PermutationInterpreter
- class eipy.interpretation.PermutationInterpreter(EI, metric=None, ensemble_predictor_keys='all', n_repeats=10, n_jobs=1, metric_greater_is_better=True)
Permuation importance based interpreter.
This method utilizes sklearn’s permutation_importance function.
- EIEnsembleIntegration class object
Fitted EnsembleIntegration model, i.e. with model_building=True.
- metricfunction
sklearn-like metric function. If None, the fmax score is used.
- n_repeatsint, default=10
Number of repeats in PermutationImportance.
- ensemble_predictor_keys: default=’all’
Ensemble predictor keys used in EnsembleIntegration. If ‘all’ then all ensemble predictors seen by EI are interpreted. Recommended to pass a subset of ensemble_predctor keys as a list.
- metric_greater_is_better: default=True
Metric greater is better.
- Returns
- self
Feature rankings of final ensemble models trained with EnsembleIntegration.
- Attributes
- ensemble_feature_rankingpandas.DataFrame
Feature rankings for each ensemble method.
- LFRpandas.DataFrame
Local feature rankings for each base predictor.
- LMRpandas.Dataframe
self.LMR = None
Methods
local_feature_rank(X_dict, y)Local Feature Ranks (LFRs) for each base predictor
local_model_rank(ensemble_predictor_keys)Local Model Ranks (LMRs)
rank_product_score(X_dict, y)Compute feature ranking of ensemble methods using LFR and LMR.
- rank_product_score(X_dict, y)
Compute feature ranking of ensemble methods using LFR and LMR.
- Parameters
- X_dictdict
Dictionary of X modalities. Keys and n_features must match those seen by EnsembleIntegration.fit_base().
- yarray of shape (n_samples,)
Target vector relative to X.
- Returns
- self
Feature ranking of ensemble methods
- local_feature_rank(X_dict, y)
Local Feature Ranks (LFRs) for each base predictor
- Parameters
- X_dictdict
Dictionary of X modalities. Keys and n_features must match those seen by EnsembleIntegration.fit_base().
- yarray of shape (n_samples,)
Target vector relative to X.
- Returns
- self
Local feature ranks.
- local_model_rank(ensemble_predictor_keys)
Local Model Ranks (LMRs)
- Parameters
- ensemble_predictor_keyslist of str
List of ensemble predictor keys that will be used to select ensembles classifiers to interpret.
- Returns
- self
Local model ranks.