search

zelros / cinnamon

CinnaMon is a Python library which offers a number of tools to detect, explain, and correct data drift in a machine learning system
MIT License
76 stars 9 forks source link

what if there is no groundtruth? #15

Open zabir-nabil opened 1 year ago

zabir-nabil commented 1 year ago

can it calculate the drift if there is no groundtruth data / y_target?

YohannLeFaou commented 1 year ago

Looking for instance at this example: https://github.com/zelros/cinnamon/blob/master/examples/breast_cancer_XGBoost_ModelDriftExplainer.ipynb

You can fit the drift explainer without passing any value for y1 and y2:

drift_explainer.fit(X1=X_train, X2=X_test)

You can then analyze drift between X1 and X2, but methods drift_explainer.get_target_drift(), drift_explainer.get_performance_metrics_drift() will not work.