regresseasy 0.6

Regression Model Performances

RegressModel

RegressModel is a Python package designed for streamlined regression analysis. It facilitates the process of training and evaluating multiple regression models on a dataset, providing a comprehensive comparison of their performance. This package is particularly useful for data scientists and machine learning practitioners who need to quickly assess the best regression model for their specific data.

Features

• Easy comparison of various regression models.
• Includes common models like Linear Regression, Lasso, Ridge, SVR, Random Forest, Decision Tree, and XGBoost.
• Outputs performance metrics like R2 score, RMSE, MSE, MAE, and cross-validation scores.

Installation

You can install RegressModel using pip:

pip install regressmodel

Usage

Here is a simple example of how to use RegressModel:

from regressmodel.core import RegModelling
import pandas as pd
from sklearn.model_selection import train_test_split

# Sample dataset
df = pd.read_csv('your_dataset.csv')
X = df.drop('target_column', axis=1)
y = df['target_column']

# Splitting the dataset
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

# Example usage of the function
# Ensure that you have defined X_train, y_train, X_test, y_test before calling this function
model_results = reg_modelling(X_train, y_train, X_test, y_test)

# # Access specific model results, e.g., Linear Regression
print("Linear Regression Results:", model_results["LinearRegression"])


# The results variable will contain performance metrics of the models

Requirements

Python 3.6 or higher scikit-learn numpy xgboost

Contributing

Contributions to RegressModel are welcome! Feel free to fork the repository and submit pull requests.

License

RegressModel is licensed under the MIT License - see the LICENSE file for details.

Author

SaiKoushikGandikota