Welcome to Shapash’s documentation !¶

Shapash is a Python library which aims to make machine learning interpretable and understandable to everyone. Shapash provides several types of visualization which displays explicit labels that everyone can understand. Data Scientists can more easily understand their models and share their results. End users can understand the decision proposed by a model using a summary of the most influential criteria. The project was developed by MAIF Data Scientists.

Company	GitHub Account	Website
	MAIF GitHub	MAIF

The objectives of Shapash:¶

To display clear and understandable results: Plots and outputs use explicit labels for each feature and its modalities:

_images/shapash-contribution_plot-example.png

To allow Data Scientists to quickly understand their models by using a webapp to easily navigate between global and local explainability, and understand how the different features contribute: Live Demo shapash-monitor

To Summarize and export the local explanation: Shapash proposes a short and clear local explanation. It allows each user, whatever their Data backround is, to understand a local prediction of a supervised model, thanks to a summarized and straightforward explanation

summary_df.head()

pred	proba	feature_1	value_1	contribution_1	feature_2	value_2	contribution_2	feature_3	value_3	contribution_3	feature_4	value_4	contribution_4	feature_5	value_5	contribution_5	feature_6	value_6	contribution_6	feature_7	value_7	contribution_7
Moderately Expensive	0.9949	Ground living area square feet	1792	0.3093	Interior finish of the garage?	Rough Finished	0.2755	Size of garage in square feet	564	0.2077	Full bathrooms above grade	2	0.1827	Physical locations within Ames city limits	College Creek	0.1709	Overall material and finish of the house	7	0.1640	Height of the basement	Good (90-99 inches)	0.1396
Moderately Expensive	0.8769	Second floor square feet	720	0.1833	Full bathrooms above grade	2	0.1551	Ground living area square feet	2192	0.1519	Remodel date	1997	0.1431	Type 1 finished square feet	378	0.1424	First Floor square feet	1052	0.1278	Half baths above grade	1	0.1277
Cheap	0.9973	Ground living area square feet	900	0.8189	Size of garage in square feet	280	0.5616	Total square feet of basement area	882	0.4091	Remodel date	1967	0.3490	Full bathrooms above grade	1	0.3248	Overall material and finish of the house	5	0.3180	First Floor square feet	900	0.2478
Cheap	0.9987	Ground living area square feet	630	0.8164	Size of garage in square feet	0	0.5877	Total square feet of basement area	630	0.4312	Remodel date	1970	0.3557	Overall material and finish of the house	4	0.3175	Full bathrooms above grade	1	0.3130	General zoning classification	Residential Medium Density	0.1784
Cheap	0.8524	Ground living area square feet	1188	0.9421	Remodel date	1959	0.4234	Overall material and finish of the house	5	0.3785	Full bathrooms above grade	1	0.3738	Number of fireplaces	0	0.1687	Rating of basement finished area	Average Rec Room	0.1302	Wood deck area in square feet	0	0.1225

To freeze different aspects of a data science project as a basis of an audit report

To discuss results: Shapash allows Data Scientists to easily share and discuss their results with non-Data users

Shapash features:¶

Compatible with Shap & Lime
Uses Shap backend to display results in a few lines of code
Encoders objects and features dictionaries used for clear results
Compatible with category_encoders & Sklearn ColumnTransformer
Visualizations of global and local explainability
Webapp to easily navigate from global to local
Select subsets for further analysis of explainability by filtering on explanatory and additional features, correct or wrong predictions
Summarizes local explanation
Offers several parameters in order to sum up in the most suitable way for your use case
Exports your local summaries to Pandas DataFrames
Usable for Regression, Binary Classification or Multiclass
Compatible with most of sklearn, lightgbm, catboost, xgboost models
Relevant for exploration and also deployment (through an API or in Batch mode) for operational use
Freeze different aspects of a data science project as a basis of an audit report

Shapash is easy to install and use: It provides a SmartExplainer class to understand your model and summarize explanation with a simple syntax.

High adaptability: Very few arguments are required to display results. But the more you work on cleaning and documenting the data, the clearer the results will be for the end user.

License is Apache Software License 2.0