choose-models

📦 Models Class – Regression Models Playground

This class helps you quickly test different regression algorithms (OLS, SGD, BGD) on any DataFrame and target.

✂️ split_data(ratio=0.8, randomState=42)

Splits data into train/test based on ratio. Uses self.target_column to separate X and y.

📈 Linear_Regression_OLS(get_equation=False, plot=False, accuracy=True)

Trains simple (1D) Linear Regression using Ordinary Least Squares.

• get_equation: print learned line
• plot: visualize
• accuracy: print score

📊 MLinear_Regression_OLS(get_equation=False, plot=False, accuracy=True)

Multi-feature version of OLS Linear Regression.

⚡ MLinear_Regression_SGD(epochs=100, learning_rate=0.01, plot=False, accuracy=True, get_equation=False)

Multi-feature SGD Linear Regression. Trains with Stochastic Gradient Descent.

🌀 Linear_Regression_BGD(epochs=100, learning_rate=0.01, plot=False, accuracy=True, get_equation=False)

Simple (1D) Linear Regression using Batch Gradient Descent.

💪 MLinear_Regression_BGD(epochs=100, learning_rate=0.01, plot=False, accuracy=True, get_equation=False)

Multi-feature BGD-based Linear Regression.

🧼 Standard_Scale(features=None)

Standardizes features (z-score normalization). Applies to entire DataFrame if no features are specified. Re-splits data after scaling.

🚀 Linear_Regression_SGD(epochs=100, learning_rate=0.01, plot=False, accuracy=True, get_equation=False)

Simple (1D) Linear Regression using SGD.

📤 Extract_Data()

Returns (X_train, X_test, y_train, y_test) — useful for external use.

🆕 Set_DF(newDF, target_column)

Reset the class with a new DataFrame and target column.

🔀 Select_Model(model=None, batch_size=None, method=None, epochs=100, learning_rate=0.01, plot=False, accuracy=True, get_equation=False)

Unified interface to select and run any regression model or method.

• model: Choose 'linear_regression' ('lr') or 'multi_linear_regression' ('mlr')
• method: Specify algorithm ('ols', 'bgd', 'sgd', 'mgbd')
• batch_size: For mini-batch gradient descent ('mgbd')
• Other arguments are passed to the underlying method

Raises ValueError if model or method is unknown.