gitter-badger/modeltools

Python tools for model exploration

ModelTools

Included is a set of Python tools to do basic data exploration and model analysis. These tools are broken up into two main parts, exploratory data analysis and model/variable selection. The exploratory data analysis component has tools to analyze the distributions of your variables and correlations among them. The model comparison tools include univariate analysis, model comparison, and variable selection. Note: this is still in development and the code has not been thoroughly tested. Unit tests still need to be built. A large source of motivation and code is from this insightful Kaggle kernel. The goal here is to expand the set of tools and generalize the code so that it can easily be used for other supervised learning problems.

Usage

There are two pairs of notebooks included to demonstrate typical usage. There is one example of classification and one of regression. The data sets can be downloaded from Kaggle, classification and regression. A .yml file is included for any dependencies.