Linear and Logistic Regression

linear_logistic_regression.py

@@ -0,0 +1,218 @@
+import numpy as np
+import matplotlib.pyplot as plt
+import seaborn as sns
+from sklearn.linear_model import LogisticRegression
+from sklearn.model_selection import train_test_split
+from sklearn.preprocessing import StandardScaler
+from sklearn.linear_model import LinearRegression
+from sklearn.metrics import classification_report
+from sklearn.metrics import mean_squared_error, r2_score
+import pandas as pd
+
+
+#to import from csv file
+def load_csv(path):
+    df = pd.read_csv(path)
+    print(df.head())
+    print(df.describe())
+    print(f"Missing values:\n{df.isnull().sum()}")
+    return df
+
+#normalization
+class LinearRegressionSupervisedML():
+
+    def normalization(self,X):
+        """
+        ARG:
+        :param X: X is the input data
+        :return: normalized X
+        """
+        scaler = StandardScaler()
+        X_normalized = scaler.fit_transform(X)
+        return X_normalized
+
+
+    def LinearRegression(self,X,y,test=0.2):
+        """
+        Arguments:
+        :param X(ndarray(m,)): input data
+        :param y: output data
+        :param test: size of test set
+        :return: model, x_train, x_test,y_train, y_test
+        """
+        X_normalized = self.normalization(X)
+        X_train, X_test, y_train, y_test = train_test_split(X_normalized, y, test_size=test, random_state=42)
+        model = LinearRegression()
+        model.fit(X_train, y_train)
+        return model, X_train, X_test, y_train, y_test
+
+    def PlotlinearRegression(self, model, X_test, y_test, feature_name='Feature'):
+        """
+        Args----
+            model: Trained LinearRegression model
+            X_test (np.array): Test features
+            y_test (np.array): Actual test targets
+            feature_name (str): Label for x-axis
+        """
+        y_pred = model.predict(X_test)
+
+        plt.figure(figsize=(10, 6))
+        plt.scatter(X_test, y_test, color='blue', label='Actual', alpha=0.7)
+        plt.plot(X_test, y_pred, color='red', linewidth=2, label='Prediction')
+
+        # Showing errors with dashed lines...
+        for i in range(len(X_test)):
+            plt.plot([X_test[i], X_test[i]], [y_test[i], y_pred[i]], color='gray', linestyle='--', linewidth=0.5)
+
+        plt.title('Linear Regression: Actual vs. Predicted')
+        plt.xlabel(feature_name)
+        plt.ylabel('Target')
+        plt.legend()
+        plt.show()
+        #print report
+        mse = mean_squared_error(y_test, y_pred)
+        r2 = r2_score(y_test, y_pred)
+        print(f"Mean Squared Error: {mse:.4f}")
+        print(f"R-squared Score: {r2:.4f}")
+
+    def plot_train_test_split(self, X_train, X_test, y_train, y_test, feature_name='Feature'):
+        """
+        Args----
+            X_train (np.array): Train features
+            X_test (np.array): Test features
+            y_train (np.array): Train targets
+            y_test (np.array): Test targets
+            feature_name (str): Label for x-axis
+        """
+        plt.figure(figsize=(10, 6))
+        plt.scatter(X_train, y_train, color='blue', label='Train', alpha=0.7)
+        plt.scatter(X_test, y_test, color='green', label='Test', alpha=0.7)
+        plt.title('Train vs. Test Split')
+        plt.xlabel(feature_name)
+        plt.ylabel('Target')
+        plt.legend()
+        plt.show()
+
+class LogisticRegressionSupervisedML():
+    def normalization(self,X):
+        """
+        ARG:
+        :param X: X is the input data
+        :return: normalized X
+        """
+        scaler = StandardScaler()
+        X_normalized = scaler.fit_transform(X)
+        return X_normalized
+
+    def LogisticRegression(self,X,y,test=0.2):
+        """
+        Arguments:
+        :param X(ndarray(m,)): input data
+        :param y: output data
+        :param test: size of test set
+        :return: model, x_train, x_test,y_train, y_test
+        """
+        X_normalized = self.normalization(X)
+        X_train, X_test, y_train, y_test = train_test_split(X_normalized, y, test_size=test, random_state=42)
+        model = LogisticRegression()
+        model.fit(X_train, y_train.ravel())
+        return model, X_train, X_test, y_train, y_test
+
+    def PlotlogisticRegression(self, model, X_test, y_test, feature_name='Feature'):
+        """
+        Args----
+            model: Trained LinearRegression model
+            X_test (np.array): Test features
+            y_test (np.array): Actual test targets
+            feature_name (str): Label for x-axis
+        """
+        y_test = y_test.ravel()  # Ensure it's 1D
+        y_prob = model.predict_proba(X_test)[:, 1]
+        y_pred = (y_prob >= 0.5).astype(int)
+
+        # Sort for smoother logistic curve
+        sorted_idx = X_test.flatten().argsort()
+        X_sorted = X_test.flatten()[sorted_idx]
+        y_prob_sorted = y_prob[sorted_idx]
+
+        plt.figure(figsize=(10, 6))
+        plt.scatter(X_test, y_test, color='blue', label='Actual Labels', alpha=0.6)
+        plt.plot(X_sorted, y_prob_sorted, color='red', label='Logistic Curve', linewidth=2)
+
+        for i in range(len(X_test)):
+            x_val = float(X_test[i][0])  # or np.squeeze(X_test[i])
+            plt.plot([x_val, x_val], [y_test[i], y_prob[i]], color="gray", linestyle='--', linewidth=0.5)
+
+        plt.title('Logistic Regression: Actual vs. Predicted Probability')
+        plt.xlabel(feature_name)
+        plt.ylabel('Probability / Label')
+        plt.legend()
+        plt.grid(True)
+        plt.show()
+
+        # Print classification report
+        report = classification_report(y_test, y_pred)
+        print(report)
+
+    def plot_train_test_split(self, X_train, X_test, y_train, y_test, feature_name='Feature'):
+        """
+        Args----
+            X_train (np.array): Train features
+            X_test (np.array): Test features
+            y_train (np.array): Train targets
+            y_test (np.array): Test targets
+            feature_name (str): Label for x-axis
+        """
+        plt.figure(figsize=(10, 6))
+        plt.scatter(X_train, y_train, color='blue', label='Train', alpha=0.7)
+        plt.scatter(X_test, y_test, color='green', label='Test', alpha=0.7)
+        plt.title('Train vs. Test Split')
+        plt.xlabel(feature_name)
+        plt.ylabel('Target')
+        plt.legend()
+        plt.show()
+
+
+if __name__ == '__main__':
+    print("<--- Running Linear Regression Example --->")
+
+    # Generating simple sample data..
+    np.random.seed(0)
+    X_linear = 4 * np.random.rand(100, 1)
+    y_linear = 5 + 2 * X_linear + np.random.randn(100, 1)
+
+    # Initialize our ML class.
+    ml_module = LinearRegressionSupervisedML()
+
+
+    # Training the model..
+    linear_model, X_train_lin, X_test_lin, y_train_lin, y_test_lin = ml_module.LinearRegression(X_linear, y_linear)
+
+    # Visualizing the data split ...
+    ml_module.plot_train_test_split(X_train_lin, X_test_lin, y_train_lin, y_test_lin, feature_name='X')
+
+    # Ploting model predictions ..
+    ml_module.PlotlinearRegression(linear_model, X_test_lin, y_test_lin, feature_name='X')
+
+if __name__ == '__main__':
+    print("<--- Running Logistic Regression Example --->")
+
+    # Generating simple sample data..
+    np.random.seed(0)
+    X_linear = 4 * np.random.rand(100, 1)
+    y_bin = 5 + 2 * X_linear + np.random.randn(100, 1)
+    y_linear = (y_bin > y_bin.mean()).astype(int)
+    # Initialize our ML class.
+    ml_module = LogisticRegressionSupervisedML()
+
+
+    # Training the model..
+    linear_model, X_train_lin, X_test_lin, y_train_lin, y_test_lin = ml_module.LogisticRegression(X_linear, y_linear)
+
+    # Visualizing the data split ...
+    ml_module.plot_train_test_split(X_train_lin, X_test_lin, y_train_lin, y_test_lin, feature_name='X')
+
+    # Ploting model predictions ..
+    ml_module.PlotlogisticRegression(linear_model, X_test_lin, y_test_lin, feature_name='X')
+
+