ChatGPT_Project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyP7W06oZtRERFrcmB+rif3l",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/IvyWang845/ChatGPT-Project/blob/main/ChatGPT_Project.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ef9U24pUIPco"
      },
      "outputs": [],
      "source": [
        "!pip install -U -q PyDrive\n",
        "\n",
        "from pydrive.auth import GoogleAuth\n",
        "from pydrive.drive import GoogleDrive\n",
        "from google.colab import auth\n",
        "from oauth2client.client import GoogleCredentials\n",
        "\n",
        "auth.authenticate_user()\n",
        "gauth = GoogleAuth()\n",
        "gauth.credentials = GoogleCredentials.get_application_default()\n",
        "drive = GoogleDrive(gauth)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "file = drive.CreateFile({'id':'1WrPLHqA4FTOTUZAorV-RCPcmg0A0CEib'}) # replace the id with id of file you want to access https://drive.google.com/file/d/1WrPLHqA4FTOTUZAorV-RCPcmg0A0CEib/view?usp=share_link\n",
        "file.GetContentFile('Biomass_data.txt')  # tab-separated"
      ],
      "metadata": {
        "id": "QCeolG0cIl03"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 1: Load Data"
      ],
      "metadata": {
        "id": "LXJ3JCDxJF17"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import nltk\n",
        "# import gensim\n",
        "\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Bjl8fhO0JOuy",
        "outputId": "2cf9b3e2-ea6b-4091-8eb7-793e29878746"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n",
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 63
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load data into dataframe\n",
        "df = pd.read_csv('Biomass_data.txt', encoding= 'unicode_escape', sep='\\t', error_bad_lines=False)"
      ],
      "metadata": {
        "id": "ovuhgKcxJUei",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a1d007e3-e7e8-430a-b166-505636a5fd0c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/IPython/core/interactiveshell.py:3326: FutureWarning: The error_bad_lines argument has been deprecated and will be removed in a future version.\n",
            "\n",
            "\n",
            "  exec(code_obj, self.user_global_ns, self.user_ns)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df.isna().sum()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FXihg5GvJbWl",
        "outputId": "7a33749a-e96a-4601-cfd6-f1233ad31370"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Type         0\n",
              "Type_num     0\n",
              "Responses    0\n",
              "DESPC        0\n",
              "DESSC        0\n",
              "            ..\n",
              "WRDHYPv      0\n",
              "WRDHYPnv     0\n",
              "RDFRE        0\n",
              "RDFKGL       0\n",
              "RDL2         0\n",
              "Length: 111, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# generate text dataset\n",
        "TF_data = df.loc[:, 'Responses'].tolist()"
      ],
      "metadata": {
        "id": "2moqGn5eJh9u"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 2: Tokenizing and Stemming"
      ],
      "metadata": {
        "id": "8rM93qzQKI3y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# Use nltk's English stopwords.\n",
        "stopwords = nltk.corpus.stopwords.words('english') #stopwords.append(\"n't\")\n",
        "stopwords.append(\"'s\")\n",
        "stopwords.append(\"'m\")\n",
        "stopwords.append(\"br\") #html <br>\n",
        "\n",
        "print (\"We use \" + str(len(stopwords)) + \" stop-words from nltk library.\")\n",
        "print (stopwords[:10])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cvOC1rwyKNFj",
        "outputId": "45d86fdc-f63f-4894-c5fe-508137422c4d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "We use 182 stop-words from nltk library.\n",
            "['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', \"you're\"]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from nltk.stem.snowball import SnowballStemmer\n",
        "# from nltk.stem import WordNetLemmatizer\n",
        "\n",
        "stemmer = SnowballStemmer(\"english\")\n",
        "\n",
        "# tokenization and stemming\n",
        "def tokenization_and_stemming(text):\n",
        "    tokens = []\n",
        "    # exclude stop words and tokenize the document, generate a list of string\n",
        "    for word in nltk.word_tokenize(text):\n",
        "        if word.lower() not in stopwords:\n",
        "            tokens.append(word.lower())\n",
        "\n",
        "    filtered_tokens = []\n",
        "\n",
        "    # filter out any tokens not containing letters (e.g., numeric tokens, raw punctuation)\n",
        "    for token in tokens:\n",
        "        if token.isalpha():\n",
        "            filtered_tokens.append(token)\n",
        "\n",
        "    # stemming\n",
        "    stems = [stemmer.stem(t) for t in filtered_tokens]\n",
        "    return stems"
      ],
      "metadata": {
        "id": "ZaFpQ2blKPvw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tokenization_and_stemming(TF_data[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aqEBTJnPKRjm",
        "outputId": "a1eac2a9-09ce-46cb-a8c7-6c64a5222682"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['increas',\n",
              " 'water',\n",
              " 'demand',\n",
              " 'corn',\n",
              " 'requir',\n",
              " 'water',\n",
              " 'grow',\n",
              " 'compar',\n",
              " 'bean',\n",
              " 'shift',\n",
              " 'crop',\n",
              " 'could',\n",
              " 'result',\n",
              " 'increas',\n",
              " 'water',\n",
              " 'usag',\n",
              " 'put',\n",
              " 'pressur',\n",
              " 'limit',\n",
              " 'water',\n",
              " 'sourc',\n",
              " 'area']"
            ]
          },
          "metadata": {},
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "TF_data[0]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        },
        "id": "UbjjX-RUKVIw",
        "outputId": "4d07a395-6962-4682-c9f8-0ccabcd5d3a8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Increased water demand: As corn requires more water to grow compared to beans, the shift in crops could result in an increase in water usage, putting pressure on the limited water sources in the area.'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# TF-IDF"
      ],
      "metadata": {
        "id": "XPPk4DpwKYZk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "# define vectorizer parameters\n",
        "# TfidfVectorizer will help us to create tf-idf matrix\n",
        "# max_df : maximum document frequency for the given word, don't keep when a word appears in the documents at this frequency rate\n",
        "# min_df : minimum document frequency for the given word, don't keep when a word appears in the documents at this frequency rate\n",
        "# max_features: maximum number of words\n",
        "# use_idf: if not true, we only calculate tf\n",
        "# stop_words : built-in stop words\n",
        "# tokenizer: how to tokenize the document\n",
        "# ngram_range: (min_value, max_value), eg. (1, 3) means the result will include 1-gram, 2-gram, 3-gram\n",
        "tfidf_model = TfidfVectorizer(max_df=0.99, max_features=1000,\n",
        "                                 min_df=0.1, stop_words='english',\n",
        "                                 use_idf=True, tokenizer=tokenization_and_stemming, ngram_range=(1,1))\n",
        "\n",
        "tfidf_matrix = tfidf_model.fit_transform(TF_data) #fit the vectorizer to synopses\n",
        "\n",
        "print (\"In total, there are \" + str(tfidf_matrix.shape[0]) + \\\n",
        "      \" reviews and \" + str(tfidf_matrix.shape[1]) + \" terms.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CS7nuEzMKbrz",
        "outputId": "011e92a2-4da2-4ce3-8d07-815e930d0298"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/sklearn/feature_extraction/text.py:396: UserWarning: Your stop_words may be inconsistent with your preprocessing. Tokenizing the stop words generated tokens ['afterward', 'alon', 'alreadi', 'alway', 'anoth', 'anyon', 'anyth', 'anywher', 'becam', 'becom', 'besid', 'cri', 'describ', 'els', 'elsewher', 'empti', 'everi', 'everyon', 'everyth', 'everywher', 'fifti', 'forti', 'henc', 'hereaft', 'herebi', 'howev', 'hundr', 'inde', 'mani', 'meanwhil', 'moreov', 'nobodi', 'noon', 'noth', 'nowher', 'otherwis', 'perhap', 'pleas', 'sever', 'sinc', 'sincer', 'sixti', 'someon', 'someth', 'sometim', 'somewher', 'thenc', 'thereaft', 'therebi', 'therefor', 'togeth', 'twelv', 'twenti', 'whatev', 'whenc', 'whenev', 'wherea', 'whereaft', 'wherebi', 'wherev'] not in stop_words.\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "In total, there are 611 reviews and 30 terms.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_matrix"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gq_rx45XKgfu",
        "outputId": "59a6a966-d280-493e-97b0-c532f641a96d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<611x30 sparse matrix of type '<class 'numpy.float64'>'\n",
              "\twith 4135 stored elements in Compressed Sparse Row format>"
            ]
          },
          "metadata": {},
          "execution_count": 72
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_matrix.toarray() #todense() return a copy, not inplace"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "cjTDYTIaKkuc",
        "outputId": "7a67bdf2-2c65-4d29-f8bd-569eaec30ab5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.        , 0.22559299, 0.2074027 , ..., 0.28731548, 0.        ,\n",
              "        0.44145075],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.24175587,\n",
              "        0.62263586],\n",
              "       [0.        , 0.        , 0.        , ..., 0.35729612, 0.42630894,\n",
              "        0.5489737 ],\n",
              "       ...,\n",
              "       [0.        , 0.        , 0.16996528, ..., 0.        , 0.14046592,\n",
              "        0.09044157],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        0.        ],\n",
              "       [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "        1.        ]])"
            ]
          },
          "metadata": {},
          "execution_count": 73
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_matrix.todense()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dXuVdzSHKmI6",
        "outputId": "6f14b76b-7a2b-4bae-9360-2df1a060300e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "matrix([[0.        , 0.22559299, 0.2074027 , ..., 0.28731548, 0.        ,\n",
              "         0.44145075],\n",
              "        [0.        , 0.        , 0.        , ..., 0.        , 0.24175587,\n",
              "         0.62263586],\n",
              "        [0.        , 0.        , 0.        , ..., 0.35729612, 0.42630894,\n",
              "         0.5489737 ],\n",
              "        ...,\n",
              "        [0.        , 0.        , 0.16996528, ..., 0.        , 0.14046592,\n",
              "         0.09044157],\n",
              "        [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "         0.        ],\n",
              "        [0.        , 0.        , 0.        , ..., 0.        , 0.        ,\n",
              "         1.        ]])"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_matrix"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1Q29bQ1WMlH8",
        "outputId": "a1aa754b-433a-43f8-fc0d-c529ede83e0c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<611x30 sparse matrix of type '<class 'numpy.float64'>'\n",
              "\twith 4135 stored elements in Compressed Sparse Row format>"
            ]
          },
          "metadata": {},
          "execution_count": 75
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# words\n",
        "tf_selected_words = tfidf_model.get_feature_names()"
      ],
      "metadata": {
        "id": "x4Q9B0IWKoQi",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f109669d-400f-4ea0-94bd-a430ce4d5a91"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/sklearn/utils/deprecation.py:87: FutureWarning: Function get_feature_names is deprecated; get_feature_names is deprecated in 1.0 and will be removed in 1.2. Please use get_feature_names_out instead.\n",
            "  warnings.warn(msg, category=FutureWarning)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tf_selected_words"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pFte2tsOKq53",
        "outputId": "59ef68ae-632a-4bf2-f0f1-837071b36a60"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['agricultur',\n",
              " 'area',\n",
              " 'bean',\n",
              " 'biomass',\n",
              " 'chang',\n",
              " 'communiti',\n",
              " 'corn',\n",
              " 'crop',\n",
              " 'decreas',\n",
              " 'energi',\n",
              " 'farm',\n",
              " 'food',\n",
              " 'gas',\n",
              " 'grow',\n",
              " 'increas',\n",
              " 'limit',\n",
              " 'mean',\n",
              " 'natur',\n",
              " 'need',\n",
              " 'produc',\n",
              " 'product',\n",
              " 'requir',\n",
              " 'result',\n",
              " 'river',\n",
              " 'sourc',\n",
              " 'switch',\n",
              " 'tradeoff',\n",
              " 'usag',\n",
              " 'use',\n",
              " 'water']"
            ]
          },
          "metadata": {},
          "execution_count": 77
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 3: Log Regression Model Training and Result Evaluation"
      ],
      "metadata": {
        "id": "3pmR2mHSK5U9"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_tf = df[\"Type_num\"]\n",
        "X_tf = tfidf_matrix"
      ],
      "metadata": {
        "id": "0pcsLLJTLpD_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_tf"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1tErvgq0ioIU",
        "outputId": "49a3ef95-3806-41aa-f583-2682bab5e43c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<611x30 sparse matrix of type '<class 'numpy.float64'>'\n",
              "\twith 4135 stored elements in Compressed Sparse Row format>"
            ]
          },
          "metadata": {},
          "execution_count": 79
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import model_selection\n",
        "X_train, X_test, y_train, y_test = model_selection.train_test_split(X_tf, y_tf, test_size=0.25, stratify = y_tf, random_state=1) #stratified sampling,\n",
        "\n",
        "print('training data has ' + str(X_train.shape[0]) + ' observation with ' + str(X_train.shape[1]) + ' terms')\n",
        "print('test data has ' + str(X_test.shape[0]) + ' observation with ' + str(X_test.shape[1]) + ' terms')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "km24kbw6Ld6r",
        "outputId": "23120169-6735-4055-c47a-c9d1958236db"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "training data has 458 observation with 30 terms\n",
            "test data has 153 observation with 30 terms\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Part 3.1: Model Training"
      ],
      "metadata": {
        "id": "sXoNtTUpLFV0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "# Logistic Regression\n",
        "classifier_logistic = LogisticRegression()"
      ],
      "metadata": {
        "id": "uPlG8Mj0Kuhz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "classifier_logistic.fit(X_train, y_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "g_nuq3mFLSXl",
        "outputId": "f2b234e1-a5c6-46de-b409-c0aa927988c8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 82
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic.predict_proba(X_test)[:,0] #level 1 prob"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CUMM1KdmPAHW",
        "outputId": "9bb81538-493a-40d3-cc03-f2b4c32489e0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.9129877 , 0.05960471, 0.05804984, 0.01423298, 0.03303313,\n",
              "       0.03507937, 0.24120611, 0.72215986, 0.0284623 , 0.47378221,\n",
              "       0.01751924, 0.41360503, 0.0159542 , 0.03555594, 0.02575838,\n",
              "       0.018158  , 0.00720061, 0.16738899, 0.01906898, 0.02802304,\n",
              "       0.16589622, 0.2095208 , 0.03793172, 0.05568895, 0.57117665,\n",
              "       0.0343285 , 0.07979531, 0.03253936, 0.30507165, 0.08032346,\n",
              "       0.10348613, 0.03889204, 0.03848769, 0.02911759, 0.39525437,\n",
              "       0.04576784, 0.01849791, 0.02780793, 0.01309061, 0.01994023,\n",
              "       0.09166682, 0.02266424, 0.01411052, 0.05025222, 0.07363688,\n",
              "       0.04707441, 0.04798437, 0.02225059, 0.02484133, 0.02371816,\n",
              "       0.00786931, 0.02287675, 0.02465188, 0.05020825, 0.0764967 ,\n",
              "       0.05430258, 0.02583236, 0.07669711, 0.16819243, 0.05523721,\n",
              "       0.15986794, 0.09978955, 0.13494382, 0.60980693, 0.04817296,\n",
              "       0.24946839, 0.025329  , 0.07748808, 0.05345116, 0.04548723,\n",
              "       0.02912163, 0.02970459, 0.04784498, 0.24674055, 0.20269435,\n",
              "       0.11277929, 0.113427  , 0.01157525, 0.17236241, 0.16630415,\n",
              "       0.3866825 , 0.0857087 , 0.03411813, 0.01028222, 0.16466155,\n",
              "       0.01431575, 0.2591523 , 0.21898805, 0.0389018 , 0.79398579,\n",
              "       0.05701139, 0.07100268, 0.04287349, 0.01345881, 0.09868799,\n",
              "       0.03457706, 0.05742537, 0.02026835, 0.21090532, 0.07501327,\n",
              "       0.89093074, 0.01568081, 0.04223272, 0.07434537, 0.06132996,\n",
              "       0.87730892, 0.02359097, 0.7916078 , 0.04277538, 0.02796634,\n",
              "       0.01692916, 0.02166307, 0.02579786, 0.07735482, 0.01640891,\n",
              "       0.05961303, 0.13740418, 0.02139958, 0.04021478, 0.08747784,\n",
              "       0.29752992, 0.04079218, 0.0564331 , 0.46115928, 0.02944699,\n",
              "       0.22233293, 0.03647126, 0.01647396, 0.01459183, 0.00621707,\n",
              "       0.06219731, 0.22609478, 0.08075984, 0.01313304, 0.02879949,\n",
              "       0.08315151, 0.01967898, 0.15635256, 0.62997806, 0.60476362,\n",
              "       0.06875605, 0.19001772, 0.02207224, 0.00919395, 0.02266424,\n",
              "       0.04874439, 0.03500698, 0.2465715 , 0.14161151, 0.2265924 ,\n",
              "       0.01550852, 0.01314351, 0.27663349])"
            ]
          },
          "metadata": {},
          "execution_count": 83
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic.score(X_test, y_test) #accuracy"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Fu0zB0ffPLNr",
        "outputId": "23635c33-b554-4c28-b475-a21f81eba782"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9084967320261438"
            ]
          },
          "metadata": {},
          "execution_count": 84
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_axis = classifier_logistic.predict_proba(X_tf)[:,0]"
      ],
      "metadata": {
        "id": "lvg9yNONiMCS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import roc_auc_score\n",
        "auc = roc_auc_score(y_test, classifier_logistic.predict_proba(X_test)[:,1])\n",
        "auc"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sc4NH2w21iQX",
        "outputId": "4e8fba7d-cac7-4933-8f99-37b47b877332"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.938345864661654"
            ]
          },
          "metadata": {},
          "execution_count": 86
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics import roc_curve, auc\n",
        "def Find_Optimal_Cutoff(target, predicted):\n",
        "  fpr, tpr, threshold = roc_curve(y_test, classifier_logistic.predict_proba(X_test)[:,1])\n",
        "  i = np.arange(len(tpr))\n",
        "  roc = pd.DataFrame({'tf' : pd.Series(tpr-(1-fpr), index=i), 'threshold' : pd.Series(threshold, index=i)})\n",
        "  roc_t = roc.iloc[(roc.tf-0).abs().argsort()[:1]]\n",
        "  return list(roc_t['threshold'])"
      ],
      "metadata": {
        "id": "qaOGybFf4KBH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold = Find_Optimal_Cutoff(y_test, classifier_logistic.predict_proba(X_test)[:,0])\n",
        "threshold"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0CMwg7NQDTHm",
        "outputId": "3fd7336f-c14a-4036-f5ff-f3cfae38e17c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.8326110110038114]"
            ]
          },
          "metadata": {},
          "execution_count": 88
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def to_labels(pos_probs, threshold):\n",
        " return (pos_probs >= threshold).astype('int')\n",
        "a = to_labels(classifier_logistic.predict_proba(X_test)[:,1],0.832)\n",
        "\n",
        "from sklearn.metrics import precision_score, recall_score, f1_score, accuracy_score\n",
        "accuracy_score(y_test, a)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vkcCClk1En6u",
        "outputId": "e44f575f-fc91-45ad-810a-1c83fb7765d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8627450980392157"
            ]
          },
          "metadata": {},
          "execution_count": 89
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "precision_score(y_test, a)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SuzfHUEZJqli",
        "outputId": "34d87962-119e-48c8-a788-e5304152127b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9666666666666667"
            ]
          },
          "metadata": {},
          "execution_count": 90
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "recall_score(y_test, a)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CgGG472LG3GE",
        "outputId": "41260c7d-1103-4ab1-dbc2-fc59f443a218"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.8721804511278195"
            ]
          },
          "metadata": {},
          "execution_count": 91
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "f1_score(y_test, a)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_TeZ2dHBJuka",
        "outputId": "5c2d91c4-20bb-456a-8ba5-c1c7b82b124c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.91699604743083"
            ]
          },
          "metadata": {},
          "execution_count": 92
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 4: Coh-metrix Features Random Forest Select Important Features"
      ],
      "metadata": {
        "id": "zNXn8GNCV6Xf"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "to_drop = [\"Type\", \"Type_num\", \"Responses\"]\n",
        "y_cm = df[\"Type_num\"]\n",
        "X_cm = df.drop(to_drop, axis=1)"
      ],
      "metadata": {
        "id": "GDjQijYmWnIC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import model_selection\n",
        "X_train_cm, X_test_cm, y_train_cm, y_test_cm = model_selection.train_test_split(X_cm, y_cm, test_size=0.25, stratify = y_cm, random_state=1) #stratified sampling,\n",
        "\n",
        "print('training data has ' + str(X_train_cm.shape[0]) + ' observation with ' + str(X_train_cm.shape[1]) + ' features')\n",
        "print('test data has ' + str(X_test_cm.shape[0]) + ' observation with ' + str(X_test_cm.shape[1]) + ' features')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yKz7sULVXpJY",
        "outputId": "60904a7e-7392-4294-bc5d-fbd5786d7479"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "training data has 458 observation with 108 features\n",
            "test data has 153 observation with 108 features\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "# Random Forest\n",
        "classifier_RF = RandomForestClassifier(n_estimators=500,\n",
        "                                random_state=1)\n",
        "classifier_RF.fit(X_train_cm, y_train_cm)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rRAvgIsKWUGi",
        "outputId": "8f44e1e2-6f21-4ffc-9c5c-a5899cbf0ad2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "RandomForestClassifier(n_estimators=500, random_state=1)"
            ]
          },
          "metadata": {},
          "execution_count": 95
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "# helper function for printing out grid search results\n",
        "def print_grid_search_metrics(gs):\n",
        "    print (\"Best score: \" + str(gs.best_score_))\n",
        "    print (\"Best parameters set:\")\n",
        "    best_parameters = gs.best_params_\n",
        "    for param_name in sorted(best_parameters.keys()):\n",
        "        print(param_name + ':' + str(best_parameters[param_name]))"
      ],
      "metadata": {
        "id": "qdyc_rsZYRme"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "importances = classifier_RF.feature_importances_"
      ],
      "metadata": {
        "id": "oc7CxNizYmdZ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "feat_labels"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hRHltdNX_w4x",
        "outputId": "e6452cca-9d10-4793-8e0a-027e58f4809a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Index(['DESPC', 'DESSC', 'DESWC', 'DESPL', 'DESPLd', 'DESSL', 'DESSLd',\n",
              "       'DESWLsy', 'DESWLsyd', 'DESWLlt',\n",
              "       ...\n",
              "       'WRDCNCc', 'WRDIMGc', 'WRDMEAc', 'WRDPOLc', 'WRDHYPn', 'WRDHYPv',\n",
              "       'WRDHYPnv', 'RDFRE', 'RDFKGL', 'RDL2'],\n",
              "      dtype='object', length=108)"
            ]
          },
          "metadata": {},
          "execution_count": 99
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.barh(feat_labels, classifier_RF.feature_importances_)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "-5Qii6PZ_Mkc",
        "outputId": "89954077-788f-4f32-eafe-faeb23f91dd5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<BarContainer object of 108 artists>"
            ]
          },
          "metadata": {},
          "execution_count": 100
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sorted_indices = np.argsort(importances)[::-1]\n",
        "\n",
        "feat_labels = df.columns[3:]\n",
        "\n",
        "for f in range(X_train_cm.shape[1]):\n",
        "    print(\"%2d) %-*s %f\" % (f + 1, 30,\n",
        "                            feat_labels[sorted_indices[f]],\n",
        "                            importances[sorted_indices[f]]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "J_I-d-_qaonL",
        "outputId": "471f0c2f-0c35-4d67-dfa3-7926af13323a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            " 1) WRDFAMc                        0.079821\n",
            " 2) RDFRE                          0.077987\n",
            " 3) WRDFRQc                        0.072434\n",
            " 4) WRDNOUN                        0.063416\n",
            " 5) WRDAOAc                        0.061628\n",
            " 6) WRDHYPnv                       0.057767\n",
            " 7) DESWLlt                        0.051321\n",
            " 8) RDFKGL                         0.047347\n",
            " 9) DESWLsyd                       0.034287\n",
            "10) DESWLsy                        0.031330\n",
            "11) DESWLltd                       0.029975\n",
            "12) DESWC                          0.024928\n",
            "13) DESSL                          0.020304\n",
            "14) SMINTEp                        0.017249\n",
            "15) DRVP                           0.016526\n",
            "16) LDTTRa                         0.015176\n",
            "17) PCSYNz                         0.013556\n",
            "18) PCSYNp                         0.012181\n",
            "19) RDL2                           0.011705\n",
            "20) SMCAUSv                        0.010914\n",
            "21) LDTTRc                         0.010612\n",
            "22) PCVERBp                        0.010380\n",
            "23) WRDFRQa                        0.009267\n",
            "24) SYNLE                          0.008508\n",
            "25) WRDFRQmc                       0.008402\n",
            "26) PCVERBz                        0.007825\n",
            "27) WRDADV                         0.007353\n",
            "28) CNCLogic                       0.006811\n",
            "29) DRPP                           0.006410\n",
            "30) PCCONNp                        0.005940\n",
            "31) WRDHYPn                        0.005793\n",
            "32) CNCAdd                         0.005597\n",
            "33) SYNNP                          0.005394\n",
            "34) SMCAUSvp                       0.005314\n",
            "35) PCNARz                         0.005273\n",
            "36) WRDPOLc                        0.005185\n",
            "37) DRGERUND                       0.004901\n",
            "38) PCTEMPz                        0.004708\n",
            "39) LDMTLD                         0.004651\n",
            "40) WRDCNCc                        0.004354\n",
            "41) PCCONNz                        0.004293\n",
            "42) SYNSTRUTt                      0.004153\n",
            "43) PCNARp                         0.003871\n",
            "44) CNCPos                         0.003677\n",
            "45) PCREFp                         0.003622\n",
            "46) SMCAUSwn                       0.003571\n",
            "47) DRNP                           0.003460\n",
            "48) CNCADC                         0.003393\n",
            "49) SMCAUSlsa                      0.003212\n",
            "50) PCCNCz                         0.003157\n",
            "51) SYNSTRUTa                      0.003110\n",
            "52) PCDCz                          0.003060\n",
            "53) WRDHYPv                        0.003054\n",
            "54) WRDADJ                         0.002918\n",
            "55) LSAGNd                         0.002863\n",
            "56) LSASSp                         0.002803\n",
            "57) WRDPRO                         0.002790\n",
            "58) LSASS1                         0.002752\n",
            "59) PCREFz                         0.002591\n",
            "60) CNCCaus                        0.002562\n",
            "61) PCDCp                          0.002451\n",
            "62) DESSLd                         0.002416\n",
            "63) CNCNeg                         0.002331\n",
            "64) WRDMEAc                        0.002313\n",
            "65) DRINF                          0.002254\n",
            "66) PCCNCp                         0.002247\n",
            "67) PCTEMPp                        0.002193\n",
            "68) WRDIMGc                        0.002182\n",
            "69) DRAP                           0.002176\n",
            "70) SYNMEDwrd                      0.002136\n",
            "71) SYNMEDpos                      0.002098\n",
            "72) CNCAll                         0.001956\n",
            "73) SYNMEDlem                      0.001897\n",
            "74) DESSC                          0.001843\n",
            "75) WRDVERB                        0.001728\n",
            "76) SMCAUSr                        0.001619\n",
            "77) CNCTempx                       0.001577\n",
            "78) LSAGN                          0.001549\n",
            "79) DESPL                          0.001450\n",
            "80) CNCTemp                        0.001375\n",
            "81) DRPVAL                         0.001342\n",
            "82) SMTEMP                         0.001312\n",
            "83) CRFCWO1                        0.001109\n",
            "84) CRFCWOa                        0.001032\n",
            "85) WRDPRP3p                       0.001021\n",
            "86) CRFNO1                         0.000661\n",
            "87) LSASSpd                        0.000630\n",
            "88) DRNEG                          0.000598\n",
            "89) CRFANP1                        0.000527\n",
            "90) CRFNOa                         0.000514\n",
            "91) SMINTEr                        0.000494\n",
            "92) CRFANPa                        0.000464\n",
            "93) CRFCWO1d                       0.000450\n",
            "94) CRFAO1                         0.000438\n",
            "95) LSASS1d                        0.000423\n",
            "96) WRDPRP1s                       0.000385\n",
            "97) CRFAOa                         0.000354\n",
            "98) CRFSO1                         0.000332\n",
            "99) CRFSOa                         0.000265\n",
            "100) CRFCWOad                       0.000238\n",
            "101) WRDPRP2                        0.000111\n",
            "102) WRDPRP1p                       0.000069\n",
            "103) LSAPP1d                        0.000000\n",
            "104) LDVOCD                         0.000000\n",
            "105) DESPLd                         0.000000\n",
            "106) LSAPP1                         0.000000\n",
            "107) WRDPRP3s                       0.000000\n",
            "108) DESPC                          0.000000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "sorted_indices"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NxClVQhVa1LV",
        "outputId": "824a920b-7228-4dbb-b2c3-a83e0b8903ca"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 97, 105,  93,  83,  96, 104,   9, 106,   8,   7,  10,   2,   5,\n",
              "        62,  76,  48,  13,  14, 107,  60,  47,  22,  94,  68,  95,  21,\n",
              "        86,  53,  78,  24, 102,  57,  69,  61,  11, 101,  81,  25,  49,\n",
              "        98,  23,  74,  12,  58,  18,  66,  75,  54,  65,  15,  73,  19,\n",
              "       103,  85,  46,  41,  87,  39,  17,  52,  20,   6,  59, 100,  82,\n",
              "        16,  26,  99,  77,  71,  70,  51,  72,   1,  84,  63,  56,  45,\n",
              "         3,  55,  79,  67,  33,  35,  92,  27,  42,  80,  37,  30,  64,\n",
              "        38,  34,  28,  40,  88,  31,  29,  32,  36,  90,  89,  44,  50,\n",
              "         4,  43,  91,   0])"
            ]
          },
          "metadata": {},
          "execution_count": 102
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 5: Logistic Regression on CM features"
      ],
      "metadata": {
        "id": "ftqpNmXVcLJ8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_selected_cm = df[\"Type_num\"]\n",
        "X_selected_cm = X_cm.loc[:,[\"WRDFAMc\", \"WRDFRQc\", \"WRDNOUN\", \"DESWLlt\", \"WRDHYPnv\", \"DESWLltd\", \"LDTTRa\", \"WRDADV\", \"DESSL\", \"DRVP\", \"DESSLd\", \"SYNLE\", \"SYNSTRUTt\", \"DESWC\", \"SMCAUSv\", \"CNCLogic\", \"PCDCp\"]]"
      ],
      "metadata": {
        "id": "BXA77cYwcGT7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_cm"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 487
        },
        "id": "5O7IE3ZPd5cb",
        "outputId": "633fd9b0-7f34-4e91-867e-45108f3b36ea"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     DESPC  DESSC  DESWC  DESPL  DESPLd  DESSL  DESSLd  DESWLsy  DESWLsyd  \\\n",
              "0      1.0    1.0   35.0    1.0     0.0   35.0   0.000    1.510     0.610   \n",
              "1      1.0    1.0   32.0    1.0     0.0   33.0   0.000    1.970     1.200   \n",
              "2      1.0    1.0   37.0    1.0     0.0   37.0   0.000    1.870     0.980   \n",
              "3      1.0    1.0   34.0    1.0     0.0   34.0   0.000    2.060     1.010   \n",
              "4      1.0    1.0   42.0    1.0     0.0   43.0   0.000    1.950     1.060   \n",
              "..     ...    ...    ...    ...     ...    ...     ...      ...       ...   \n",
              "606    1.0    1.0   15.0    1.0     0.0   15.0   0.000    1.600     0.828   \n",
              "607    1.0    2.0   51.0    2.0     0.0   26.0   7.071    1.392     0.750   \n",
              "608    1.0    2.0   41.0    2.0     0.0   21.0   7.071    1.415     0.805   \n",
              "609    1.0    1.0    3.0    1.0     0.0    3.0   0.000    2.000     1.000   \n",
              "610    1.0    1.0   18.0    1.0     0.0   18.0   0.000    1.389     0.502   \n",
              "\n",
              "     DESWLlt  ...     WRDCNCc     WRDIMGc     WRDMEAc  WRDPOLc  WRDHYPn  \\\n",
              "0      4.630  ...  459.850000  459.570000  457.100000    3.590    6.740   \n",
              "1      5.560  ...  491.910000  469.230000  481.000000    4.210    5.930   \n",
              "2      5.700  ...  417.190000  449.630000  433.780000    3.670    6.790   \n",
              "3      6.210  ...  416.810000  459.560000  521.000000    3.230    6.450   \n",
              "4      6.070  ...  443.560000  444.850000  452.360000    4.180    7.250   \n",
              "..       ...  ...         ...         ...         ...      ...      ...   \n",
              "606    4.867  ...  395.625000  440.625000  521.599976    3.813    6.729   \n",
              "607    4.137  ...  385.614990  398.833008  418.614990    4.618    7.628   \n",
              "608    4.293  ...  481.200012  486.000000  473.071014    4.116    7.105   \n",
              "609    6.000  ...  523.000000  526.000000  508.000000    1.667   10.167   \n",
              "610    4.444  ...  415.375000  443.000000  449.143005    6.000    5.971   \n",
              "\n",
              "     WRDHYPv  WRDHYPnv      RDFRE  RDFKGL    RDL2  \n",
              "0      2.540     2.810  43.230000  15.930   1.870  \n",
              "1      2.210     2.590   7.780000  20.120   0.200  \n",
              "2      0.720     3.280  11.500000  20.850  -1.200  \n",
              "3      2.120     2.940   0.000000  21.970  -5.880  \n",
              "4      1.880     3.590   0.000000  23.820  -0.590  \n",
              "..       ...       ...        ...     ...     ...  \n",
              "606    1.875     2.758  56.250000   9.140   4.330  \n",
              "607    1.225     1.726  63.188999  10.781  12.683  \n",
              "608    1.498     2.714  66.319000   9.102  17.480  \n",
              "609    0.000     6.778  34.590000   9.180 -11.057  \n",
              "610    1.219     1.385  71.056000   7.820  12.055  \n",
              "\n",
              "[611 rows x 108 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-fc403d3d-9612-4b4e-bb96-1549fdd697cd\">\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>DESPC</th>\n",
              "      <th>DESSC</th>\n",
              "      <th>DESWC</th>\n",
              "      <th>DESPL</th>\n",
              "      <th>DESPLd</th>\n",
              "      <th>DESSL</th>\n",
              "      <th>DESSLd</th>\n",
              "      <th>DESWLsy</th>\n",
              "      <th>DESWLsyd</th>\n",
              "      <th>DESWLlt</th>\n",
              "      <th>...</th>\n",
              "      <th>WRDCNCc</th>\n",
              "      <th>WRDIMGc</th>\n",
              "      <th>WRDMEAc</th>\n",
              "      <th>WRDPOLc</th>\n",
              "      <th>WRDHYPn</th>\n",
              "      <th>WRDHYPv</th>\n",
              "      <th>WRDHYPnv</th>\n",
              "      <th>RDFRE</th>\n",
              "      <th>RDFKGL</th>\n",
              "      <th>RDL2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>35.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.510</td>\n",
              "      <td>0.610</td>\n",
              "      <td>4.630</td>\n",
              "      <td>...</td>\n",
              "      <td>459.850000</td>\n",
              "      <td>459.570000</td>\n",
              "      <td>457.100000</td>\n",
              "      <td>3.590</td>\n",
              "      <td>6.740</td>\n",
              "      <td>2.540</td>\n",
              "      <td>2.810</td>\n",
              "      <td>43.230000</td>\n",
              "      <td>15.930</td>\n",
              "      <td>1.870</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>32.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>33.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.970</td>\n",
              "      <td>1.200</td>\n",
              "      <td>5.560</td>\n",
              "      <td>...</td>\n",
              "      <td>491.910000</td>\n",
              "      <td>469.230000</td>\n",
              "      <td>481.000000</td>\n",
              "      <td>4.210</td>\n",
              "      <td>5.930</td>\n",
              "      <td>2.210</td>\n",
              "      <td>2.590</td>\n",
              "      <td>7.780000</td>\n",
              "      <td>20.120</td>\n",
              "      <td>0.200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>37.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>37.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.870</td>\n",
              "      <td>0.980</td>\n",
              "      <td>5.700</td>\n",
              "      <td>...</td>\n",
              "      <td>417.190000</td>\n",
              "      <td>449.630000</td>\n",
              "      <td>433.780000</td>\n",
              "      <td>3.670</td>\n",
              "      <td>6.790</td>\n",
              "      <td>0.720</td>\n",
              "      <td>3.280</td>\n",
              "      <td>11.500000</td>\n",
              "      <td>20.850</td>\n",
              "      <td>-1.200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>34.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>2.060</td>\n",
              "      <td>1.010</td>\n",
              "      <td>6.210</td>\n",
              "      <td>...</td>\n",
              "      <td>416.810000</td>\n",
              "      <td>459.560000</td>\n",
              "      <td>521.000000</td>\n",
              "      <td>3.230</td>\n",
              "      <td>6.450</td>\n",
              "      <td>2.120</td>\n",
              "      <td>2.940</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>21.970</td>\n",
              "      <td>-5.880</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>42.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>43.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.950</td>\n",
              "      <td>1.060</td>\n",
              "      <td>6.070</td>\n",
              "      <td>...</td>\n",
              "      <td>443.560000</td>\n",
              "      <td>444.850000</td>\n",
              "      <td>452.360000</td>\n",
              "      <td>4.180</td>\n",
              "      <td>7.250</td>\n",
              "      <td>1.880</td>\n",
              "      <td>3.590</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>23.820</td>\n",
              "      <td>-0.590</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>15.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.600</td>\n",
              "      <td>0.828</td>\n",
              "      <td>4.867</td>\n",
              "      <td>...</td>\n",
              "      <td>395.625000</td>\n",
              "      <td>440.625000</td>\n",
              "      <td>521.599976</td>\n",
              "      <td>3.813</td>\n",
              "      <td>6.729</td>\n",
              "      <td>1.875</td>\n",
              "      <td>2.758</td>\n",
              "      <td>56.250000</td>\n",
              "      <td>9.140</td>\n",
              "      <td>4.330</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>607</th>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>51.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>26.0</td>\n",
              "      <td>7.071</td>\n",
              "      <td>1.392</td>\n",
              "      <td>0.750</td>\n",
              "      <td>4.137</td>\n",
              "      <td>...</td>\n",
              "      <td>385.614990</td>\n",
              "      <td>398.833008</td>\n",
              "      <td>418.614990</td>\n",
              "      <td>4.618</td>\n",
              "      <td>7.628</td>\n",
              "      <td>1.225</td>\n",
              "      <td>1.726</td>\n",
              "      <td>63.188999</td>\n",
              "      <td>10.781</td>\n",
              "      <td>12.683</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>608</th>\n",
              "      <td>1.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>41.0</td>\n",
              "      <td>2.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>21.0</td>\n",
              "      <td>7.071</td>\n",
              "      <td>1.415</td>\n",
              "      <td>0.805</td>\n",
              "      <td>4.293</td>\n",
              "      <td>...</td>\n",
              "      <td>481.200012</td>\n",
              "      <td>486.000000</td>\n",
              "      <td>473.071014</td>\n",
              "      <td>4.116</td>\n",
              "      <td>7.105</td>\n",
              "      <td>1.498</td>\n",
              "      <td>2.714</td>\n",
              "      <td>66.319000</td>\n",
              "      <td>9.102</td>\n",
              "      <td>17.480</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>609</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>6.000</td>\n",
              "      <td>...</td>\n",
              "      <td>523.000000</td>\n",
              "      <td>526.000000</td>\n",
              "      <td>508.000000</td>\n",
              "      <td>1.667</td>\n",
              "      <td>10.167</td>\n",
              "      <td>0.000</td>\n",
              "      <td>6.778</td>\n",
              "      <td>34.590000</td>\n",
              "      <td>9.180</td>\n",
              "      <td>-11.057</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>610</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>18.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>1.389</td>\n",
              "      <td>0.502</td>\n",
              "      <td>4.444</td>\n",
              "      <td>...</td>\n",
              "      <td>415.375000</td>\n",
              "      <td>443.000000</td>\n",
              "      <td>449.143005</td>\n",
              "      <td>6.000</td>\n",
              "      <td>5.971</td>\n",
              "      <td>1.219</td>\n",
              "      <td>1.385</td>\n",
              "      <td>71.056000</td>\n",
              "      <td>7.820</td>\n",
              "      <td>12.055</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>611 rows × 108 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-fc403d3d-9612-4b4e-bb96-1549fdd697cd')\"\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-fc403d3d-9612-4b4e-bb96-1549fdd697cd 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-fc403d3d-9612-4b4e-bb96-1549fdd697cd');\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": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_selected_cm"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 488
        },
        "id": "v2jRSdL-d8B2",
        "outputId": "638531f4-14c1-4924-f863-a471be549d06"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        WRDFAMc  WRDFRQc     WRDNOUN  DESWLlt  WRDHYPnv  DESWLltd  LDTTRa  \\\n",
              "0    583.860000    2.110  400.000000    4.630     2.810     2.160   0.740   \n",
              "1    583.000000    2.040  406.250000    5.560     2.590     3.010   0.790   \n",
              "2    568.940000    1.970  513.510000    5.700     3.280     3.030   0.700   \n",
              "3    563.440000    1.760  411.760000    6.210     2.940     2.900   0.790   \n",
              "4    556.500000    2.000  500.000000    6.070     3.590     3.200   0.700   \n",
              "..          ...      ...         ...      ...       ...       ...     ...   \n",
              "606  588.125000    2.223  400.000000    4.867     2.758     2.416   0.867   \n",
              "607  580.221985    2.360  215.686996    4.137     1.726     2.333   0.769   \n",
              "608  577.267029    2.221  341.463013    4.293     2.714     2.358   0.738   \n",
              "609  575.500000    1.530  666.666016    6.000     6.778     3.000   1.000   \n",
              "610  593.778015    2.571  222.223007    4.444     1.385     1.947   0.778   \n",
              "\n",
              "        WRDADV  DESSL        DRVP  DESSLd  SYNLE  SYNSTRUTt  DESWC    SMCAUSv  \\\n",
              "0     0.000000   35.0  171.430000   0.000   17.0      0.000   35.0  28.570000   \n",
              "1     0.000000   33.0   93.750000   0.000   10.0      0.000   32.0  31.250000   \n",
              "2     0.000000   37.0   81.080000   0.000   12.0      0.000   37.0  27.030000   \n",
              "3    29.410000   34.0  147.060000   0.000   10.0      0.000   34.0  29.410000   \n",
              "4    23.810000   43.0  119.050000   0.000   12.0      0.000   42.0  23.810000   \n",
              "..         ...    ...         ...     ...    ...        ...    ...        ...   \n",
              "606   0.000000   15.0  133.332993   0.000    5.0      0.000   15.0  66.667000   \n",
              "607  98.039001   26.0  176.470993   7.071    1.0      0.028   51.0  19.608000   \n",
              "608  24.389999   21.0  243.901993   7.071    6.0      0.067   41.0  24.389999   \n",
              "609   0.000000    3.0    0.000000   0.000    0.0      0.000    3.0   0.000000   \n",
              "610  55.556000   18.0  111.111000   0.000    6.0      0.000   18.0   0.000000   \n",
              "\n",
              "       CNCLogic       PCDCp  \n",
              "0      0.000000    1.830000  \n",
              "1     62.500000   73.240000  \n",
              "2    162.160000  100.000000  \n",
              "3     29.410000   24.830000  \n",
              "4     71.430000   79.670000  \n",
              "..          ...         ...  \n",
              "606   66.667000   46.020000  \n",
              "607   58.824001   73.239998  \n",
              "608   73.170998   94.059998  \n",
              "609    0.000000    0.080000  \n",
              "610   55.556000   99.720001  \n",
              "\n",
              "[611 rows x 17 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-e2842a03-ae58-46a2-8fc6-50c46af88101\">\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>WRDFAMc</th>\n",
              "      <th>WRDFRQc</th>\n",
              "      <th>WRDNOUN</th>\n",
              "      <th>DESWLlt</th>\n",
              "      <th>WRDHYPnv</th>\n",
              "      <th>DESWLltd</th>\n",
              "      <th>LDTTRa</th>\n",
              "      <th>WRDADV</th>\n",
              "      <th>DESSL</th>\n",
              "      <th>DRVP</th>\n",
              "      <th>DESSLd</th>\n",
              "      <th>SYNLE</th>\n",
              "      <th>SYNSTRUTt</th>\n",
              "      <th>DESWC</th>\n",
              "      <th>SMCAUSv</th>\n",
              "      <th>CNCLogic</th>\n",
              "      <th>PCDCp</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>583.860000</td>\n",
              "      <td>2.110</td>\n",
              "      <td>400.000000</td>\n",
              "      <td>4.630</td>\n",
              "      <td>2.810</td>\n",
              "      <td>2.160</td>\n",
              "      <td>0.740</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>35.0</td>\n",
              "      <td>171.430000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>17.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>35.0</td>\n",
              "      <td>28.570000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>1.830000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>583.000000</td>\n",
              "      <td>2.040</td>\n",
              "      <td>406.250000</td>\n",
              "      <td>5.560</td>\n",
              "      <td>2.590</td>\n",
              "      <td>3.010</td>\n",
              "      <td>0.790</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>33.0</td>\n",
              "      <td>93.750000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>10.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>32.0</td>\n",
              "      <td>31.250000</td>\n",
              "      <td>62.500000</td>\n",
              "      <td>73.240000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>568.940000</td>\n",
              "      <td>1.970</td>\n",
              "      <td>513.510000</td>\n",
              "      <td>5.700</td>\n",
              "      <td>3.280</td>\n",
              "      <td>3.030</td>\n",
              "      <td>0.700</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>37.0</td>\n",
              "      <td>81.080000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>12.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>37.0</td>\n",
              "      <td>27.030000</td>\n",
              "      <td>162.160000</td>\n",
              "      <td>100.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>563.440000</td>\n",
              "      <td>1.760</td>\n",
              "      <td>411.760000</td>\n",
              "      <td>6.210</td>\n",
              "      <td>2.940</td>\n",
              "      <td>2.900</td>\n",
              "      <td>0.790</td>\n",
              "      <td>29.410000</td>\n",
              "      <td>34.0</td>\n",
              "      <td>147.060000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>10.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>34.0</td>\n",
              "      <td>29.410000</td>\n",
              "      <td>29.410000</td>\n",
              "      <td>24.830000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>556.500000</td>\n",
              "      <td>2.000</td>\n",
              "      <td>500.000000</td>\n",
              "      <td>6.070</td>\n",
              "      <td>3.590</td>\n",
              "      <td>3.200</td>\n",
              "      <td>0.700</td>\n",
              "      <td>23.810000</td>\n",
              "      <td>43.0</td>\n",
              "      <td>119.050000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>12.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>42.0</td>\n",
              "      <td>23.810000</td>\n",
              "      <td>71.430000</td>\n",
              "      <td>79.670000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606</th>\n",
              "      <td>588.125000</td>\n",
              "      <td>2.223</td>\n",
              "      <td>400.000000</td>\n",
              "      <td>4.867</td>\n",
              "      <td>2.758</td>\n",
              "      <td>2.416</td>\n",
              "      <td>0.867</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>15.0</td>\n",
              "      <td>133.332993</td>\n",
              "      <td>0.000</td>\n",
              "      <td>5.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>15.0</td>\n",
              "      <td>66.667000</td>\n",
              "      <td>66.667000</td>\n",
              "      <td>46.020000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>607</th>\n",
              "      <td>580.221985</td>\n",
              "      <td>2.360</td>\n",
              "      <td>215.686996</td>\n",
              "      <td>4.137</td>\n",
              "      <td>1.726</td>\n",
              "      <td>2.333</td>\n",
              "      <td>0.769</td>\n",
              "      <td>98.039001</td>\n",
              "      <td>26.0</td>\n",
              "      <td>176.470993</td>\n",
              "      <td>7.071</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.028</td>\n",
              "      <td>51.0</td>\n",
              "      <td>19.608000</td>\n",
              "      <td>58.824001</td>\n",
              "      <td>73.239998</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>608</th>\n",
              "      <td>577.267029</td>\n",
              "      <td>2.221</td>\n",
              "      <td>341.463013</td>\n",
              "      <td>4.293</td>\n",
              "      <td>2.714</td>\n",
              "      <td>2.358</td>\n",
              "      <td>0.738</td>\n",
              "      <td>24.389999</td>\n",
              "      <td>21.0</td>\n",
              "      <td>243.901993</td>\n",
              "      <td>7.071</td>\n",
              "      <td>6.0</td>\n",
              "      <td>0.067</td>\n",
              "      <td>41.0</td>\n",
              "      <td>24.389999</td>\n",
              "      <td>73.170998</td>\n",
              "      <td>94.059998</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>609</th>\n",
              "      <td>575.500000</td>\n",
              "      <td>1.530</td>\n",
              "      <td>666.666016</td>\n",
              "      <td>6.000</td>\n",
              "      <td>6.778</td>\n",
              "      <td>3.000</td>\n",
              "      <td>1.000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>3.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.080000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>610</th>\n",
              "      <td>593.778015</td>\n",
              "      <td>2.571</td>\n",
              "      <td>222.223007</td>\n",
              "      <td>4.444</td>\n",
              "      <td>1.385</td>\n",
              "      <td>1.947</td>\n",
              "      <td>0.778</td>\n",
              "      <td>55.556000</td>\n",
              "      <td>18.0</td>\n",
              "      <td>111.111000</td>\n",
              "      <td>0.000</td>\n",
              "      <td>6.0</td>\n",
              "      <td>0.000</td>\n",
              "      <td>18.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>55.556000</td>\n",
              "      <td>99.720001</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>611 rows × 17 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e2842a03-ae58-46a2-8fc6-50c46af88101')\"\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-e2842a03-ae58-46a2-8fc6-50c46af88101 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-e2842a03-ae58-46a2-8fc6-50c46af88101');\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": 105
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import model_selection\n",
        "X_train_s_cm, X_test_s_cm, y_train_s_cm, y_test_s_cm = model_selection.train_test_split(X_selected_cm, y_selected_cm, test_size=0.25, stratify = y_selected_cm, random_state=1) #stratified sampling,\n",
        "print('training data has ' + str(X_train_s_cm.shape[0]) + ' observation with ' + str(X_train_s_cm.shape[1]) + ' features')\n",
        "print('test data has ' + str(X_test_s_cm.shape[0]) + ' observation with ' + str(X_test_s_cm.shape[1]) + ' features')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WlKnywCjdS27",
        "outputId": "59282f4f-dab0-4b87-9fac-a24621b4a4c8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "training data has 458 observation with 17 features\n",
            "test data has 153 observation with 17 features\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.linear_model import LogisticRegression\n",
        "# Logistic Regression\n",
        "classifier_logistic_cm = LogisticRegression()"
      ],
      "metadata": {
        "id": "XZ0VpP2qeo9p"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "classifier_logistic_cm.fit(X_train_s_cm, y_train_s_cm)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rBOK4qiDeySs",
        "outputId": "ebff0753-3346-4c8e-d76c-4a323d980ec1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
            "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
            "\n",
            "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
            "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
            "Please also refer to the documentation for alternative solver options:\n",
            "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
            "  n_iter_i = _check_optimize_result(\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 109
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic_cm.predict_proba(X_test_s_cm)[:,0] #level 1 prob"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O_AbZv2zfHuP",
        "outputId": "9d0b08b8-07b3-4c22-e1d4-7eccc85b997a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([8.80069485e-01, 7.27145652e-05, 2.44826212e-03, 6.49346667e-05,\n",
              "       2.27916411e-02, 7.57656240e-01, 5.63528918e-03, 8.97157602e-01,\n",
              "       6.92915966e-02, 9.12837107e-01, 5.59437138e-04, 6.40276939e-06,\n",
              "       1.91261590e-02, 9.37997397e-04, 2.44495992e-06, 1.28657074e-04,\n",
              "       6.26128111e-04, 1.79815257e-01, 2.34744732e-03, 8.63238427e-02,\n",
              "       8.94644173e-01, 2.77989580e-03, 1.77071217e-03, 5.24323521e-04,\n",
              "       9.66828003e-01, 2.67897389e-04, 1.93413162e-02, 5.10501302e-05,\n",
              "       2.97091833e-03, 7.10524534e-03, 6.76219309e-04, 1.22790440e-04,\n",
              "       5.84503532e-03, 5.91954832e-04, 9.94090334e-01, 1.20721452e-01,\n",
              "       1.30854456e-05, 6.40902234e-04, 7.77832712e-04, 9.01250385e-06,\n",
              "       2.25640277e-03, 4.28614111e-05, 2.66131653e-03, 2.32302174e-01,\n",
              "       7.57400555e-03, 3.89795731e-06, 6.33814096e-02, 5.52160569e-03,\n",
              "       1.33707102e-02, 6.11904977e-02, 9.26000843e-05, 8.62411314e-05,\n",
              "       3.22013432e-05, 2.03942578e-02, 9.78626162e-01, 7.59046086e-01,\n",
              "       4.33824021e-04, 2.89519368e-03, 1.20120743e-01, 5.37686620e-02,\n",
              "       8.59098985e-01, 1.09092515e-01, 3.26870293e-08, 7.01468428e-01,\n",
              "       3.90067682e-03, 6.19963370e-01, 2.69864497e-04, 1.22796370e-06,\n",
              "       2.24216797e-03, 2.33364694e-05, 5.80177808e-04, 2.49236344e-01,\n",
              "       1.54683275e-05, 2.21832303e-02, 6.32889058e-04, 1.38459487e-03,\n",
              "       9.33944768e-04, 1.15091427e-03, 5.19972708e-04, 7.14403236e-01,\n",
              "       9.37908690e-01, 2.92094224e-03, 7.66510707e-03, 6.45492396e-05,\n",
              "       3.92200914e-04, 4.37498678e-05, 7.17570685e-01, 2.09036033e-03,\n",
              "       3.32522781e-01, 8.76555606e-01, 1.29171221e-06, 6.83511255e-03,\n",
              "       1.59122141e-03, 1.18802632e-02, 9.99999286e-01, 1.10261873e-02,\n",
              "       3.43033722e-03, 4.28794666e-03, 5.76713255e-03, 6.10450986e-02,\n",
              "       6.62910608e-01, 4.92181981e-02, 3.39959428e-05, 3.11411781e-02,\n",
              "       1.27442617e-02, 8.29941027e-01, 2.43717785e-04, 6.62160303e-01,\n",
              "       8.18314649e-02, 8.14033545e-02, 1.54011806e-04, 1.00079864e-02,\n",
              "       4.87186959e-03, 7.07462507e-03, 3.97462242e-05, 1.08329892e-01,\n",
              "       8.54908101e-02, 6.57766864e-03, 2.72520206e-03, 1.00597000e-02,\n",
              "       8.76688499e-03, 2.43815770e-03, 1.66715064e-03, 7.84982497e-01,\n",
              "       4.80167761e-04, 1.18172649e-02, 6.62009253e-01, 4.65000306e-04,\n",
              "       1.35217155e-03, 1.58854594e-04, 1.73546510e-02, 9.89707915e-01,\n",
              "       1.13748775e-02, 9.28644661e-04, 1.71667920e-04, 2.74751081e-02,\n",
              "       3.42007587e-02, 5.26752745e-02, 1.23996797e-05, 6.26914722e-06,\n",
              "       1.23758948e-03, 2.66067122e-02, 1.92620196e-04, 4.65966231e-03,\n",
              "       1.57171645e-04, 3.99562189e-02, 8.55421713e-06, 6.24306724e-03,\n",
              "       1.65578495e-04, 3.15869108e-01, 4.07451327e-01, 7.49132822e-05,\n",
              "       1.49580147e-02])"
            ]
          },
          "metadata": {},
          "execution_count": 110
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic_cm.score(X_test_s_cm, y_test_s_cm)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rpe53rgOgcpg",
        "outputId": "b4f91f0a-2487-49f8-ae04-b4a5b258a673"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9411764705882353"
            ]
          },
          "metadata": {},
          "execution_count": 111
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_axis = classifier_logistic_cm.predict_proba(X_selected_cm)[:,0]"
      ],
      "metadata": {
        "id": "qMsRDNZLh8av"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "threshold_cm = Find_Optimal_Cutoff(y_test_s_cm, classifier_logistic_cm.predict_proba(X_test_s_cm)[:,0])\n",
        "threshold_cm"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rJA9p_RNKPpi",
        "outputId": "ef8afcd7-78be-4cad-f039-007caa66e5f3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.8326110110038114]"
            ]
          },
          "metadata": {},
          "execution_count": 113
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "b = to_labels(classifier_logistic_cm.predict_proba(X_test_s_cm)[:,1],0.832)\n",
        "accuracy_score(y_test_s_cm, b)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "K3-jQhYiKruy",
        "outputId": "fe59e92d-a658-42cb-9ce3-f5f0008ffa28"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9411764705882353"
            ]
          },
          "metadata": {},
          "execution_count": 114
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(precision_score(y_test_s_cm, b))\n",
        "print(recall_score(y_test_s_cm, b))\n",
        "print(f1_score(y_test_s_cm, b))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jV4CyuNELOhJ",
        "outputId": "fbd02cb7-72f7-4647-d0da-dc42d3f9e57b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.0\n",
            "0.9323308270676691\n",
            "0.9649805447470817\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "roc_auc_score(y_test_s_cm,classifier_logistic_cm.predict_proba(X_test_s_cm)[:,1])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Yms-cTfwRnV-",
        "outputId": "c2e8f1e2-8fe7-4e6f-f2c9-68ad6c38f55c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9714285714285713"
            ]
          },
          "metadata": {},
          "execution_count": 116
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 6: Plot CM and TFIDF as axis to differentiate two types of responses"
      ],
      "metadata": {
        "id": "y2KxfyjLgiWe"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plot_data = pd.DataFrame({'Coh-Metrix':X_axis, 'TF-IDF':y_axis, 'Type':df[\"Type_num\"]})"
      ],
      "metadata": {
        "id": "pWOxbNowjSyw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plot_data"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 424
        },
        "id": "6E4mlLgSj9BL",
        "outputId": "615c8599-01b8-40b8-d796-e7e860f9ebf7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "     Coh-Metrix    TF-IDF  Type\n",
              "0      0.392840  0.041504     0\n",
              "1      0.332523  0.038902     0\n",
              "2      0.530420  0.097838     0\n",
              "3      0.665076  0.284574     0\n",
              "4      0.989708  0.226095     0\n",
              "..          ...       ...   ...\n",
              "606    0.002980  0.377267     1\n",
              "607    0.002724  0.091429     1\n",
              "608    0.019341  0.079795     1\n",
              "609    0.946095  0.098688     1\n",
              "610    0.000159  0.006217     1\n",
              "\n",
              "[611 rows x 3 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-6c63c8df-f66a-4a78-b3fe-d264d5bfeaef\">\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>Coh-Metrix</th>\n",
              "      <th>TF-IDF</th>\n",
              "      <th>Type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.392840</td>\n",
              "      <td>0.041504</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.332523</td>\n",
              "      <td>0.038902</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.530420</td>\n",
              "      <td>0.097838</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.665076</td>\n",
              "      <td>0.284574</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.989708</td>\n",
              "      <td>0.226095</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>606</th>\n",
              "      <td>0.002980</td>\n",
              "      <td>0.377267</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>607</th>\n",
              "      <td>0.002724</td>\n",
              "      <td>0.091429</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>608</th>\n",
              "      <td>0.019341</td>\n",
              "      <td>0.079795</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>609</th>\n",
              "      <td>0.946095</td>\n",
              "      <td>0.098688</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>610</th>\n",
              "      <td>0.000159</td>\n",
              "      <td>0.006217</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>611 rows × 3 columns</p>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-6c63c8df-f66a-4a78-b3fe-d264d5bfeaef')\"\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-6c63c8df-f66a-4a78-b3fe-d264d5bfeaef 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-6c63c8df-f66a-4a78-b3fe-d264d5bfeaef');\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": 118
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Use the 'hue' argument to provide a factor variable\n",
        "sns.lmplot(x=\"Coh-Metrix\", y=\"TF-IDF\", data=plot_data, fit_reg=False, hue='Type', legend=False)\n",
        "\n",
        "# Move the legend to an empty part of the plot\n",
        "plt.legend(loc='lower right')\n",
        "\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 369
        },
        "id": "Bbdw_eL4ghDu",
        "outputId": "42453184-25ae-4761-8764-6d803fa71419"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 360x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 7: Design and develop a machine-based algorithm to classify response types"
      ],
      "metadata": {
        "id": "VRRLH50YLqEk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_all = df[\"Type_num\"]"
      ],
      "metadata": {
        "id": "--vsXNLzLotY"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df1 = pd.DataFrame(tfidf_matrix.todense(), columns = tf_selected_words)\n",
        "print(df1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5aoWX_IBOJwl",
        "outputId": "26546fd9-f132-42bd-8f25-31643ba61ee6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "     agricultur      area      bean   biomass     chang  communiti      corn  \\\n",
            "0      0.000000  0.225593  0.207403  0.000000  0.000000   0.000000  0.153542   \n",
            "1      0.000000  0.000000  0.000000  0.000000  0.000000   0.000000  0.216561   \n",
            "2      0.000000  0.000000  0.000000  0.000000  0.000000   0.000000  0.190940   \n",
            "3      0.350588  0.000000  0.000000  0.000000  0.000000   0.304715  0.179241   \n",
            "4      0.000000  0.000000  0.183307  0.000000  0.538434   0.000000  0.271408   \n",
            "..          ...       ...       ...       ...       ...        ...       ...   \n",
            "606    0.000000  0.000000  0.000000  0.000000  0.000000   0.000000  0.000000   \n",
            "607    0.000000  0.000000  0.332754  0.000000  0.000000   0.000000  0.246341   \n",
            "608    0.000000  0.000000  0.169965  0.174769  0.000000   0.000000  0.251654   \n",
            "609    0.000000  0.000000  0.000000  0.000000  0.000000   0.000000  0.000000   \n",
            "610    0.000000  0.000000  0.000000  0.000000  0.000000   0.000000  0.000000   \n",
            "\n",
            "         crop  decreas    energi  ...   product    requir    result     river  \\\n",
            "0    0.256552  0.00000  0.000000  ...  0.000000  0.259211  0.272729  0.000000   \n",
            "1    0.000000  0.00000  0.000000  ...  0.306341  0.000000  0.000000  0.400198   \n",
            "2    0.000000  0.00000  0.000000  ...  0.540197  0.000000  0.000000  0.000000   \n",
            "3    0.000000  0.00000  0.000000  ...  0.253549  0.000000  0.318376  0.331232   \n",
            "4    0.000000  0.00000  0.000000  ...  0.575889  0.000000  0.000000  0.000000   \n",
            "..        ...      ...       ...  ...       ...       ...       ...       ...   \n",
            "606  0.000000  0.00000  0.255516  ...  0.666193  0.000000  0.418263  0.000000   \n",
            "607  0.000000  0.00000  0.267307  ...  0.000000  0.000000  0.000000  0.000000   \n",
            "608  0.000000  0.47291  0.136536  ...  0.000000  0.000000  0.000000  0.000000   \n",
            "609  0.000000  0.00000  0.000000  ...  0.000000  0.000000  0.000000  0.000000   \n",
            "610  0.000000  0.00000  0.000000  ...  0.000000  0.000000  0.000000  0.000000   \n",
            "\n",
            "        sourc    switch  tradeoff      usag       use     water  \n",
            "0    0.273774  0.000000  0.000000  0.287315  0.000000  0.441451  \n",
            "1    0.000000  0.000000  0.000000  0.000000  0.241756  0.622636  \n",
            "2    0.000000  0.000000  0.000000  0.357296  0.426309  0.548974  \n",
            "3    0.000000  0.000000  0.000000  0.335404  0.000000  0.386503  \n",
            "4    0.241967  0.000000  0.000000  0.000000  0.000000  0.390164  \n",
            "..        ...       ...       ...       ...       ...       ...  \n",
            "606  0.000000  0.000000  0.000000  0.440633  0.000000  0.169254  \n",
            "607  0.000000  0.000000  0.413017  0.000000  0.000000  0.000000  \n",
            "608  0.000000  0.242756  0.000000  0.000000  0.140466  0.090442  \n",
            "609  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  \n",
            "610  0.000000  0.000000  0.000000  0.000000  0.000000  1.000000  \n",
            "\n",
            "[611 rows x 30 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "X_all = pd.concat([X_selected_cm, df1], axis=1)\n",
        "print(X_all)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nU6w1Ub3O-qi",
        "outputId": "ee8585f3-7df9-40b0-c590-c1d7a0928014"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "        WRDFAMc  WRDFRQc     WRDNOUN  DESWLlt  WRDHYPnv  DESWLltd  LDTTRa  \\\n",
            "0    583.860000    2.110  400.000000    4.630     2.810     2.160   0.740   \n",
            "1    583.000000    2.040  406.250000    5.560     2.590     3.010   0.790   \n",
            "2    568.940000    1.970  513.510000    5.700     3.280     3.030   0.700   \n",
            "3    563.440000    1.760  411.760000    6.210     2.940     2.900   0.790   \n",
            "4    556.500000    2.000  500.000000    6.070     3.590     3.200   0.700   \n",
            "..          ...      ...         ...      ...       ...       ...     ...   \n",
            "606  588.125000    2.223  400.000000    4.867     2.758     2.416   0.867   \n",
            "607  580.221985    2.360  215.686996    4.137     1.726     2.333   0.769   \n",
            "608  577.267029    2.221  341.463013    4.293     2.714     2.358   0.738   \n",
            "609  575.500000    1.530  666.666016    6.000     6.778     3.000   1.000   \n",
            "610  593.778015    2.571  222.223007    4.444     1.385     1.947   0.778   \n",
            "\n",
            "        WRDADV  DESSL        DRVP  ...   product    requir    result  \\\n",
            "0     0.000000   35.0  171.430000  ...  0.000000  0.259211  0.272729   \n",
            "1     0.000000   33.0   93.750000  ...  0.306341  0.000000  0.000000   \n",
            "2     0.000000   37.0   81.080000  ...  0.540197  0.000000  0.000000   \n",
            "3    29.410000   34.0  147.060000  ...  0.253549  0.000000  0.318376   \n",
            "4    23.810000   43.0  119.050000  ...  0.575889  0.000000  0.000000   \n",
            "..         ...    ...         ...  ...       ...       ...       ...   \n",
            "606   0.000000   15.0  133.332993  ...  0.666193  0.000000  0.418263   \n",
            "607  98.039001   26.0  176.470993  ...  0.000000  0.000000  0.000000   \n",
            "608  24.389999   21.0  243.901993  ...  0.000000  0.000000  0.000000   \n",
            "609   0.000000    3.0    0.000000  ...  0.000000  0.000000  0.000000   \n",
            "610  55.556000   18.0  111.111000  ...  0.000000  0.000000  0.000000   \n",
            "\n",
            "        river     sourc    switch  tradeoff      usag       use     water  \n",
            "0    0.000000  0.273774  0.000000  0.000000  0.287315  0.000000  0.441451  \n",
            "1    0.400198  0.000000  0.000000  0.000000  0.000000  0.241756  0.622636  \n",
            "2    0.000000  0.000000  0.000000  0.000000  0.357296  0.426309  0.548974  \n",
            "3    0.331232  0.000000  0.000000  0.000000  0.335404  0.000000  0.386503  \n",
            "4    0.000000  0.241967  0.000000  0.000000  0.000000  0.000000  0.390164  \n",
            "..        ...       ...       ...       ...       ...       ...       ...  \n",
            "606  0.000000  0.000000  0.000000  0.000000  0.440633  0.000000  0.169254  \n",
            "607  0.000000  0.000000  0.000000  0.413017  0.000000  0.000000  0.000000  \n",
            "608  0.000000  0.000000  0.242756  0.000000  0.000000  0.140466  0.090442  \n",
            "609  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  \n",
            "610  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  1.000000  \n",
            "\n",
            "[611 rows x 47 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import model_selection\n",
        "X_train_all, X_test_all, y_train_all, y_test_all = model_selection.train_test_split(X_all, y_all, test_size=0.25, stratify = y_all, random_state=1) #stratified sampling,\n",
        "print('training data has ' + str(X_train_all.shape[0]) + ' observation with ' + str(X_train_all.shape[1]) + ' features')\n",
        "print('test data has ' + str(X_test_all.shape[0]) + ' observation with ' + str(X_test_all.shape[1]) + ' features')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3VWlV8DYPbPH",
        "outputId": "c5f7fcad-7b80-4eec-a726-952a8da60edb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "training data has 458 observation with 47 features\n",
            "test data has 153 observation with 47 features\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic_all = LogisticRegression()\n",
        "classifier_logistic_all.fit(X_train_all, y_train_all)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ch5xPg58PLXA",
        "outputId": "36e03a64-fe3f-4012-fb24-50d53bb6e06c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.8/dist-packages/sklearn/linear_model/_logistic.py:814: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
            "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
            "\n",
            "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
            "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
            "Please also refer to the documentation for alternative solver options:\n",
            "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
            "  n_iter_i = _check_optimize_result(\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LogisticRegression()"
            ]
          },
          "metadata": {},
          "execution_count": 124
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "threshold_all = Find_Optimal_Cutoff(y_test_all, classifier_logistic_all.predict_proba(X_test_all)[:,0])\n",
        "threshold_all"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eZZOd6OHP4yO",
        "outputId": "a4df8623-f9aa-454c-b655-f399bbf662b2"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.8326110110038114]"
            ]
          },
          "metadata": {},
          "execution_count": 125
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "c = to_labels(classifier_logistic_all.predict_proba(X_test_all)[:,1],0.832)\n",
        "accuracy_score(y_test_all, c)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vlexXltWQPdH",
        "outputId": "d7853d7a-6318-432a-e02b-626627648cb8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9477124183006536"
            ]
          },
          "metadata": {},
          "execution_count": 126
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(precision_score(y_test_all, c))\n",
        "print(recall_score(y_test_all, c))\n",
        "print(f1_score(y_test_all, c))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "JDkK-YJMQZlC",
        "outputId": "653e080c-2690-4986-9050-4ff9f1fa75f7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.0\n",
            "0.9398496240601504\n",
            "0.9689922480620156\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "roc_auc_score(y_test_all, classifier_logistic_all.predict_proba(X_test_all)[:,1])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-x6pXQ4nRWdK",
        "outputId": "0d9f0aa7-37fb-499e-bc17-6668ce2ce899"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9766917293233082"
            ]
          },
          "metadata": {},
          "execution_count": 128
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.metrics._plot.confusion_matrix import confusion_matrix\n",
        "confusion_matrix(y_test_all, c)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1ZGXeVnwrxJS",
        "outputId": "32be100e-9275-439e-ead4-fef80e8dec5f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 20,   0],\n",
              "       [  8, 125]])"
            ]
          },
          "metadata": {},
          "execution_count": 131
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Part 7: Design and develop a machine-based algorithm to classify response types"
      ],
      "metadata": {
        "id": "8GC5ouwdWlmy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#https://drive.google.com/file/d/10tiiwMd4k5rjNheGJI144fSq3UYpP4yW/view?usp=share_link\n",
        "file = drive.CreateFile({'id':'10tiiwMd4k5rjNheGJI144fSq3UYpP4yW'})\n",
        "file.GetContentFile('Peach_test.txt')  # tab-separated"
      ],
      "metadata": {
        "id": "gMeuDFjdXEkW"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test_df = pd.read_csv('Peach_test.txt', encoding= 'unicode_escape', sep='\\t', error_bad_lines=False)"
      ],
      "metadata": {
        "id": "wpCY8ji7lGPf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test_df.isna().sum()"
      ],
      "metadata": {
        "id": "RoMx6A8EqUfM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# generate text dataset\n",
        "TF_test_data = test_df.loc[:, 'Responses'].tolist()"
      ],
      "metadata": {
        "id": "NyyzdEgSqWqb"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Use nltk's English stopwords.\n",
        "stopwords = nltk.corpus.stopwords.words('english') #stopwords.append(\"n't\")\n",
        "stopwords.append(\"'s\")\n",
        "stopwords.append(\"'m\")\n",
        "stopwords.append(\"br\") #html <br>\n",
        "stopwords.append(\"watch\")\n",
        "\n",
        "print (\"We use \" + str(len(stopwords)) + \" stop-words from nltk library.\")\n",
        "print (stopwords[:10])"
      ],
      "metadata": {
        "id": "wdkE5wAuqeps"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from nltk.stem.snowball import SnowballStemmer\n",
        "# from nltk.stem import WordNetLemmatizer\n",
        "\n",
        "stemmer = SnowballStemmer(\"english\")\n",
        "\n",
        "# tokenization and stemming\n",
        "def tokenization_and_stemming(text):\n",
        "    tokens = []\n",
        "    # exclude stop words and tokenize the document, generate a list of string\n",
        "    for word in nltk.word_tokenize(text):\n",
        "        if word.lower() not in stopwords:\n",
        "            tokens.append(word.lower())\n",
        "\n",
        "    filtered_tokens = []\n",
        "\n",
        "    # filter out any tokens not containing letters (e.g., numeric tokens, raw punctuation)\n",
        "    for token in tokens:\n",
        "        if token.isalpha():\n",
        "            filtered_tokens.append(token)\n",
        "\n",
        "    # stemming\n",
        "    stems = [stemmer.stem(t) for t in filtered_tokens]\n",
        "    return stems"
      ],
      "metadata": {
        "id": "BqgMqd77qhT7"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tokenization_and_stemming(TF_test_data[0])"
      ],
      "metadata": {
        "id": "5CJeU6XRqojH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "TF_test_data[0]"
      ],
      "metadata": {
        "id": "KXx716bTqvPx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "# define vectorizer parameters\n",
        "# TfidfVectorizer will help us to create tf-idf matrix\n",
        "# max_df : maximum document frequency for the given word, don't keep when a word appears in the documents at this frequency rate\n",
        "# min_df : minimum document frequency for the given word, don't keep when a word appears in the documents at this frequency rate\n",
        "# max_features: maximum number of words\n",
        "# use_idf: if not true, we only calculate tf\n",
        "# stop_words : built-in stop words\n",
        "# tokenizer: how to tokenize the document\n",
        "# ngram_range: (min_value, max_value), eg. (1, 3) means the result will include 1-gram, 2-gram, 3-gram\n",
        "tfidf_test_model = TfidfVectorizer(max_df=0.99, max_features=1000,\n",
        "                                 min_df=0.1, stop_words='english',\n",
        "                                 use_idf=True, tokenizer=tokenization_and_stemming, ngram_range=(1,1))\n",
        "\n",
        "tfidf_test_matrix = tfidf_test_model.fit_transform(TF_test_data) #fit the vectorizer to synopses\n",
        "\n",
        "print (\"In total, there are \" + str(tfidf_matrix.shape[0]) + \\\n",
        "      \" reviews and \" + str(tfidf_matrix.shape[1]) + \" terms.\")"
      ],
      "metadata": {
        "id": "8ETX8GKvqz_S"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_test_matrix"
      ],
      "metadata": {
        "id": "FNpgjs3YrFUr"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_test_matrix.toarray()"
      ],
      "metadata": {
        "id": "Yiaq-3JCrLcp"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tfidf_test_matrix.todense()"
      ],
      "metadata": {
        "id": "StCr9GDqrQQX"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tf_test_words = tfidf_test_model.get_feature_names()"
      ],
      "metadata": {
        "id": "6RPMZjlzrXIj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "tf_test_words"
      ],
      "metadata": {
        "id": "TbIvHsNxrfyo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Part 7.1 Content-level predictors only**"
      ],
      "metadata": {
        "id": "atFIn6ZbkV3C"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "y_test_tf = test_df[\"Type_num\"]\n",
        "X_test_tf = tfidf_test_matrix"
      ],
      "metadata": {
        "id": "jGMFz1wQk0we"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "rom sklearn import model_selection\n",
        "X_train_peach, X_test_peach, y_train_peach, y_test_peach = model_selection.train_test_split(X_test_tf, y_test_tf, test_size=0.25, stratify = y_test_tf, random_state=1) #stratified sampling,\n",
        "\n",
        "print('training data has ' + str(X_train.shape[0]) + ' observation with ' + str(X_train.shape[1]) + ' terms')\n",
        "print('test data has ' + str(X_test.shape[0]) + ' observation with ' + str(X_test.shape[1]) + ' terms')"
      ],
      "metadata": {
        "id": "DxYd6267s5iz"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Logistic Regression\n",
        "classifier_logistic = LogisticRegression()\n",
        "classifier_logistic.fit(X_train_peach, y_train_peach)"
      ],
      "metadata": {
        "id": "2K1w8jSMuU72"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "classifier_logistic.score(X_test_peach, y_test_peach)"
      ],
      "metadata": {
        "id": "bGkh70d5uZqS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "# helper function for printing out grid search results\n",
        "def print_grid_search_metrics(gs):\n",
        "    print (\"Best score: \" + str(gs.best_score_))\n",
        "    print (\"Best parameters set:\")\n",
        "    best_parameters = gs.best_params_\n",
        "    for param_name in sorted(best_parameters.keys()):\n",
        "        print(param_name + ':' + str(best_parameters[param_name]))"
      ],
      "metadata": {
        "id": "8-Wmb3pDj7Xy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "parameters = {\n",
        "    'penalty':('l2','l1'),\n",
        "    'C':(0.01, 0.05, 0.1, 0.2, 1)\n",
        "}\n",
        "Grid_LR = GridSearchCV(LogisticRegression(solver='liblinear'),parameters, cv=5)\n",
        "Grid_LR.fit(X_train_peach, y_train_peach)"
      ],
      "metadata": {
        "id": "_ViKUSFUkvkU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# the best hyperparameter combination\n",
        "print_grid_search_metrics(Grid_LR)"
      ],
      "metadata": {
        "id": "jKortLi-teNK"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# best model\n",
        "best_LR_model = Grid_LR.best_estimator_"
      ],
      "metadata": {
        "id": "zjjQONcktmJu"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "best_LR_model.predict(X_test_peach)"
      ],
      "metadata": {
        "id": "em1bXAQBtooq"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "best_LR_model.score(X_test_peach, y_test_peach)"
      ],
      "metadata": {
        "id": "y66KlN4utpbI"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Part 7.2 Coh-metrix predictors only**"
      ],
      "metadata": {
        "id": "vtjVLFLGvI4I"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "to_drop = [\"Type\", \"Type_num\", \"Responses\"]\n",
        "y_test_cm = test_df[\"Type_num\"]\n",
        "X_test_cm = test_df.drop(to_drop, axis=1)"
      ],
      "metadata": {
        "id": "tq3gt5oVvrPj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "y_peach_cm = test_df[\"Type_num\"]\n",
        "X_peach_cm = X_test_cm.iloc[:,[97, 105, 93, 83, 96, 104, 9, 106, 8, 7, 10, 2, 5, 62, 76, 48]]"
      ],
      "metadata": {
        "id": "guXRbXR1vPoR"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn import model_selection\n",
        "X_train_peach_cm, X_test_peach_cm, y_train_peach_cm, y_test_peach_cm = model_selection.train_test_split(X_peach_cm, y_peach_cm, test_size=0.25, stratify = y_peach_cm, random_state=1) #stratified sampling,\n",
        "\n",
        "print('training data has ' + str(X_train_peach_cm.shape[0]) + ' observation with ' + str(X_train_peach_cm.shape[1]) + ' features')\n",
        "print('test data has ' + str(X_test_peach_cm.shape[0]) + ' observation with ' + str(X_test_peach_cm.shape[1]) + ' features')"
      ],
      "metadata": {
        "id": "G_5L1J9sv_Cx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "ufTh8VKBwOdJ"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}