              "      <td>Normandy</td>\n",
              "      <td>MO</td>\n",
              "      <td>63121.0</td>\n",
              "      <td>38.704765</td>\n",
              "      <td>-90.299811</td>\n",
              "      <td>4.0</td>\n",
              "      <td>11</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>5 rows × 69 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ed60cbb1-41c8-408e-95ef-0b1ac9c5ef7c')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-ed60cbb1-41c8-408e-95ef-0b1ac9c5ef7c button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-ed60cbb1-41c8-408e-95ef-0b1ac9c5ef7c');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rvmQxAhOMOMb"
      },
      "source": [
        "## Load Reviews Datasets"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df =pd.read_csv('/content/gdrive/MyDrive/Data/yelp_reviews_clean_version4_st_louis.csv')"
      ],
      "metadata": {
        "id": "R936IjTMMOMb"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df = df.groupby(['business_id'])['text'].apply(lambda x: ','.join(x)).reset_index()"
      ],
      "metadata": {
        "id": "VFClgfecy16O"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "pd.set_option('display.width', None)\n",
        "pd.set_option('display.max_colwidth', -1)"
      ],
      "metadata": {
        "id": "BJBaUViez3Np",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d718fba7-d5cd-4dfc-ae33-b0cd8ff23af3"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:2: FutureWarning: Passing a negative integer is deprecated in version 1.0 and will not be supported in future version. Instead, use None to not limit the column width.\n",
            "  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def clean_doc(doc): \n",
        "    #split review into individual words\n",
        "    tokens=doc.split()\n",
        "    re_punc = re.compile('[%s]' % re.escape(string.punctuation))\n",
        "    # remove punctuation from each word\n",
        "    tokens = [re_punc.sub('', w) for w in tokens]\n",
        "\n",
        "    #word lemmzatization    \n",
        "    lemm = WordNetLemmatizer()\n",
        "    tokens=[lemm.lemmatize(word) for word in tokens]\n",
        "\n",
        "    # remove remaining tokens that are not alphabetic\n",
        "    tokens = [word for word in tokens if word.isalpha()]\n",
        "    # filter out short tokens\n",
        "    tokens = [word for word in tokens if len(word) > 3]\n",
        "    #lowercase all words\n",
        "    tokens = [word.lower() for word in tokens]\n",
        "    # filter out stop words\n",
        "    stop_words = set(stopwords.words('english'))\n",
        "    tokens = [w for w in tokens if not w in stop_words]     \n",
        "\n",
        "    # filter out common words\n",
        "    tokens = [w for w in tokens if not w in ['louis', 'restaurant', 'also', 'youre']] \n",
        "    # # word stemming    \n",
        "    # ps=PorterStemmer()\n",
        "    # tokens=[ps.stem(word) for word in tokens]\n",
        "\n",
        "    return tokens\n",
        "\n",
        "def final_processed_text(doc):\n",
        "    #this is a function to join the processed text back\n",
        "    ' '.join(doc)\n",
        "    return doc"
      ],
      "metadata": {
        "id": "9aA2s_9J1hao"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "combined = pd.merge(df, df_stlouis, on ='business_id')"
      ],
      "metadata": {
        "id": "xKhcQFDj0t3l"
      },
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "combined.head(2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "UdlbBD1X1qsj",
        "outputId": "3f5f36b0-3017-458c-f2a8-a09aa7a7e58c"
      },
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "              business_id  \\\n",
              "0  -1wnOZFsvsJPEXh_hWcmMA   \n",
              "1  -2gmbMDzKgYZ_8DOnJtPyw   \n",
              "\n",
              "                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          text  \\\n",
              "0  Tried their chicken adobo tacos tonight. Food looked and tasted fresh. Chicken seasoning was very good. The chips tasted like they were freshly made. Overall it was a wonderful experience. The staff was also super friendly. I'll definitely be back!,Amazing atmosphere only to be outdone by the tasty food. Can't wait for my next meal!,The best burgers in O'Fallon hands down! The Monday special is a fantastic deal. Great staff and family friendly.,Nice clean sports bar.  Had Fish Fry Friday and it was great!!!  New place so come check it out!!!,One of O'fallon's best family friendly neighborhood bar and restaurant. It's nice to have such a gem in the downtown. Prices are very reasonable and whether you are there for a few beers with friends or bringing your kids for a dinner, this place is the perfect. \\n\\nThe food is hand prepared and you can tell they take the time and energy to ensure it's fresh and tasty. Everything is great but a special shout out to the chips and salsa and burgers. Also they have a pretty legit fish fry on Friday's. \\n\\nLastly, whether you are a bud light drinker, prefer the craft brews, or a wino, they have it all.,Just went there for taco Tuesday! They were very busy but still had a few tables available. The way she says were amazing and worked their butts off. Taco plates were only eight dollars each for two shrimp tacos and chips and salsa. They were absolutely delicious this will be my new Tuesday dinner location,Brought my family in for lunch, we got seated and received drinks quickly. Our server was very nice and recommended a couple different items. We ordered and then waited, and waited, and waited. Finally after about 30 min I asked how much longer and did not get an answer. So we waited a few more minutes. Finally our server came over and told us they had moved our ticket up in line and we were next. Unfortunately, after waiting over 45 min the burgers and wraps we had ordered didn't seem worth it. We ended up leaving and never ate. Very disappointing experience.,My boyfriend and I stopped in about 12 am while he was in town working an event.  I have to tell you all the staff were great.  Since our names are Tom and Tracee, we got introduced to the owners who happen to have the same name.  They were very grateful we stopped in.  I also got the opportunity to try St Louis favorite root beer! We had their pizza, which was very delicious! Next time we are in town we will definitely stop by again,Swung in for a quick, late lunch.\\nI was immediately impressed with the crowd in a Monday afternoon.\\nI was under the impression that I would get a cold beer and some so/so bar food.\\nBoy! I was wrong!\\nI asked for a menu and the Bartender asked if I wanted to hear the special.\\nI said \"ok,\" she said \"$5 burger with fries.\" I didn't need the menu any longer.\\nMy drink was then made and was made correctly! (The vodka to soda, to ice, to glass, ratio was perfect!)\\nMy meal arrived as I ordered a second (and final) drink.\\nThe lettuce, tomatoes, pickles, and onions were served on the side. The burger was perfectly grilled as was the bun.\\nThe fries were fresh cut, homemade with skin on!\\nI can't believe this was a $5.00 meal!\\nI will definitely be back!,Was severly dissapointed and underwhelmed. Not a fan of pre-packaged food. My expectations were too high maybe. Portion size was good. Food quality and taste was poor at best...never again. Not even if starving.,Clean and updated bar/restaurant. Great service and great food. Burgers are hand patted and buns are grilled. Delicious! Amber at the bar was fantastic. Fries are fresh cut and plentiful.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                   \n",
              "1  Used to be my favorite.  Ate there one hot summer day last year. Not the coolest temp in the dinning area and staff looked a bit over warm too. Got my favorite beef and broccoli which tasted good I thought.  Couple hours later at home got sicker than a dog. ER staff said \"tainted food or lack of sanitation\" caused my vomiting and pain.  They took the name of where I ate but I did not file a report.  I Don't Advise eating there. I will never go there again!    It was not worth a trip to the ER and my stomach \\nBeing pumped out.     BEWARE.,The place had great food. This was the tops as far as what I've experienced in Chinese cuisine in the area. The people were nice and the food was reasonably priced. It was quite busy while my family was there. I recommend it.,This place has typical Chinese dishes. average and not at all outstanding food but even the mediocrity of the food is ruined by the owners attitude towards employees and then in turn from all the employees to the customers. I will never go here again.,I order at 3:44 and they said 40 minutes.  So far it has been an hour and no call.  I called\\nStore and no apology from staff. I was told sorry I wasn't here when you placed your order. \\n???customer service???   Really?  It is now after 5pm and I still don't have my ordered food.  I don't mind waiting, but this is silly.  The place is only a few minutes away.    I would have spent\\nmy money someplace else.  I called and the owner was upset with me.  I just called to cancel the order and he said he could not cancel because I paid with credit card.  REALLY!!!  Customer Service????,The man on the phone was friendly and helped me figure out the best portions for my order.  The delivery was fast and the food was the best I've had around here for your typical Chinese delivery.  This is now my go to Chinese!,Food is mediocre and the guy answering the phone was rude!  I've ordered the same thing a few times and it's always different,Awesome, great, fast delivery, food was fantastic, exactly what I ordered and it was hot (that doesn't always happen other places).  The Chicken Fried Rice was superb and the Crab Rangoon was incredible.  Definitely give this place a try...... I'm sure you'll like it.  Try to remember this is Collinsville, IL and they're trying as hard as they can.,just the kind of cruddy \"chinese food\" american's have come to expect that \"real\" chinese deign not to eat. chinese food in china is great. i feel bad seeing what passes as \"healthy\" chinese food in these parts.\\n\\ni got the snow pea/beef offering. they sure don't skimp on the rice. there was more rice than entree. (in china, you get one small bowl of rice...totaling less than 1/4 bulk of the entree) the beef that was stringy and unchewable was passable. the snow peas were good. the brown sauce of the platter had the consistency of old motor oil.\\n\\nthe egg drop soup came hot in a syrofoam cup. it tasted, well...chemically. i told the counter lady and she countered with \"that's the way we always make it\" and offered another choice.\\n\\nmy choice is to not return.,why is the staff always so angry? the profit margin in Chinese shops is huge. lighten up, both the Chinese and the local staff. bad vibes...,The food is cheaply priced, but always good. The inside could use renovating, but who eats Chinese food inside a restaurant anyways?,We ordered 3 different meals for lunch.  Pineapple Chicken , Pork Egg Foo Young and Cashew Chicken.  Comes with choice of fried rice or white.  Also soup or egg roll or crab Rangoon.  Our meal today was hot and delicious, a huge amount of food for two people.  We heated a plate for supper and it was delicious again.  Everything was great, we've bought many meals from them and never had a bad meal yet.  