PyInterpret: A Unified Python Library for Machine Learning Model Interpretation
PyInterpret is a comprehensive Python library that unifies fragmented explainability tools under one consistent API. It provides modular coverage of both global and local explanations across different data modalities, consolidating capabilities from state-of-the-art tools like SHAP, LIME, and others.
🎯 Key Features
- Unified API: Consistent interface across all interpretation methods
- Local Attribution: SHAP, LIME, and other instance-level explanations
- Global Insights: Permutation importance, partial dependence plots
- Modular Architecture: Easy extension and customization
- Multiple Data Types: Support for tabular, text, image, and time-series data
- Framework Integration: Works seamlessly with scikit-learn, pandas, and other ML libraries
- Professional Quality: Comprehensive testing, documentation, and error handling
🚀 Quick Start
Installation
# Basic installation
pip install pyinterpret
# With SHAP support
pip install pyinterpret[shap]
# With LIME support
pip install pyinterpret[lime]
# With all optional dependencies
pip install pyinterpret[all]
Basic Usage
from pyinterpret import SHAPExplainer, LIMEExplainer, PermutationImportanceExplainer
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_classification
import pandas as pd
# Create sample data and train model
X, y = make_classification(n_samples=1000, n_features=10, random_state=42)
X_df = pd.DataFrame(X, columns=[f'feature_{i}' for i in range(X.shape[1])])
model = RandomForestClassifier(random_state=42)
model.fit(X_df, y)
# Local explanation with SHAP
shap_explainer = SHAPExplainer(model, explainer_type='tree')
shap_result = shap_explainer.explain_instance(X_df.iloc[0])
print("SHAP attributions:", shap_result.attributions)
print("Feature names:", shap_result.feature_names)
# Global explanation with Permutation Importance
perm_explainer = PermutationImportanceExplainer(model, scoring='accuracy')
perm_result = perm_explainer.explain_global(X_df, y)
print("Most important features:", perm_result.feature_names[:5])
print("Importance scores:", perm_result.attributions[:5])
1 maintainer
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
pyinterpret-0.1.1.tar.gz
(353.1 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file pyinterpret-0.1.1.tar.gz.
File metadata
- Download URL: pyinterpret-0.1.1.tar.gz
- Upload date:
- Size: 353.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
1a6eaa39aa9d8017b2125c84185b9e870f8ed0ea35d97ae91ae551d65fff774d
|
|
| MD5 |
d6230a53db4dbfc6a7dafe0aa11b65e7
|
|
| BLAKE2b-256 |
00a4225145571cab032f97f1ad25e11fe0eb9a071c2baf22d0d0df4f7be87bbb
|
Provenance
The following attestation bundles were made for pyinterpret-0.1.1.tar.gz:
Publisher:
publish.yml on mowne67/pyinterpret
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
pyinterpret-0.1.1.tar.gz -
Subject digest:
1a6eaa39aa9d8017b2125c84185b9e870f8ed0ea35d97ae91ae551d65fff774d - Sigstore transparency entry: 2617618246
- Sigstore integration time:
-
Permalink:
mowne67/pyinterpret@e8bc986a02a0e4aa67e24cfe61a59bcc57aa1033 -
Branch / Tag:
refs/tags/v0.1.1 - Owner: https://github.com/mowne67
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@e8bc986a02a0e4aa67e24cfe61a59bcc57aa1033 -
Trigger Event:
release
-
Statement type:
File details
Details for the file pyinterpret-0.1.1-py3-none-any.whl.
File metadata
- Download URL: pyinterpret-0.1.1-py3-none-any.whl
- Upload date:
- Size: 31.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b11128a1ed7887b8c69696bf85ae07e8af752f5ba0dc9c55eea239ad2ce0abf5
|
|
| MD5 |
d0dfadea19832cc9473353f37fe74509
|
|
| BLAKE2b-256 |
8c1291c6e7a953bf589f67dbb98a01f9b1c15789b41efbc05e72800bfc7a43e2
|
Provenance
The following attestation bundles were made for pyinterpret-0.1.1-py3-none-any.whl:
Publisher:
publish.yml on mowne67/pyinterpret
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
pyinterpret-0.1.1-py3-none-any.whl -
Subject digest:
b11128a1ed7887b8c69696bf85ae07e8af752f5ba0dc9c55eea239ad2ce0abf5 - Sigstore transparency entry: 2617618253
- Sigstore integration time:
-
Permalink:
mowne67/pyinterpret@e8bc986a02a0e4aa67e24cfe61a59bcc57aa1033 -
Branch / Tag:
refs/tags/v0.1.1 - Owner: https://github.com/mowne67
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@e8bc986a02a0e4aa67e24cfe61a59bcc57aa1033 -
Trigger Event:
release
-
Statement type:
