explaineval 0.2.5

Unified evaluation and explainability framework for ML models

📊 ExplainEval

ExplainEval is a unified evaluation and explainability framework for machine learning models. It supports classification, regression, NLP, and time series tasks, providing a consistent API to evaluate models, explain predictions, and generate interactive reports.

---

🚀 Features

• ✅ Easy evaluation for Classification, Regression, NLP, and Time Series models
• 🔍 Built-in explainability using SHAP and LIME
• 📊 Confusion matrix, ROC curves, MAE, RMSE, and more
• 📋 Auto-generated HTML reports for stakeholders
• 🔁 Model comparison interface
• 📦 Compatible with scikit-learn, XGBoost, LightGBM, Transformers

---

📦 Installation

Install via PyPI:

pip install explaineval

Or install from source:

git clone https://github.com/yourname/explaineval.git
cd explaineval
pip install -e .

---

🔧 Supported Tasks

Task Type	Models Supported	Explainability
Classification	scikit-learn, XGBoost, LightGBM, etc.	SHAP, LIME
Regression	Linear, Tree-based, Boosting, etc.	SHAP
NLP	Sklearn Pipelines, Transformers	SHAP, Attention
Time Series	ARIMA, LSTM, XGBoost, etc.	SHAP

---

🧠 Usage Example

🔍 Classification Example

from explaineval.main import EvalX
from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split

X, y = load_iris(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
model = RandomForestClassifier().fit(X_train, y_train)

evalx = EvalX(model, task="classification", background_data=X_train)
evalx.evaluate(X_test, y_test)
evalx.explain(X_test[:5])
evalx.generate_report("classification_report.html")

📈 Regression Example

from sklearn.datasets import load_diabetes
from sklearn.ensemble import GradientBoostingRegressor

X, y = load_diabetes(return_X_y=True)
model = GradientBoostingRegressor().fit(X, y)
evalx = EvalX(model, task="regression", background_data=X)
evalx.evaluate(X, y)
evalx.explain(X[:5])
evalx.generate_report("regression_report.html")

---

📋 Auto Reports

Call generate_report("filename.html") to save a clean HTML report including:

• Task summary
• Evaluation metrics
• SHAP/LIME explanation visualizations

---

🧪 Model Comparison

from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier

model1 = RandomForestClassifier().fit(X_train, y_train)
model2 = GradientBoostingClassifier().fit(X_train, y_train)

results = evalx.compare([model1, model2], X_test, y_test)
print(results)

---

📄 License

MIT License © 2025 Shaheer Zaman Khan

---

🤝 Contributing

We welcome contributions! To contribute:

1. Fork the repo
2. Create your feature branch (git checkout -b feature/YourFeature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin feature/YourFeature)
5. Open a pull request

---

📬 Contact

Have questions or feedback? Open an issue or contact us at [e-mail]