I ordered it on the Internet for the first time and it was simple to pick up as we were coming home.,Order food for delivery, advised I was in a wheelchair they needed extra time to get to the door. Delivery man left before getting to the door then ask me to come and get my refund from the store, instead of placing it back on my debit card. I spoke with manager he told me I was a idiot and hung up on me. Terrible customer service and will never give this restaurant my business again.,In town on a business trip and my usual go to place Peking Garden was closed for the week. The owners were on vacation. So I decided to give this place a whirl.\\n\\nI had the pork chow mein and a few other side items and they were all really really tasty and I would definitely come here again if Peking Garden is closed. \\n\\nI didn't have a problem with the staff like some other reviews had but they really didn't seem to care and I wasn't expecting much and besides I took my order to go.,This is my go to place for cheap delicious Chinese food. \\n\\nI wish I could talk about their menu in full, but the first time I ate from here I got the mushroom chicken and haven't been tempted to try anything else. I go with the fried rice with decent chunks of chicken giant pieces of halved button mushrooms, crunchy snap peas, carrots, and water chestnuts. All of this is served with soy sauce. My only complaint is the rice is sometimes a bit overcooked/dry but I can over look that. Possibly the real reason I keep coming back is the crab rangoon. With your order you get the choice of a side of crab rangoon or egg rolls. There is nothing unique about the crab rangoon, they just serve it with that golden brown crunchy exterior, and the sweet creamy crab filling you expect. For me they are kinda addictive.\\n\\nThere are 3 ways to get your dose of New China: Dine-In, Delivery, or Take Out. I've never dined in, small dining area with one large table that could probably seat 6, delivery in my experiences has been very slow, so I recommend take out. \\n\\nIt might not be the best Chinese food in the area, but for the portion and the price it works for me.,This all occurred on Monday, the 23 of April. First time ever eating from this restaurant, I had ordered pork fried rice and pot stickers. My mother had ordered Lou mien and crab Rangoon's, my sister had the same. It was all take out, I'm not sure if it was delivered. I thought it was a bit odd because the rice was yellow, I never had yellow rice before. Though I know there are different breads of rice. The rice was a little bland, though the pot stickers and the dipping sauce was very tasteful. Later that night I had decided to eat up the rest of my food to eat. The next day towards the evening, do you know what hit me. Instant throwing up and diarrhea. It takes up to 24 to 48 hours for food poisoning to occur in the body. All three of us were hit with food poisoning, my mother, my sister, and I. Be careful from now on eating from this restaurant, I would really consider it especially if you read all these reviews.,Food is pretty good, I liked it a lot better until I started getting the same delivery guy.  I won't order from there anymore unless I'm picking it up myself.  He is always late, and my food smells horribly of cigarette smoke.  Can't even enjoy it anymore.  It's so bad that I have to leave the bag outside or my house stinks of cigarettes.  It's a shame, I would give it 3-1/2 or 4 if not for that.,This food is AWESOME!  Everytime I order it is on time and fresh and excellent!  The beef and broccolli is outstanding!  Shrimp fried rice and crab rangoon are also excellent.  Delivery drivers are also so nice too!,New China does not honor their coupons. Jerks that work there would rather loose a LOT of business than honor a coupon when I was spending 40 bucks on lunch. There are many other places to eat. They will never get another dime of my money. Worst customer service ever,Only giving 1 star because I cannot give 0.  Called and was put on hold for 5 minutes then hung up on.  When I called back I was asked if I can hold again.  When I told them what happened the girl who answered stated they were too busy, I asked I should call order from somewhere else she said that's up to you and hung up again.  Completely rude and incompetent girl that answers your phones.,New to collinsville and searching for a new Chinese delivery restaurant. This is NOT it.  \\n\\nI had the single most disgusting dish of my life from here.  Chicken chow mein. There was nearly no chicken in the dish.  The vegetables consisted of onion and bok choi with a sliver of celery.  Mainly onion. The flavor was non existent.  Just onion and Barry washed bok choi.  There was no sauce but it came with crunchy noodles.   Seriously gross.  It got tossed.  And I'm not a picky person. \\n\\nThe fried chicken (for my little kids) tasted like it was cooked in dirty grease.  Twice.  \\n\\nThe crab Rangoon had a decent flavor on the inside but again, dirty oil flavor. \\n\\nWe also order beef and broccoli.  That wasn't terrible. The beef was super chewy but the flavor wasn't awful. Tons of broccoli. \\n\\nI'm one and done with this place.  It was so bad I won't bother with another chance.,WARNING....do not order food from here.....haven't gotten food here in a long time. Last time, ordered egg foo yung and the bottom patty was melted into the styrofoam container. Ordered crab stix...bit into one, and it was still wrapped in plastic....NEVER AGAIN,Worst Chinese food I've had in a long time. We ordered from them and they gave us none of the food we actually ordered. Tried to be a good sport and all of the food was TERRIBLE. I called and they said it wasn't their problem. Rude staff. Please do not order from here,Found a hair in my food and was only given half of my money back. One of the people was laughing about me finding a hair in my food.  Tried to contact manager, was told that the two english speaking inviduals there were \"considered managers since the other people working that day didn't speak english\". I asked for the owners name, was given the run around and a fake name, researched on the internet and found out the REAL owner's name. Tried to contact the owner, wrote a letter, no response.,New China on Beltline Rd in Collinsville has good food BUT every time I try to place an order for delivery over the phone something terrible ALWAYS happens. One of two things always happens when you order over the phone, either they take your order wrong and you don't get what you ordered when it arrives at your house or the delivery driver cannot find your address and drives around with your order in his/her car for 30+ minutes. \\n\\nThe worst part about the service here is when they make a mistake the managers never want to fix the problem (this has happened three times now over 2 years) and they want to argue with the customers point of view. The managers need to learn that the customer is always right, especially someone who has ordered over 50 times in the last few years. \\n\\nOn my last order when the delivery driver got lost i asked them to remake the food and I would come pick it up because the sweet and sour chicken would be soggy and not good after sitting in the drivers car in the sealed plastic container for over 35 minutes and the manager would not remake the food.\\n\\nSo overall, terrible delivery and even worse managing...,I always get super fresh food here. I've tried others, but I'm always going back here.  I think once maybe I had bad fried rice like maybe the oil was old or something,but I've been going here since it opened and I'm surprised it has low ratings.  It's mostly to go. Personally I don't really like eating there, but my food is always still hot even 30-40 minutes later.  I even think the food is as fresh at lunchtime or dinner.   Plus, love the fact that they deliver to Maryville so I don't have to get out if I'm sick or if there's a foot of snow on the ground.  It's my go to Chinese place.  \\n\\nMy favorites are the General TSO, garlic chicken, broccoli chicken, chicken fried rice and their egg rolls.  Not a fan of their over stuffed crab Rangoon as I like my feeling to be hot and runny they need to fry and more stuff them less.  It's always good and consistent for me.  I'm not expecting high in PF Chang's.   \n",
              "\n",
              "                name                    address          city state  \\\n",
              "0  1st Street Lounge  119 W 1st St               O'Fallon      IL     \n",
              "1  New China          1099 Belt Line Rd, Unit G  Collinsville  IL     \n",
              "\n",
              "   postal_code   latitude  longitude  stars  ...  Chicken Wings  Sports Bars  \\\n",
              "0  62269.0      38.591460 -89.911523  4.5    ...  0              0             \n",
              "1  62234.0      38.692212 -89.985062  2.5    ...  0              0             \n",
              "\n",
              "  JapaneseTotal HasTV_NA RestaurantsPriceRange2_NA  \\\n",
              "0  0             0.0      1.0                        \n",
              "1  0             0.0      0.0                        \n",
              "\n",
              "   RestaurantsReservations_NA  garage_NA  validated_NA lot_NA  valet_NA  \n",
              "0  1.0                         1.0        1.0           1.0    1.0       \n",
              "1  0.0                         1.0        1.0           1.0    1.0       \n",
              "\n",
              "[2 rows x 70 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-095fc317-f37c-4ca5-a1f7-c0c45d43e775\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>business_id</th>\n",
              "      <th>text</th>\n",
              "      <th>name</th>\n",
              "      <th>address</th>\n",
              "      <th>city</th>\n",
              "      <th>state</th>\n",
              "      <th>postal_code</th>\n",
              "      <th>latitude</th>\n",
              "      <th>longitude</th>\n",
              "      <th>stars</th>\n",
              "      <th>...</th>\n",
              "      <th>Chicken Wings</th>\n",
              "      <th>Sports Bars</th>\n",
              "      <th>JapaneseTotal</th>\n",
              "      <th>HasTV_NA</th>\n",
              "      <th>RestaurantsPriceRange2_NA</th>\n",
              "      <th>RestaurantsReservations_NA</th>\n",
              "      <th>garage_NA</th>\n",
              "      <th>validated_NA</th>\n",
              "      <th>lot_NA</th>\n",
              "      <th>valet_NA</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>-1wnOZFsvsJPEXh_hWcmMA</td>\n",
