Heart_Disease_Prediction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Heart Disease Prediction",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "\n",
        "# import warnings filter\n",
        "from warnings import simplefilter\n",
        "# ignore all future warnings\n",
        "simplefilter(action='ignore', category = FutureWarning)\n",
        "# import warnings filter\n",
        "from warnings import simplefilter\n",
        "# ignore all future warnings\n",
        "simplefilter(action='ignore', category = FutureWarning)"
      ],
      "metadata": {
        "id": "RPgWZrh_6-jE"
      },
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df = pd.read_csv('https://raw.githubusercontent.com/cpluz-shrijayan/heartDiseasePrediction/main/cleveland_Heart_Data.csv', header = None)\n",
        "\n",
        "df.columns = ['age', 'sex', 'cp', 'trestbps', 'chol',\n",
        "              'fbs', 'restecg', 'thalach', 'exang', \n",
        "              'oldpeak', 'slope', 'ca', 'thal', 'target']"
      ],
      "metadata": {
        "id": "2j2sth2k751i"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "### 1 = male, 0 = female\n",
        "df.isnull().sum()\n",
        "\n",
        "df['target'] = df.target.map({0: 0, 1: 1, 2: 1, 3: 1, 4: 1})\n",
        "df['sex'] = df.sex.map({0: 'female', 1: 'male'})\n",
        "df['thal'] = df.thal.fillna(df.thal.mean())\n",
        "df['ca'] = df.ca.fillna(df.ca.mean())"
      ],
      "metadata": {
        "id": "Ss7SGJaX8D-Y"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n"
      ],
      "metadata": {
        "id": "SO8L9q1d8HYE"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# distribution of target vs age \n",
        "sns.set_context(\"paper\", font_scale = 2, rc = {\"font.size\": 20,\"axes.titlesize\": 25,\"axes.labelsize\": 20}) \n",
        "sns.catplot(kind = 'count', data = df, x = 'age', hue = 'target', order = df['age'].sort_values().unique())\n",
        "plt.title('Variation of Age for each target class')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 373
        },
        "id": "jHTKoTiC8KIi",
        "outputId": "12151bfa-0868-47d3-baae-e8be37b16f77"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 434.485x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# barplot of age vs sex with hue = target\n",
        "sns.catplot(kind = 'bar', data = df, y = 'age', x = 'sex', hue = 'target')\n",
        "plt.title('Distribution of age vs sex with the target class')\n",
        "plt.show()\n",
        "\n",
        "df['sex'] = df.sex.map({'female': 0, 'male': 1})"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 373
        },
        "id": "Cp-PeC1N8Me_",
        "outputId": "9a635389-10ed-454c-ff3d-e1c2a384a073"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 434.485x360 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# data preprocessing\n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
        "\n",
        "from sklearn.preprocessing import StandardScaler as ss\n",
        "sc = ss()\n",
        "X_train = sc.fit_transform(X_train)\n",
        "X_test = sc.transform(X_test)"
      ],
      "metadata": {
        "id": "G8VMwJ_E8QBk"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#SVM\n",
        "from sklearn.svm import SVC\n",
        "classifier = SVC(kernel = 'rbf')\n",
        "classifier.fit(X_train, y_train)\n",
        "# Predicting the Test set results\n",
        "y_pred = classifier.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = classifier.predict(X_train)\n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "\n",
        "print()\n",
        "print('Accuracy for training set for svm = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for svm = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4Pvs7Jeu8T8S",
        "outputId": "2bae9e58-f221-4487-b9e8-803374499836"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for svm = 0.6694214876033058\n",
            "Accuracy for test set for svm = 0.5737704918032787\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#   Naive Bayes \n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
        "\n",
        "from sklearn.naive_bayes import GaussianNB\n",
        "classifier = GaussianNB()\n",
        "classifier.fit(X_train, y_train)\n",
        "\n",
        "\n",
        "# Predicting the Test set results\n",
        "y_pred = classifier.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = classifier.predict(X_train)\n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "\n",
        "print()\n",
        "print('Accuracy for training set for Naive Bayes = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for Naive Bayes = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5eWlXV4E8bgm",
        "outputId": "be661172-85ec-4f97-e54a-47576d4e2634"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for Naive Bayes = 0.8677685950413223\n",
            "Accuracy for test set for Naive Bayes = 0.7868852459016393\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Logistic Regression\n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
        "\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "classifier = LogisticRegression()\n",
        "classifier.fit(X_train, y_train)\n",
        "\n",
        "# Predicting the Test set results\n",
        "y_pred = classifier.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = classifier.predict(X_train)\n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "\n",
        "print()\n",
        "print('Accuracy for training set for Logistic Regression = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for Logistic Regression = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NDaMNadc8lW9",
        "outputId": "35108dd8-7c9b-49a6-8394-7c9c4eb86b8f"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for Logistic Regression = 0.8636363636363636\n",
            "Accuracy for test set for Logistic Regression = 0.8032786885245902\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/sklearn/linear_model/_logistic.py:818: 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",
            "  extra_warning_msg=_LOGISTIC_SOLVER_CONVERGENCE_MSG,\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#   Decision Tree\n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
        "\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "classifier = DecisionTreeClassifier()\n",
        "classifier.fit(X_train, y_train)\n",
        "\n",
        "# Predicting the Test set results\n",
        "y_pred = classifier.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = classifier.predict(X_train)\n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "\n",
        "print()\n",
        "print('Accuracy for training set for Decision Tree = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for Decision Tree = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "R3f0-2Je8qVc",
        "outputId": "5ce3f609-90b4-4c48-dff6-b6454bdad18c"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for Decision Tree = 1.0\n",
            "Accuracy for test set for Decision Tree = 0.7704918032786885\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#  Random Forest\n",
        "X = df.iloc[:, :-1].values\n",
        "y = df.iloc[:, -1].values\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)\n",
        "\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "classifier = RandomForestClassifier(n_estimators = 10)\n",
        "classifier.fit(X_train, y_train)\n",
        "\n",
        "# Predicting the Test set results\n",
        "y_pred = classifier.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = classifier.predict(X_train)\n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "\n",
        "print()\n",
        "print('Accuracy for training set for Random Forest = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for Random Forest = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gqnSraS0855H",
        "outputId": "66c41d2e-d230-4bd9-a675-edd1cc02c825"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for Random Forest = 1.0\n",
            "Accuracy for test set for Random Forest = 0.7704918032786885\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# applying lightGBM\n",
        "import lightgbm as lgb\n",
        "\n",
        "d_train = lgb.Dataset(X_train, label = y_train)\n",
        "params = {}\n",
        "\n",
        "clf = lgb.train(params, d_train, 100)\n",
        "#Prediction\n",
        "y_pred = clf.predict(X_test)\n",
        "#convert into binary values\n",
        "for i in range(0, len(y_pred)):\n",
        "    if y_pred[i]>= 0.5:       # setting threshold to .5\n",
        "       y_pred[i]=1\n",
        "    else:  \n",
        "       y_pred[i]=0\n",
        "       \n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = clf.predict(X_train)\n",
        "\n",
        "for i in range(0, len(y_pred_train)):\n",
        "    if y_pred_train[i]>= 0.5:       # setting threshold to .5\n",
        "       y_pred_train[i]=1\n",
        "    else:  \n",
        "       y_pred_train[i]=0\n",
        "       \n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "print()\n",
        "print('Accuracy for training set for LightGBM = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for LightGBM = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0l7eRfY989tp",
        "outputId": "2d4e7526-3994-4051-fe77-e685be5789ef"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for LightGBM = 0.9958677685950413\n",
            "Accuracy for test set for LightGBM = 0.7704918032786885\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# applying XGBoost\n",
        "\n",
        "#from sklearn.model_selection import train_test_split\n",
        "#X_train, X_test, y_train, y_test = train_test_split(X, target, test_size = 0.20, random_state = 0)\n",
        "\n",
        "from xgboost import XGBClassifier\n",
        "xg = XGBClassifier()\n",
        "xg.fit(X_train, y_train)\n",
        "y_pred = xg.predict(X_test)\n",
        "\n",
        "from sklearn.metrics import confusion_matrix\n",
        "cm_test = confusion_matrix(y_pred, y_test)\n",
        "\n",
        "y_pred_train = xg.predict(X_train)\n",
        "\n",
        "for i in range(0, len(y_pred_train)):\n",
        "    if y_pred_train[i]>= 0.5:       # setting threshold to .5\n",
        "       y_pred_train[i]=1\n",
        "    else:  \n",
        "       y_pred_train[i]=0\n",
        "       \n",
        "cm_train = confusion_matrix(y_pred_train, y_train)\n",
        "print()\n",
        "print('Accuracy for training set for XGBoost = {}'.format((cm_train[0][0] + cm_train[1][1])/len(y_train)))\n",
        "print('Accuracy for test set for XGBoost = {}'.format((cm_test[0][0] + cm_test[1][1])/len(y_test)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2m_P47jn9EeG",
        "outputId": "03f89a84-2a51-4ecd-df2b-8707c7fd7503"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Accuracy for training set for XGBoost = 0.987603305785124\n",
            "Accuracy for test set for XGBoost = 0.7540983606557377\n"
          ]
        }
      ]
    }
  ]
}