Metadata-Version: 2.1
Name: pyregression
Version: 0.0.8
Summary: Machine learning regressions methods
Home-page: https://gitlab.com/pyregression/pyregression
Author: Emiliano Lopez
Author-email: emiliano.lopez@gmail.com
License: UNKNOWN
Description: # PyRegression
        
        PyRegression is a python library for using in machine learning regressions. The first minimal launched version include nonlinear regression with K-Nearest Neighbor (KNN), Random Forest (RFO) and Support Vector Regression (SVR). Of course, you can use PyRegression for Multiple Linear Regression (MLR) too.
        
        The nonlinear methods include the searching of the best hyperparameters using k-fold cross validation. For more details visit the official documentation in pyregression.gitlab.io
        
        ## Install
        
        ``pip install pyregression``
        
        or if you prefer (recommended) install in a virtual environment:
        
        ```bash
        python3 -m venv venv-pyregression
        cd venv-pyregression
        source bin/activate
        
        pip install pyregression
        # install dependencies
        pip install pandas
        pip install sklearn
        ```
        
        ## Example code
        
        ```python
        # import regression methods
        from pyregression.regressors import MLR, KNN, RFO, SVR
        
        # import dataset example
        from pyregression.datasets import load_soilmoisture
        
        # load dataset example
        X,y = load_soilmoisture() 
        
        # do the regressions (test_size = 0.6 by default)
        mlr = MLR(X,y)
        knn = KNN(X,y)
        rfo = RFO(X,y)
        svr = SVR(X,y)
        
        # print predicted values
        for regression in (mlr, knn, rfo, svr):
            print(regression["y_pred"])
        ```
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