              "      <td>Tried their chicken adobo tacos tonight. Food looked and tasted fresh. Chicken seasoning was very good. The chips tasted like they were freshly made. Overall it was a wonderful experience. The staff was also super friendly. I'll definitely be back!,Amazing atmosphere only to be outdone by the tasty food. Can't wait for my next meal!,The best burgers in O'Fallon hands down! The Monday special is a fantastic deal. Great staff and family friendly.,Nice clean sports bar.  Had Fish Fry Friday and it was great!!!  New place so come check it out!!!,One of O'fallon's best family friendly neighborhood bar and restaurant. It's nice to have such a gem in the downtown. Prices are very reasonable and whether you are there for a few beers with friends or bringing your kids for a dinner, this place is the perfect. \\n\\nThe food is hand prepared and you can tell they take the time and energy to ensure it's fresh and tasty. Everything is great but a special shout out to the chips and salsa and burgers. Also they have a pretty legit fish fry on Friday's. \\n\\nLastly, whether you are a bud light drinker, prefer the craft brews, or a wino, they have it all.,Just went there for taco Tuesday! They were very busy but still had a few tables available. The way she says were amazing and worked their butts off. Taco plates were only eight dollars each for two shrimp tacos and chips and salsa. They were absolutely delicious this will be my new Tuesday dinner location,Brought my family in for lunch, we got seated and received drinks quickly. Our server was very nice and recommended a couple different items. We ordered and then waited, and waited, and waited. Finally after about 30 min I asked how much longer and did not get an answer. So we waited a few more minutes. Finally our server came over and told us they had moved our ticket up in line and we were next. Unfortunately, after waiting over 45 min the burgers and wraps we had ordered didn't seem worth it. We ended up leaving and never ate. Very disappointing experience.,My boyfriend and I stopped in about 12 am while he was in town working an event.  I have to tell you all the staff were great.  Since our names are Tom and Tracee, we got introduced to the owners who happen to have the same name.  They were very grateful we stopped in.  I also got the opportunity to try St Louis favorite root beer! We had their pizza, which was very delicious! Next time we are in town we will definitely stop by again,Swung in for a quick, late lunch.\\nI was immediately impressed with the crowd in a Monday afternoon.\\nI was under the impression that I would get a cold beer and some so/so bar food.\\nBoy! I was wrong!\\nI asked for a menu and the Bartender asked if I wanted to hear the special.\\nI said \"ok,\" she said \"$5 burger with fries.\" I didn't need the menu any longer.\\nMy drink was then made and was made correctly! (The vodka to soda, to ice, to glass, ratio was perfect!)\\nMy meal arrived as I ordered a second (and final) drink.\\nThe lettuce, tomatoes, pickles, and onions were served on the side. The burger was perfectly grilled as was the bun.\\nThe fries were fresh cut, homemade with skin on!\\nI can't believe this was a $5.00 meal!\\nI will definitely be back!,Was severly dissapointed and underwhelmed. Not a fan of pre-packaged food. My expectations were too high maybe. Portion size was good. Food quality and taste was poor at best...never again. Not even if starving.,Clean and updated bar/restaurant. Great service and great food. Burgers are hand patted and buns are grilled. Delicious! Amber at the bar was fantastic. Fries are fresh cut and plentiful.</td>\n",
              "      <td>1st Street Lounge</td>\n",
              "      <td>119 W 1st St</td>\n",
              "      <td>O'Fallon</td>\n",
              "      <td>IL</td>\n",
              "      <td>62269.0</td>\n",
              "      <td>38.591460</td>\n",
              "      <td>-89.911523</td>\n",
              "      <td>4.5</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>-2gmbMDzKgYZ_8DOnJtPyw</td>\n",
              "      <td>Used to be my favorite.  Ate there one hot summer day last year. Not the coolest temp in the dinning area and staff looked a bit over warm too. Got my favorite beef and broccoli which tasted good I thought.  Couple hours later at home got sicker than a dog. ER staff said \"tainted food or lack of sanitation\" caused my vomiting and pain.  They took the name of where I ate but I did not file a report.  I Don't Advise eating there. I will never go there again!    It was not worth a trip to the ER and my stomach \\nBeing pumped out.     BEWARE.,The place had great food. This was the tops as far as what I've experienced in Chinese cuisine in the area. The people were nice and the food was reasonably priced. It was quite busy while my family was there. I recommend it.,This place has typical Chinese dishes. average and not at all outstanding food but even the mediocrity of the food is ruined by the owners attitude towards employees and then in turn from all the employees to the customers. I will never go here again.,I order at 3:44 and they said 40 minutes.  So far it has been an hour and no call.  I called\\nStore and no apology from staff. I was told sorry I wasn't here when you placed your order. \\n???customer service???   Really?  It is now after 5pm and I still don't have my ordered food.  I don't mind waiting, but this is silly.  The place is only a few minutes away.    I would have spent\\nmy money someplace else.  I called and the owner was upset with me.  I just called to cancel the order and he said he could not cancel because I paid with credit card.  REALLY!!!  Customer Service????,The man on the phone was friendly and helped me figure out the best portions for my order.  The delivery was fast and the food was the best I've had around here for your typical Chinese delivery.  This is now my go to Chinese!,Food is mediocre and the guy answering the phone was rude!  I've ordered the same thing a few times and it's always different,Awesome, great, fast delivery, food was fantastic, exactly what I ordered and it was hot (that doesn't always happen other places).  The Chicken Fried Rice was superb and the Crab Rangoon was incredible.  Definitely give this place a try...... I'm sure you'll like it.  Try to remember this is Collinsville, IL and they're trying as hard as they can.,just the kind of cruddy \"chinese food\" american's have come to expect that \"real\" chinese deign not to eat. chinese food in china is great. i feel bad seeing what passes as \"healthy\" chinese food in these parts.\\n\\ni got the snow pea/beef offering. they sure don't skimp on the rice. there was more rice than entree. (in china, you get one small bowl of rice...totaling less than 1/4 bulk of the entree) the beef that was stringy and unchewable was passable. the snow peas were good. the brown sauce of the platter had the consistency of old motor oil.\\n\\nthe egg drop soup came hot in a syrofoam cup. it tasted, well...chemically. i told the counter lady and she countered with \"that's the way we always make it\" and offered another choice.\\n\\nmy choice is to not return.,why is the staff always so angry? the profit margin in Chinese shops is huge. lighten up, both the Chinese and the local staff. bad vibes...,The food is cheaply priced, but always good. The inside could use renovating, but who eats Chinese food inside a restaurant anyways?,We ordered 3 different meals for lunch.  Pineapple Chicken , Pork Egg Foo Young and Cashew Chicken.  Comes with choice of fried rice or white.  Also soup or egg roll or crab Rangoon.  Our meal today was hot and delicious, a huge amount of food for two people.  We heated a plate for supper and it was delicious again.  Everything was great, we've bought many meals from them and never had a bad meal yet.  I ordered it on the Internet for the first time and it was simple to pick up as we were coming home.,Order food for delivery, advised I was in a wheelchair they needed extra time to get to the door. Delivery man left before getting to the door then ask me to come and get my refund from the store, instead of placing it back on my debit card. I spoke with manager he told me I was a idiot and hung up on me. Terrible customer service and will never give this restaurant my business again.,In town on a business trip and my usual go to place Peking Garden was closed for the week. The owners were on vacation. So I decided to give this place a whirl.\\n\\nI had the pork chow mein and a few other side items and they were all really really tasty and I would definitely come here again if Peking Garden is closed. \\n\\nI didn't have a problem with the staff like some other reviews had but they really didn't seem to care and I wasn't expecting much and besides I took my order to go.,This is my go to place for cheap delicious Chinese food. \\n\\nI wish I could talk about their menu in full, but the first time I ate from here I got the mushroom chicken and haven't been tempted to try anything else. I go with the fried rice with decent chunks of chicken giant pieces of halved button mushrooms, crunchy snap peas, carrots, and water chestnuts. All of this is served with soy sauce. My only complaint is the rice is sometimes a bit overcooked/dry but I can over look that. Possibly the real reason I keep coming back is the crab rangoon. With your order you get the choice of a side of crab rangoon or egg rolls. There is nothing unique about the crab rangoon, they just serve it with that golden brown crunchy exterior, and the sweet creamy crab filling you expect. For me they are kinda addictive.\\n\\nThere are 3 ways to get your dose of New China: Dine-In, Delivery, or Take Out. I've never dined in, small dining area with one large table that could probably seat 6, delivery in my experiences has been very slow, so I recommend take out. \\n\\nIt might not be the best Chinese food in the area, but for the portion and the price it works for me.,This all occurred on Monday, the 23 of April. First time ever eating from this restaurant, I had ordered pork fried rice and pot stickers. My mother had ordered Lou mien and crab Rangoon's, my sister had the same. It was all take out, I'm not sure if it was delivered. I thought it was a bit odd because the rice was yellow, I never had yellow rice before. Though I know there are different breads of rice. The rice was a little bland, though the pot stickers and the dipping sauce was very tasteful. Later that night I had decided to eat up the rest of my food to eat. The next day towards the evening, do you know what hit me. Instant throwing up and diarrhea. It takes up to 24 to 48 hours for food poisoning to occur in the body. All three of us were hit with food poisoning, my mother, my sister, and I. Be careful from now on eating from this restaurant, I would really consider it especially if you read all these reviews.,Food is pretty good, I liked it a lot better until I started getting the same delivery guy.  I won't order from there anymore unless I'm picking it up myself.  He is always late, and my food smells horribly of cigarette smoke.  Can't even enjoy it anymore.  It's so bad that I have to leave the bag outside or my house stinks of cigarettes.  It's a shame, I would give it 3-1/2 or 4 if not for that.,This food is AWESOME!  Everytime I order it is on time and fresh and excellent!  The beef and broccolli is outstanding!  Shrimp fried rice and crab rangoon are also excellent.  Delivery drivers are also so nice too!,New China does not honor their coupons. Jerks that work there would rather loose a LOT of business than honor a coupon when I was spending 40 bucks on lunch. There are many other places to eat. They will never get another dime of my money. Worst customer service ever,Only giving 1 star because I cannot give 0.  Called and was put on hold for 5 minutes then hung up on.  When I called back I was asked if I can hold again.  When I told them what happened the girl who answered stated they were too busy, I asked I should call order from somewhere else she said that's up to you and hung up again.  Completely rude and incompetent girl that answers your phones.,New to collinsville and searching for a new Chinese delivery restaurant. This is NOT it.  \\n\\nI had the single most disgusting dish of my life from here.  Chicken chow mein. There was nearly no chicken in the dish.  The vegetables consisted of onion and bok choi with a sliver of celery.  Mainly onion. The flavor was non existent.  Just onion and Barry washed bok choi.  There was no sauce but it came with crunchy noodles.   Seriously gross.  It got tossed.  And I'm not a picky person. \\n\\nThe fried chicken (for my little kids) tasted like it was cooked in dirty grease.  Twice.  \\n\\nThe crab Rangoon had a decent flavor on the inside but again, dirty oil flavor. \\n\\nWe also order beef and broccoli.  That wasn't terrible. The beef was super chewy but the flavor wasn't awful. Tons of broccoli. \\n\\nI'm one and done with this place.  It was so bad I won't bother with another chance.,WARNING....do not order food from here.....haven't gotten food here in a long time. Last time, ordered egg foo yung and the bottom patty was melted into the styrofoam container. Ordered crab stix...bit into one, and it was still wrapped in plastic....NEVER AGAIN,Worst Chinese food I've had in a long time. We ordered from them and they gave us none of the food we actually ordered. Tried to be a good sport and all of the food was TERRIBLE. I called and they said it wasn't their problem. Rude staff. Please do not order from here,Found a hair in my food and was only given half of my money back. One of the people was laughing about me finding a hair in my food.  Tried to contact manager, was told that the two english speaking inviduals there were \"considered managers since the other people working that day didn't speak english\". I asked for the owners name, was given the run around and a fake name, researched on the internet and found out the REAL owner's name. Tried to contact the owner, wrote a letter, no response.,New China on Beltline Rd in Collinsville has good food BUT every time I try to place an order for delivery over the phone something terrible ALWAYS happens. One of two things always happens when you order over the phone, either they take your order wrong and you don't get what you ordered when it arrives at your house or the delivery driver cannot find your address and drives around with your order in his/her car for 30+ minutes. \\n\\nThe worst part about the service here is when they make a mistake the managers never want to fix the problem (this has happened three times now over 2 years) and they want to argue with the customers point of view. The managers need to learn that the customer is always right, especially someone who has ordered over 50 times in the last few years. \\n\\nOn my last order when the delivery driver got lost i asked them to remake the food and I would come pick it up because the sweet and sour chicken would be soggy and not good after sitting in the drivers car in the sealed plastic container for over 35 minutes and the manager would not remake the food.\\n\\nSo overall, terrible delivery and even worse managing...,I always get super fresh food here. I've tried others, but I'm always going back here.  I think once maybe I had bad fried rice like maybe the oil was old or something,but I've been going here since it opened and I'm surprised it has low ratings.  It's mostly to go. Personally I don't really like eating there, but my food is always still hot even 30-40 minutes later.  I even think the food is as fresh at lunchtime or dinner.   Plus, love the fact that they deliver to Maryville so I don't have to get out if I'm sick or if there's a foot of snow on the ground.  It's my go to Chinese place.  \\n\\nMy favorites are the General TSO, garlic chicken, broccoli chicken, chicken fried rice and their egg rolls.  Not a fan of their over stuffed crab Rangoon as I like my feeling to be hot and runny they need to fry and more stuff them less.  It's always good and consistent for me.  I'm not expecting high in PF Chang's.</td>\n",
              "      <td>New China</td>\n",
              "      <td>1099 Belt Line Rd, Unit G</td>\n",
              "      <td>Collinsville</td>\n",
              "      <td>IL</td>\n",
              "      <td>62234.0</td>\n",
              "      <td>38.692212</td>\n",
              "      <td>-89.985062</td>\n",
              "      <td>2.5</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2 rows × 70 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-095fc317-f37c-4ca5-a1f7-c0c45d43e775')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-095fc317-f37c-4ca5-a1f7-c0c45d43e775 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-095fc317-f37c-4ca5-a1f7-c0c45d43e775');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "combined.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VTlugZeg198D",
        "outputId": "eb900f78-c100-4523-a4fc-45fe4e7cf9f6"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(3181, 70)"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "combined['stars'] = combined['stars'].replace([1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0], [1, 2, 3, 4, 5, 6, 7, 8, 9])\n",
        "combined['stars'] = combined['stars'].astype(int)"
      ],
      "metadata": {
        "id": "S3oukk8DQxaA"
      },
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train, test = train_test_split(combined, test_size=0.3, random_state = 42)"
      ],
      "metadata": {
        "id": "VnFOdkX9eoAS"
      },
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test, val = train_test_split(test, test_size=0.5, random_state = 42)"
      ],
      "metadata": {
        "id": "-lxDrBSCey37"
      },
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df2 = train\n",
        "labels = df2.pop('stars')\n",
        "df2 = df2.iloc[:,1]\n",
        "train_ds = tf.data.Dataset.from_tensor_slices((df2, labels))\n",
        "train_ds = train_ds.batch(32)"
      ],
      "metadata": {
        "id": "cyn122OopHnt"
      },
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df2 = test\n",
        "labels = df2.pop('stars')\n",
        "df2 = df2.iloc[:,1]\n",
        "test_ds = tf.data.Dataset.from_tensor_slices((df2, labels))\n",
        "test_ds = test_ds.batch(32)"
      ],
      "metadata": {
        "id": "Bznxwyt_pHqI"
      },
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df2 = val\n",
        "labels = df2.pop('stars')\n",
        "df2 = df2.iloc[:,1]\n",
        "val_ds = tf.data.Dataset.from_tensor_slices((df2, labels))\n",
        "val_ds = val_ds.batch(32)"
      ],
      "metadata": {
        "id": "MKClyp7EpURz"
      },
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_IfqCmSJ5nVa"
      },
      "source": [
        "## Display The Number of Batches"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_ds), len(val_ds), len(test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EBcAo2QCgoym",
        "outputId": "e40c5d5a-faa2-49ae-9a7e-72c46b9b7cc5"
      },
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(70, 15, 15)"
            ]
          },
          "metadata": {},
          "execution_count": 47
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(train), len(val), len(test)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K_z0GBuliPq6",
        "outputId": "a7b0439f-f31d-4fbe-d9f5-98afe965db86"
      },
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2226, 478, 477)"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lkl6vfG95nVb"
      },
      "source": [
        "## Displaying The Shapes and Dtypes of the First Batch"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p-Oxbpeg5nVb",
        "outputId": "b51e243d-7829-4d35-835a-ba39372fd9a5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "inputs.shape: (32,)\n",
            "\n",
            "inputs.dtype: <dtype: 'string'>\n",
            "\n",
            "targets.shape: (32,)\n",
            "\n",
            "targets.dtype: <dtype: 'int64'>\n",
            "\n",
            "inputs[0]: tf.Tensor(b'BLT and Buffalo chicken grilled cheese were very good. Had chili and the fries as the side and were good as well. Will go back.,I was lucky when a local suggested this gem in this beautiful town. I have enjoyed both my experiences. The food and hospitality was above and beyond.,Stopped in on our way through for a early afternoon lunch. The food was amazing and the staff was extremely nice. It\\'s a small cute restaurant on a very unique street with lots of shops and restaurants.,I took a trip out to The Lantern with my sister and her kids after it received rave reviews in a local foodie group she belongs to.\\nWe were seated and immediately informed our orders might take a while due to a large party arriving just before us. This is truly a family run operation, with one cook shouldering all of food prep. We appreciated the heads up but settled in with some coffee and decided to stay. \\nOur food didn\\'t arrive at the same time, not surprising with one person cooking. Luckily, our server gave my niece a pack of cards to play with and the kids\\' French toast came out first so we didn\\'t have to listen to complaints of starvation. \\nI ordered the breakfast poutine, a mass of potato, sausage, sausage gravy, eggs and cheese. I was expecting French fries but got home fries/cubed potatoes, which was an better, in my opinion. My sister\\'s biscuits and gravy was a big hit with her and her picky son, who couldn\\'t shove forkfuls into his mouth fast enough. I felt his pain; my dish was huge yet I managed to eat all of it, unable to find the willpower to just put my fork down. The gravy here is really quite good and actually has a nice flavor. \\nOur server was very personable and considerate-we enjoyed talking to her when she stopped by. \\nThe menu changes monthly, which can be great but caused my sister some sadness as she was mentally bookmarking the breakfast poutine for her next visit. However, I\\'m sure she\\'ll find something to enjoy with the new menu. 4 of us (1 kids meal, 3 regular meals, 4 drinks) got out of there for under $40 minus tip. Delicious food, great service and great prices. If you go with a group be patient. Again, this is a small operation. Ask for the playing cards and get some coffee. It\\'s worth the wait.,Cute little family ran restaurant. I tried a little bit of everything : biscuits and gravy, potato\\'s, breakfast sandwich & French toast. All of them get 5 stars. I will be back and look forward to trying new things they have!,Disappointing food, and frequent spelling errors on menus. The religious tone is overwhelming and unwelcoming to anyone who is not a Christian. The food was adequate at best, but the atmosphere and religious overtones are too much.,Had no idea this place had been open as long as they have been!  And can\\'t believe nobody\\'s added them to Yelp before!  Stopped in after gym and then yoga, hoping for breakfast, which they had sold out of.  That\\'s got\\'ta tell ya something!  So, I picked the B.L.E.S.T. sandwich and it was unbelievably good!!!  Got the macaroni and cheese as my side and it was DELISH also!  They offer cold brew, which is my favorite!  CanNOT wait to get back here and try their breakfast!  Good service.  Reasonable prices.  GREAT food.,Finally made it back for breakfast!  The Rise & Shine was divine!  Be sure to have enough time to enjoy this place.  It\\'s not a fast food kind\\'a place (THANK GOODNESS!).,The Lantern was suggested to me by a couple of locals. I am very pleased that I went with their recommendation. I had the Buffalo chicken sandwich. It was wonderful! I was pleasantly surprised when I got my bill. It was less than I was expecting. I discovered that the waitress applied a \"healthcare worker\" discount (I was wearing scrubs). It was a sweet gesture and made me feel appreciated. I will definitely return the next time I am in Lebanon.,Top 5 breakfasts I have had! Biscuits were so fluffy and moist, and gravy was so flavorful. Check this spot out!', shape=(), dtype=string)\n",
            "\n",
            "targets[0]: tf.Tensor(8, shape=(), dtype=int64)\n"
          ]
        }
      ],
      "source": [
        "for inputs, targets in train_ds:\n",
        "    print(\"inputs.shape:\", inputs.shape)\n",
        "    print()\n",
        "    print(\"inputs.dtype:\", inputs.dtype)\n",
        "    print()\n",
        "    print(\"targets.shape:\", targets.shape)\n",
        "    print()\n",
        "    print(\"targets.dtype:\", targets.dtype)\n",
        "    print()\n",
        "    print(\"inputs[0]:\", inputs[0])\n",
        "    print()\n",
        "    print(\"targets[0]:\", targets[0])\n",
        "    break"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_history(history):\n",
        "  losses = history.history['loss']\n",
        "  accs = history.history['accuracy']\n",
        "  val_losses = history.history['val_loss']\n",
        "  val_accs = history.history['val_accuracy']\n",
        "  epochs = len(losses)\n",
        "\n",
        "  plt.figure(figsize=(16, 4))\n",
        "  for i, metrics in enumerate(zip([losses, accs], [val_losses, val_accs], ['Loss', 'Accuracy'])):\n",
        "    plt.subplot(1, 2, i + 1)\n",
        "    plt.plot(range(epochs), metrics[0], label='Training {}'.format(metrics[2]))\n",
        "    plt.plot(range(epochs), metrics[1], label='Validation {}'.format(metrics[2]))\n",
        "    plt.legend()\n",
        "  plt.show()"
      ],
      "metadata": {
        "id": "yzHGgRJ6JRuD"
      },
      "execution_count": 50,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def print_validation_report(test_labels, predictions):\n",
        "    print(\"Classification Report\")\n",
        "    print(classification_report(test_labels, predictions))\n",
        "    print('Accuracy Score: {:.3f}'.format(accuracy_score(test_labels, predictions)))\n",
        "    print('Mean Absolute Error: {:.3f}'.format(mean_absolute_error(test_labels, predictions)))\n",
        "    print('Root Mean Square Error: {:.3f}'.format(np.sqrt(MSE(test_labels, predictions))))\n",
        "    print('Mean Square Error: {:.3f}'.format(MSE(test_labels, predictions, squared = True)))"
      ],
      "metadata": {
        "id": "lyFZe1XEZc_9"
      },
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def plot_confusion_matrix(y_true, y_pred):\n",
        "    mtx = confusion_matrix(y_true, y_pred)\n",
        "    fig, ax = plt.subplots(figsize=(8,8))\n",
        "    sns.heatmap(mtx, annot=True, fmt='d', linewidths=.75,  cbar=False, ax=ax,cmap='Blues',linecolor='white')\n",
        "    #  square=True,\n",
        "    plt.ylabel('true label')\n",
        "    plt.xlabel('predicted label')\n",
        "    ax.xaxis.set_ticklabels(['1', '1.5', '2', '2.5', '3', '3.5', '4', '4.5', '5'])\n",
        "    ax.yaxis.set_ticklabels(['1', '1.5', '2', '2.5', '3', '3.5', '4', '4.5', '5'])"
      ],
      "metadata": {
        "id": "4tFlVmj7Z8Gh"
      },
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XH-IOhSN5nVc"
      },
      "source": [
        "## Preprocessing Datasets TextVectorization Layer"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 53,
      "metadata": {
        "id": "lk_p63Bj5nVc"
      },
      "outputs": [],
      "source": [
        "text_vectorization = TextVectorization(\n",
        "    max_tokens=1000,\n",
        "    output_mode=\"count\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 54,
      "metadata": {
        "id": "s-e9b9T35nVd"
      },
      "outputs": [],
      "source": [
        "text_only_train_ds = train_ds.map(lambda x, y: x)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2WZlxRCo5nVd"
      },
      "source": [
        "## Adapt Method - Standardize Text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "metadata": {
        "id": "p8qD1a4O5nVd"
      },
      "outputs": [],
      "source": [
        "text_vectorization.adapt(text_only_train_ds)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 56,
      "metadata": {
        "id": "Nnv09sg25nVd"
      },
      "outputs": [],
      "source": [
        "count_1gram_train_ds = train_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_1gram_val_ds = val_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_1gram_test_ds = test_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_data = [(example.numpy(), label.numpy()) for example, label in count_1gram_test_ds]\n",
        "test_y = []\n",
        "\n",
        "for i in range(len(test)//32):\n",
        "  for j in range(32):\n",
        "    test_y.append(test_data[i][1][j])\n",
        "\n",
        "for j in range(len(test)%32):\n",
        "  test_y.append(test_data[len(test)//32][1][j])\n",
        "\n",
        "test_y = np.array(test_y)"
      ],
      "metadata": {
        "id": "x3_TGvD3lY13"
      },
      "execution_count": 57,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test_y.size"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cNgx2qCNmO38",
        "outputId": "a1233e99-40a9-46ac-c100-959cd8184643"
      },
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "477"
            ]
          },
          "metadata": {},
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Count Unigram Model"
      ],
      "metadata": {
        "id": "bEDU66iggfZM"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 59,
      "metadata": {
        "id": "wh2FgcZbgfZM"
      },
      "outputs": [],
      "source": [
        "text_vectorization = TextVectorization(\n",
        "    ngrams=1,\n",
        "    max_tokens=1000,\n",
        "    output_mode=\"count\",\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 60,
      "metadata": {
        "id": "pb8FVXP4gfZM"
      },
      "outputs": [],
      "source": [
        "text_vectorization.adapt(text_only_train_ds)\n",
        "count_1gram_train_ds = train_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_1gram_val_ds = val_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_1gram_test_ds = test_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "06DLyfBHgfZM"
      },
      "source": [
        "## Build Count Unigram Model "
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=128):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "  #  x = layers.Dense(256, activation=\"relu\", kernel_regularizer=tf.keras.regularizers.L2(0.001))(x)\n",
        " #   x = layers.Dense(64, activation=\"relu\", kernel_regularizer=tf.keras.regularizers.L2(0.001))(x)\n",
        "  #  x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "\n",
        "    opt = keras.optimizers.Adam(learning_rate=0.001)\n",
        "    model.compile(optimizer=opt,\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "UZxOrBuWgfZN"
      },
      "execution_count": 71,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Unigram = get_model()\n",
        "model_count_Unigram.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "595dea2c-7aff-4161-dcbb-f5929221e59e",
        "id": "JodJLUMcgfZN"
      },
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_4 (InputLayer)        [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 128)               128128    \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 128)               0         \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 129,418\n",
            "Trainable params: 129,418\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 73,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0be91698-f426-47e3-bca0-6f966abb84fc",
        "id": "RnTkLjMXgfZN"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 5s 65ms/step - loss: 35.5959 - accuracy: 0.2336 - val_loss: 10.7133 - val_accuracy: 0.2197\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 5.1340 - accuracy: 0.3055 - val_loss: 2.6322 - val_accuracy: 0.2490\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.0109 - accuracy: 0.3149 - val_loss: 1.9683 - val_accuracy: 0.3222\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7836 - accuracy: 0.3787 - val_loss: 1.7263 - val_accuracy: 0.3745\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6500 - accuracy: 0.3801 - val_loss: 1.6743 - val_accuracy: 0.4038\n",
            "Epoch 6/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6039 - accuracy: 0.4057 - val_loss: 1.4839 - val_accuracy: 0.4414\n",
            "Epoch 7/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.4888 - accuracy: 0.4313 - val_loss: 1.4889 - val_accuracy: 0.4331\n",
            "Epoch 8/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.4324 - accuracy: 0.4533 - val_loss: 1.4137 - val_accuracy: 0.4623\n",
            "Epoch 9/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.3871 - accuracy: 0.4438 - val_loss: 1.3403 - val_accuracy: 0.4874\n",
            "Epoch 10/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.3296 - accuracy: 0.4834 - val_loss: 1.3402 - val_accuracy: 0.4707\n",
            "Epoch 11/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.2833 - accuracy: 0.4888 - val_loss: 1.3610 - val_accuracy: 0.4686\n",
            "Epoch 12/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.2087 - accuracy: 0.5153 - val_loss: 1.3123 - val_accuracy: 0.4770\n",
            "15/15 [==============================] - 1s 41ms/step - loss: 1.4043 - accuracy: 0.4633\n",
            "Test acc: 0.463\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "     tf.keras.callbacks.ModelCheckpoint(\"binary_2gram.keras\",save_best_only=True)\n",
        "    ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "history = model_count_Unigram.fit(count_1gram_train_ds.cache(),\n",
        "          validation_data=count_1gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_count_Unigram = keras.models.load_model(\"binary_2gram.keras\")\n",
        "print(f\"Test acc: {model_count_Unigram.evaluate(count_1gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Unigram.evaluate(count_1gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "fcf1ed56-1b8e-4712-94a4-e85a9c8e1994",
        "id": "M3moftFRgfZN"
      },
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 42ms/step - loss: 1.4043 - accuracy: 0.4633\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.4043138027191162, 0.4633123576641083]"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "aa0b4312-ac7a-4734-f319-74217446018a",
        "id": "mIxoOQnOgfZO"
      },
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_count_Unigram.predict(count_1gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "63675c39-db22-4c60-8fba-a549802c12ac",
        "id": "lkOY0RKzgfZO"
      },
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 41ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "6f6e744c-e350-498b-afbc-017b52b35ab9",
        "id": "x47ZfRTqgfZO"
      },
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.43      0.18      0.25        17\n",
            "           4       0.27      0.08      0.12        39\n",
            "           5       0.31      0.26      0.28        78\n",
            "           6       0.43      0.74      0.55       124\n",
            "           7       0.54      0.50      0.52       130\n",
            "           8       0.62      0.54      0.58        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.46       477\n",
            "   macro avg       0.29      0.25      0.26       477\n",
            "weighted avg       0.44      0.46      0.43       477\n",
            "\n",
            "Accuracy Score: 0.463\n",
            "Mean Absolute Error: 0.658\n",
            "Root Mean Square Error: 0.976\n",
            "Mean Square Error: 0.952\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "9ff1da8e-6055-46de-efd0-31ddfa18241a",
        "id": "r7mPQXwsgfZO"
      },
      "execution_count": 78,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2 layer model"
      ],
      "metadata": {
        "id": "pJoKI6vagfZO"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=128):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    x = layers.Dense(128, activation=\"relu\")(x)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "    model.compile(optimizer=\"adam\",\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "E6WtmunCgfZO"
      },
      "execution_count": 103,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Unigram = get_model()\n",
        "model_count_Unigram.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e9fae883-2044-4c69-afc4-a8ed46f06885",
        "id": "9bkHxsUEgfZO"
      },
      "execution_count": 104,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_9\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_10 (InputLayer)       [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_23 (Dense)            (None, 128)               128128    \n",
            "                                                                 \n",
            " dropout_12 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_24 (Dense)            (None, 128)               16512     \n",
            "                                                                 \n",
            " dropout_13 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_25 (Dense)            (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 145,930\n",
            "Trainable params: 145,930\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 105,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "84c55a37-f589-46aa-927c-3dffdbd4a7ea",
        "id": "29PPUMc-gfZP"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 4s 57ms/step - loss: 21.0795 - accuracy: 0.2296 - val_loss: 2.9979 - val_accuracy: 0.3117\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 4ms/step - loss: 3.0404 - accuracy: 0.2862 - val_loss: 1.9237 - val_accuracy: 0.3285\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.0097 - accuracy: 0.3113 - val_loss: 1.8307 - val_accuracy: 0.3515\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.8691 - accuracy: 0.3297 - val_loss: 1.7392 - val_accuracy: 0.3515\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 4ms/step - loss: 1.7383 - accuracy: 0.3401 - val_loss: 1.5950 - val_accuracy: 0.3933\n",
            "Epoch 6/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6718 - accuracy: 0.3527 - val_loss: 1.5973 - val_accuracy: 0.3828\n",
            "Epoch 7/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6087 - accuracy: 0.3477 - val_loss: 1.4772 - val_accuracy: 0.3954\n",
            "Epoch 8/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5194 - accuracy: 0.3715 - val_loss: 1.4921 - val_accuracy: 0.3724\n",
            "Epoch 9/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5139 - accuracy: 0.3971 - val_loss: 1.4067 - val_accuracy: 0.4059\n",
            "Epoch 10/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5261 - accuracy: 0.3774 - val_loss: 1.4789 - val_accuracy: 0.3954\n",
            "Epoch 11/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5053 - accuracy: 0.3962 - val_loss: 1.4433 - val_accuracy: 0.3996\n",
            "Epoch 12/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.4961 - accuracy: 0.4061 - val_loss: 1.4387 - val_accuracy: 0.4017\n",
            "15/15 [==============================] - 1s 42ms/step - loss: 1.5172 - accuracy: 0.4277\n",
            "Test acc: 0.428\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "     tf.keras.callbacks.ModelCheckpoint(\"binary_2gram.keras\",save_best_only=True)\n",
        "    ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "model_count_Unigram.fit(count_1gram_train_ds.cache(),\n",
        "          validation_data=count_1gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_count_Unigram = keras.models.load_model(\"binary_2gram.keras\")\n",
        "print(f\"Test acc: {model_count_Unigram.evaluate(count_1gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Unigram.evaluate(count_1gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "333404af-1770-4e35-8849-41c0aa26544a",
        "id": "BskjqiTbgfZP"
      },
      "execution_count": 106,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 67ms/step - loss: 1.5172 - accuracy: 0.4277\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.5171765089035034, 0.42767295241355896]"
            ]
          },
          "metadata": {},
          "execution_count": 106
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "outputId": "7eba050b-90f9-43a5-d249-6e82f1d16e14",
        "id": "Zi2wWcBtgfZP"
      },
      "execution_count": 107,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_count_Unigram.predict(count_1gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2a7e18fc-c27e-4b2d-842a-f7ad0e548f9b",
        "id": "u_vTfa1BgfZP"
      },
      "execution_count": 108,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 41ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2d2e91af-fe81-45c8-bdc9-ac41eaebffc2",
        "id": "dkX7MFO5gfZP"
      },
      "execution_count": 109,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.54      0.41      0.47        17\n",
            "           4       0.50      0.08      0.13        39\n",
            "           5       0.40      0.36      0.38        78\n",
            "           6       0.39      0.75      0.51       124\n",
            "           7       0.48      0.36      0.41       130\n",
            "           8       0.52      0.37      0.43        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.43       477\n",
            "   macro avg       0.31      0.26      0.26       477\n",
            "weighted avg       0.43      0.43      0.40       477\n",
            "\n",
            "Accuracy Score: 0.428\n",
            "Mean Absolute Error: 0.696\n",
            "Root Mean Square Error: 1.003\n",
            "Mean Square Error: 1.006\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "0e65b27b-871f-4c99-e537-31a5d3d488e3",
        "id": "GS7SRUJ_gfZP"
      },
      "execution_count": 110,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Count Bigram Model"
      ],
      "metadata": {
        "id": "m4xkHjrqqjrV"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 111,
      "metadata": {
        "id": "IKeZB55zqsKs"
      },
      "outputs": [],
      "source": [
        "text_vectorization = TextVectorization(\n",
        "    ngrams=2,\n",
        "    max_tokens=1000,\n",
        "    output_mode=\"count\",\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y5TQ4gp2qsKt"
      },
      "source": [
        "## Build Count Bigram Model "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 112,
      "metadata": {
        "id": "3qY0kpVAqsKt"
      },
      "outputs": [],
      "source": [
        "text_vectorization.adapt(text_only_train_ds)\n",
        "count_2gram_train_ds = train_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_2gram_val_ds = val_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "count_2gram_test_ds = test_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=128):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "    model.compile(optimizer=\"adam\",\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "ZS3iCRROrUBu"
      },
      "execution_count": 113,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Bigram = get_model()\n",
        "model_count_Bigram.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UNHDLNLpkMO_",
        "outputId": "ff8d6776-b96e-4de7-f19a-eb21439ac29c"
      },
      "execution_count": 114,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_10\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_11 (InputLayer)       [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_26 (Dense)            (None, 128)               128128    \n",
            "                                                                 \n",
            " dropout_14 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_27 (Dense)            (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 129,418\n",
            "Trainable params: 129,418\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 116,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ecf5d40a-e1ab-4df8-b8ef-977ed459ab30",
        "id": "yK5rCDPAqsKt"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 8s 110ms/step - loss: 82.0017 - accuracy: 0.2534 - val_loss: 26.9525 - val_accuracy: 0.1778\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 18.9955 - accuracy: 0.3396 - val_loss: 2.8448 - val_accuracy: 0.4205\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.2682 - accuracy: 0.3338 - val_loss: 1.9702 - val_accuracy: 0.2803\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.8760 - accuracy: 0.3522 - val_loss: 1.6996 - val_accuracy: 0.4100\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7016 - accuracy: 0.4052 - val_loss: 1.5288 - val_accuracy: 0.4289\n",
            "Epoch 6/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6610 - accuracy: 0.4191 - val_loss: 1.7101 - val_accuracy: 0.3326\n",
            "Epoch 7/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7397 - accuracy: 0.3904 - val_loss: 1.6195 - val_accuracy: 0.4268\n",
            "Epoch 8/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6477 - accuracy: 0.4191 - val_loss: 1.6438 - val_accuracy: 0.4142\n",
            "15/15 [==============================] - 1s 64ms/step - loss: 1.6715 - accuracy: 0.4151\n",
            "Test acc: 0.415\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "     tf.keras.callbacks.ModelCheckpoint(\"binary_2gram.keras\",save_best_only=True)\n",
        "    ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "history = model_count_Bigram.fit(count_2gram_train_ds.cache(),\n",
        "          validation_data=count_2gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_count_Bigram = keras.models.load_model(\"binary_2gram.keras\")\n",
        "print(f\"Test acc: {model_count_Bigram.evaluate(count_2gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Bigram.evaluate(count_2gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "346624c6-a7bf-4e8e-eed0-f8ef76f6b345",
        "id": "YMW7i9RBqsKu"
      },
      "execution_count": 117,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 3s 188ms/step - loss: 1.6715 - accuracy: 0.4151\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.6715388298034668, 0.4150943458080292]"
            ]
          },
          "metadata": {},
          "execution_count": 117
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "outputId": "2d733513-44c1-4ec9-928d-566e1cce1de5",
        "id": "z7Gceja-qsKu"
      },
      "execution_count": 118,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_count_Bigram.predict(count_2gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "60414d08-e5ba-4aa3-8103-43205c025ba1",
