PunditKit: A GUI for Scikit-Learn Models
Simplify. Visualise. Learn.
PunditKit is a free toolkit for exploratory data analysis and modelling of tabular data - such as a database or a spreadsheet - with a simple interface, visual diagnostics, and the choice of a large number of different models.
- Machine learning models using scikit-learn
- Interpretable explanations of predictions using lime
- Exploratory data summaries for checking datasets
- Model diagnostics for evaluating the effectiveness of different models
- Feature importance: identify key drivers of the response
- Partial dependence plots: understand relationships with a particular feature.
PunditKit is under active development. The goal is to develop an opinionated modelling framework with best practice modelling and visualisation techniques. If you encounter any problems or have any feature requests, please consider raising an issue.
Installation (via pip)
PunditKit is developed using Python. First download and install a Python 3.x distribution such as Anaconda
PunditKit can then be installed using
pip from the command line (if
is added to PATH during installation) or using Anaconda prompt.
pip install punditkit
This adds the
Modelling a dataset
Suppose you have a file called
iris.csv that you would like to model.
To use punditkit on the dataset, run:
Currently only Comma Separated Values (CSV) datasets are supported. Datasets within Excel spreadsheets can be analysed using punditkit, use Save As CSV from within Excel to convert to the right format.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Hashes for punditkit-0.0.2-py3-none-any.whl