        "id": "GDW-FNaJqsKu"
      },
      "execution_count": 119,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 58ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5d31a6e6-efbf-4985-fadd-1707e5049a53",
        "id": "eBncrCklqsKu"
      },
      "execution_count": 120,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.45      0.29      0.36        17\n",
            "           4       0.28      0.13      0.18        39\n",
            "           5       0.40      0.47      0.43        78\n",
            "           6       0.38      0.57      0.45       124\n",
            "           7       0.46      0.29      0.36       130\n",
            "           8       0.51      0.60      0.55        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.42       477\n",
            "   macro avg       0.27      0.26      0.26       477\n",
            "weighted avg       0.40      0.42      0.39       477\n",
            "\n",
            "Accuracy Score: 0.415\n",
            "Mean Absolute Error: 0.748\n",
            "Root Mean Square Error: 1.077\n",
            "Mean Square Error: 1.159\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "e0422f40-b627-4f0c-9370-688e526287bc",
        "id": "8uQVLcTGqsKu"
      },
      "execution_count": 121,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2 layer model"
      ],
      "metadata": {
        "id": "yhr1k_TLqsKv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=64):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    x = layers.Dense(64, activation=\"relu\")(x)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "    model.compile(optimizer=\"adam\",\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "xhFJ25v2qsKv"
      },
      "execution_count": 144,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Bigram = get_model()\n",
        "model_count_Bigram.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a668bf00-04e2-4ad3-e680-7a63fa9e9650",
        "id": "tG9FqbE3qsKv"
      },
      "execution_count": 145,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_15\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_16 (InputLayer)       [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_40 (Dense)            (None, 64)                64064     \n",
            "                                                                 \n",
            " dropout_23 (Dropout)        (None, 64)                0         \n",
            "                                                                 \n",
            " dense_41 (Dense)            (None, 64)                4160      \n",
            "                                                                 \n",
            " dropout_24 (Dropout)        (None, 64)                0         \n",
            "                                                                 \n",
            " dense_42 (Dense)            (None, 10)                650       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 68,874\n",
            "Trainable params: 68,874\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 146,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c941be28-e294-4d70-ba94-3ed177885272",
        "id": "vxpNVQRkqsKv"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 7s 91ms/step - loss: 173.7345 - accuracy: 0.1846 - val_loss: 12.8603 - val_accuracy: 0.2741\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 24.5644 - accuracy: 0.1905 - val_loss: 2.9692 - val_accuracy: 0.2950\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 4.2929 - accuracy: 0.2332 - val_loss: 2.1566 - val_accuracy: 0.2929\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.1780 - accuracy: 0.2695 - val_loss: 2.0783 - val_accuracy: 0.2845\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.0412 - accuracy: 0.2758 - val_loss: 1.9799 - val_accuracy: 0.2762\n",
            "15/15 [==============================] - 2s 107ms/step - loss: 2.0062 - accuracy: 0.2642\n",
            "Test acc: 0.264\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "     tf.keras.callbacks.ModelCheckpoint(\"binary_2gram.keras\",save_best_only=True)\n",
        "    ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "history = model_count_Bigram.fit(count_2gram_train_ds.cache(),\n",
        "          validation_data=count_2gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_count_Bigram = keras.models.load_model(\"binary_2gram.keras\")\n",
        "print(f\"Test acc: {model_count_Bigram.evaluate(count_2gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_count_Bigram.evaluate(count_2gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ebac4697-0808-4ce5-eaa6-34d604abe533",
        "id": "De2FYeY8qsKv"
      },
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 61ms/step - loss: 2.0062 - accuracy: 0.2642\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2.006181240081787, 0.2641509473323822]"
            ]
          },
          "metadata": {},
          "execution_count": 147
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 275
        },
        "outputId": "3aaa093c-0099-4022-e0a9-5fd8da6ae50e",
        "id": "N5ar1beAqsKv"
      },
      "execution_count": 148,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_count_Bigram.predict(count_2gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dadf313e-5928-44bb-8320-25b7f126ff45",
        "id": "aVYY3l0vqsKw"
      },
      "execution_count": 149,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 59ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b29fd24a-66b9-4758-9c84-33d17547f654",
        "id": "9cbvxHoVqsKw"
      },
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.00      0.00      0.00        17\n",
            "           4       0.00      0.00      0.00        39\n",
            "           5       0.00      0.00      0.00        78\n",
            "           6       0.26      0.99      0.42       124\n",
            "           7       0.20      0.02      0.03       130\n",
            "           8       1.00      0.01      0.03        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.26       477\n",
            "   macro avg       0.16      0.11      0.05       477\n",
            "weighted avg       0.27      0.26      0.12       477\n",
            "\n",
            "Accuracy Score: 0.264\n",
            "Mean Absolute Error: 1.132\n",
            "Root Mean Square Error: 1.484\n",
            "Mean Square Error: 2.201\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "04a3b96a-0a92-4a93-f838-5df503d333e6",
        "id": "9Wk8TT_vqsKw"
      },
      "execution_count": 151,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C3xAe8fU5nVh"
      },
      "source": [
        "# Bigrams with TF-IDF Encoding"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VX5RoLxB5nVh"
      },
      "source": [
        "## Configuring `TextVectorization` To Return TF-IDF-weighted Outputs"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 152,
      "metadata": {
        "id": "nVBOiZJ45nVi"
      },
      "outputs": [],
      "source": [
        "text_vectorization = TextVectorization(\n",
        "    ngrams=2,\n",
        "    max_tokens=1000,\n",
        "    output_mode=\"tf_idf\",\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 153,
      "metadata": {
        "id": "KklLpT7u5nVi"
      },
      "outputs": [],
      "source": [
        "text_vectorization.adapt(text_only_train_ds)\n",
        "\n",
        "tfidf_2gram_train_ds = train_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "tfidf_2gram_val_ds = val_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)\n",
        "tfidf_2gram_test_ds = test_ds.map(\n",
        "    lambda x, y: (text_vectorization(x), y),\n",
        "    num_parallel_calls=4)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hjbUrsST5nVi"
      },
      "source": [
        "## Build TF-IDF Bigram Model"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=128):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "    model.compile(optimizer=\"adam\",\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "20WmcowpwZq3"
      },
      "execution_count": 154,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_tfidf = get_model()\n",
        "model_tfidf.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8sVeW-3NwphY",
        "outputId": "0dc5fadb-b42b-4481-9ea8-76f8a6079edd"
      },
      "execution_count": 155,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_16\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_17 (InputLayer)       [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_43 (Dense)            (None, 128)               128128    \n",
            "                                                                 \n",
            " dropout_25 (Dropout)        (None, 128)               0         \n",
            "                                                                 \n",
            " dense_44 (Dense)            (None, 10)                1290      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 129,418\n",
            "Trainable params: 129,418\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 156,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tdz6qgHv5nVi",
        "outputId": "7b71318c-900a-4dcc-b49e-6e9dcca97305"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 8s 107ms/step - loss: 158.2374 - accuracy: 0.2210 - val_loss: 33.2647 - val_accuracy: 0.2029\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 62.1216 - accuracy: 0.2538 - val_loss: 12.7139 - val_accuracy: 0.3536\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 15.8303 - accuracy: 0.3320 - val_loss: 3.2128 - val_accuracy: 0.3013\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.1605 - accuracy: 0.3226 - val_loss: 1.7167 - val_accuracy: 0.3494\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.8367 - accuracy: 0.3221 - val_loss: 1.6317 - val_accuracy: 0.3766\n",
            "Epoch 6/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7729 - accuracy: 0.3392 - val_loss: 1.5517 - val_accuracy: 0.3849\n",
            "Epoch 7/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6876 - accuracy: 0.3805 - val_loss: 1.5318 - val_accuracy: 0.3975\n",
            "Epoch 8/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6663 - accuracy: 0.3886 - val_loss: 1.4890 - val_accuracy: 0.4079\n",
            "Epoch 9/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6176 - accuracy: 0.4075 - val_loss: 1.4766 - val_accuracy: 0.4456\n",
            "Epoch 10/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5437 - accuracy: 0.4245 - val_loss: 1.4109 - val_accuracy: 0.4205\n",
            "Epoch 11/20\n",
            "70/70 [==============================] - 0s 4ms/step - loss: 1.5094 - accuracy: 0.4164 - val_loss: 1.3799 - val_accuracy: 0.4477\n",
            "Epoch 12/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5005 - accuracy: 0.4182 - val_loss: 1.5638 - val_accuracy: 0.3431\n",
            "Epoch 13/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6088 - accuracy: 0.3881 - val_loss: 1.5297 - val_accuracy: 0.3661\n",
            "Epoch 14/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.5867 - accuracy: 0.3949 - val_loss: 1.4625 - val_accuracy: 0.4247\n",
            "15/15 [==============================] - 1s 59ms/step - loss: 1.5062 - accuracy: 0.4172\n",
            "Test acc: 0.417\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "    tf.keras.callbacks.ModelCheckpoint(\"tfidf_2gram.keras\",save_best_only=True)\n",
        "   ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "history = model_tfidf.fit(tfidf_2gram_train_ds.cache(),\n",
        "          validation_data=tfidf_2gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_tfidf = keras.models.load_model(\"tfidf_2gram.keras\")\n",
        "print(f\"Test acc: {model_tfidf.evaluate(tfidf_2gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_tfidf.evaluate(tfidf_2gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e7876172-c31a-4dea-dc6c-1e5836c57778",
        "id": "6WytD1kTgPt2"
      },
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 59ms/step - loss: 1.5062 - accuracy: 0.4172\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.5061997175216675, 0.4171907901763916]"
            ]
          },
          "metadata": {},
          "execution_count": 157
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "outputId": "9c01a154-1bec-491f-f170-f077a951b18b",
        "id": "1mkg4PzLgPt3"
      },
      "execution_count": 158,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_tfidf.predict(tfidf_2gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "c56b796f-de9d-41bf-8c0a-1c542bc2b2d6",
        "id": "K0V7W9GIgPt3"
      },
      "execution_count": 159,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 59ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a37566c5-19c2-4fb3-fcd9-9f374dff7c8e",
        "id": "XArGy0lSgPt3"
      },
      "execution_count": 160,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.47      0.41      0.44        17\n",
            "           4       0.55      0.15      0.24        39\n",
            "           5       0.20      0.13      0.16        78\n",
            "           6       0.40      0.80      0.53       124\n",
            "           7       0.48      0.38      0.42       130\n",
            "           8       0.54      0.40      0.46        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.42       477\n",
            "   macro avg       0.29      0.25      0.25       477\n",
            "weighted avg       0.41      0.42      0.38       477\n",
            "\n",
            "Accuracy Score: 0.417\n",
            "Mean Absolute Error: 0.711\n",
            "Root Mean Square Error: 1.017\n",
            "Mean Square Error: 1.034\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "8c7d3781-90cc-4537-8581-a1f6ec306201",
        "id": "JzGB7_IVgPt3"
      },
      "execution_count": 161,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2 layer model"
      ],
      "metadata": {
        "id": "XV96b9vskosE"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def get_model(max_tokens=1000, hidden_dim=64):\n",
        "    inputs = tf.keras.Input(shape=(max_tokens,))\n",
        "    x = layers.Dense(hidden_dim, activation=\"relu\")(inputs)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    x = layers.Dense(64, activation=\"relu\")(x)\n",
        "    x = layers.Dropout(0.2)(x)\n",
        "    outputs = layers.Dense(10, activation=\"softmax\")(x)\n",
        "    model = tf.keras.Model(inputs, outputs)\n",
        "    model.compile(optimizer=\"adam\",\n",
        "                  loss='SparseCategoricalCrossentropy',\n",
        "                  metrics=[\"accuracy\"])\n",
        "    return model"
      ],
      "metadata": {
        "id": "GqtFhZ4RkqMG"
      },
      "execution_count": 162,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_tfidf = get_model()\n",
        "model_tfidf.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "287a1bf7-4a4f-409f-e20c-08c8d8d9cbb1",
        "id": "zb9KKqrskqMH"
      },
      "execution_count": 163,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_17\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_18 (InputLayer)       [(None, 1000)]            0         \n",
            "                                                                 \n",
            " dense_45 (Dense)            (None, 64)                64064     \n",
            "                                                                 \n",
            " dropout_26 (Dropout)        (None, 64)                0         \n",
            "                                                                 \n",
            " dense_46 (Dense)            (None, 64)                4160      \n",
            "                                                                 \n",
            " dropout_27 (Dropout)        (None, 64)                0         \n",
            "                                                                 \n",
            " dense_47 (Dense)            (None, 10)                650       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 68,874\n",
            "Trainable params: 68,874\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 164,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4faa4584-6b98-41f1-afa8-33b459358046",
        "id": "AMsxO4t1k0u0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "70/70 [==============================] - 6s 84ms/step - loss: 174.5909 - accuracy: 0.1730 - val_loss: 6.2704 - val_accuracy: 0.1715\n",
            "Epoch 2/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 16.8712 - accuracy: 0.2044 - val_loss: 1.9775 - val_accuracy: 0.1987\n",
            "Epoch 3/20\n",
            "70/70 [==============================] - 0s 4ms/step - loss: 7.2762 - accuracy: 0.2363 - val_loss: 1.8451 - val_accuracy: 0.2636\n",
            "Epoch 4/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 4.4578 - accuracy: 0.2376 - val_loss: 1.8965 - val_accuracy: 0.2908\n",
            "Epoch 5/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 3.2828 - accuracy: 0.2215 - val_loss: 1.9951 - val_accuracy: 0.2720\n",
            "Epoch 6/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.5323 - accuracy: 0.2372 - val_loss: 1.9300 - val_accuracy: 0.3054\n",
            "Epoch 7/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 2.0314 - accuracy: 0.2686 - val_loss: 2.0147 - val_accuracy: 0.2657\n",
            "Epoch 8/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.9351 - accuracy: 0.2794 - val_loss: 1.7895 - val_accuracy: 0.3305\n",
            "Epoch 9/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.8506 - accuracy: 0.2934 - val_loss: 1.8258 - val_accuracy: 0.3117\n",
            "Epoch 10/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.8194 - accuracy: 0.2754 - val_loss: 1.7268 - val_accuracy: 0.3410\n",
            "Epoch 11/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7797 - accuracy: 0.2960 - val_loss: 1.7161 - val_accuracy: 0.3368\n",
            "Epoch 12/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7524 - accuracy: 0.3122 - val_loss: 1.6475 - val_accuracy: 0.3682\n",
            "Epoch 13/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7076 - accuracy: 0.3221 - val_loss: 1.6117 - val_accuracy: 0.3515\n",
            "Epoch 14/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.7008 - accuracy: 0.3163 - val_loss: 1.6067 - val_accuracy: 0.3452\n",
            "Epoch 15/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6639 - accuracy: 0.3306 - val_loss: 1.5582 - val_accuracy: 0.3724\n",
            "Epoch 16/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6725 - accuracy: 0.3374 - val_loss: 1.5551 - val_accuracy: 0.3410\n",
            "Epoch 17/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6653 - accuracy: 0.3495 - val_loss: 1.6928 - val_accuracy: 0.2971\n",
            "Epoch 18/20\n",
            "70/70 [==============================] - 0s 3ms/step - loss: 1.6380 - accuracy: 0.3621 - val_loss: 1.5223 - val_accuracy: 0.3598\n",
            "15/15 [==============================] - 1s 57ms/step - loss: 1.5767 - accuracy: 0.3438\n",
            "Test acc: 0.344\n"
          ]
        }
      ],
      "source": [
        "callbacks = [\n",
        "    tf.keras.callbacks.ModelCheckpoint(\"tfidf_2gram.keras\",save_best_only=True)\n",
        "   ,tf.keras.callbacks.EarlyStopping(monitor='val_accuracy', patience=3)\n",
        "]\n",
        "\n",
        "history = model_tfidf.fit(tfidf_2gram_train_ds.cache(),\n",
        "          validation_data=tfidf_2gram_val_ds.cache(),\n",
        "          epochs=20,\n",
        "          callbacks=callbacks)\n",
        "model_tfidf = keras.models.load_model(\"tfidf_2gram.keras\")\n",
        "print(f\"Test acc: {model_tfidf.evaluate(tfidf_2gram_test_ds)[1]:.3f}\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_tfidf.evaluate(tfidf_2gram_test_ds)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b3b832cf-600d-45d3-c1c5-9a3b4188bb31",
        "id": "9gSZ0hu-k0u1"
      },
      "execution_count": 165,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 61ms/step - loss: 1.5767 - accuracy: 0.3438\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1.5767172574996948, 0.34381550550460815]"
            ]
          },
          "metadata": {},
          "execution_count": 165
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot_history(history)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 271
        },
        "outputId": "8a9261fa-b9e5-4973-a2b6-5930c2012879",
        "id": "XtrL-K4Vk0u1"
      },
      "execution_count": 166,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1152x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pred1= model_tfidf.predict(tfidf_2gram_test_ds)\n",
        "pred1=np.argmax(pred1, axis=1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f27d86aa-8698-41c4-d2b2-7adfd41b550c",
        "id": "wn-eyqcCk0u2"
      },
      "execution_count": 167,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "15/15 [==============================] - 1s 59ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print_validation_report(test_y, pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "492fd576-44e3-46a8-d2d9-c37e2f4defc3",
        "id": "G147ux5nk0u2"
      },
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.00      0.00      0.00         2\n",
            "           2       0.00      0.00      0.00         8\n",
            "           3       0.00      0.00      0.00        17\n",
            "           4       0.00      0.00      0.00        39\n",
            "           5       0.00      0.00      0.00        78\n",
            "           6       0.28      0.60      0.38       124\n",
            "           7       0.41      0.54      0.47       130\n",
            "           8       0.45      0.29      0.35        70\n",
            "           9       0.00      0.00      0.00         9\n",
            "\n",
            "    accuracy                           0.34       477\n",
            "   macro avg       0.13      0.16      0.13       477\n",
            "weighted avg       0.25      0.34      0.28       477\n",
            "\n",
            "Accuracy Score: 0.344\n",
            "Mean Absolute Error: 0.920\n",
            "Root Mean Square Error: 1.299\n",
            "Mean Square Error: 1.688\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_classification.py:1318: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set(font_scale=1.2)\n",
        "plot_confusion_matrix(test_y,pred1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 506
        },
        "outputId": "0b76d8a8-a91c-4cb8-dbed-be23ab3c294a",
        "id": "VX38WmBak0u2"
      },
      "execution_count": 169,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 576x576 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    }
  ]
}