AutoMLExplainerSetupClass class

Definition

Placeholder class for all objects needed for interface with AzureML explain package.

AutoMLExplainerSetupClass(X_transform: typing.Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow, NoneType] = None, X_test_raw: typing.Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow, NoneType] = None, X_test_transform: typing.Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow, NoneType] = None, pipeline: typing.Union[sklearn.pipeline.Pipeline, azureml.train.automl.automl_explain_utilities.StreamingPipelineExplainabilityWrapper, NoneType] = None, estimator: typing.Union[sklearn.pipeline.Pipeline, azureml.train.automl.automl_explain_utilities.StreamingPipelineExplainabilityWrapper, NoneType] = None, featurizer: typing.Union[sklearn.pipeline.Pipeline, azureml.automl.core.featurization.streaming.streaming_featurization_transformer.StreamingFeaturizationTransformer, NoneType] = None, engineered_feature_names: typing.Union[typing.List[str], NoneType] = None, raw_feature_names: typing.Union[typing.List[str], NoneType] = None, feature_map: typing.Union[numpy.ndarray, pandas.core.frame.DataFrame, scipy.sparse.base.spmatrix, azureml.dataprep.api.dataflow.Dataflow, NoneType] = None, classes: typing.Union[typing.List[typing.Any], NoneType] = None)
Inheritance
builtins.object
AutoMLExplainerSetupClass

Parameters

X_transform
DataFrame or ndarray or scipy.sparse.csr_matrix

The featurized training features used for fitting pipelines during AutoML experiment.

X_test_raw
DataFrame or ndarray or scipy.sparse.csr_matrix

The raw test features used evaluating an AutoML trained pipeline.

X_test_transform
DataFrame or ndarray or scipy.sparse.csr_matrix

The featurized test features for evaluating an AutoML estimator.

pipeline
sklearn.pipeline

The entire fitted AutoML model.

estimator
sklearn.pipeline

The AutoML estimator including the model specific preprocessor and learner.

featurizer
sklearn.pipeline

The AutoML featurizer which does transformations from raw features to engineered features.

engineered_feature_names
List[str]

The list of names for the features generated by the AutoML featurizers.

raw_feature_names
List[Any]

The list of names for the raw features to be explained.

classes

The list of classes discovered in the labeled column in case of classification problem.

Attributes

X_test_raw

Return the raw test features used evaluating an AutoML trained pipeline.

Returns

The raw test features used evaluating an AutoML trained pipeline.

X_test_transform

Return the featurized test features for evaluating an AutoML estimator.

Returns

The featurized test features for evaluating an AutoML estimator.

X_transform

Return the featurized training features used for fitting pipelines during AutoML experiment.

Returns

The featurized training features used for fitting pipelines during AutoML experiment.

automl_estimator

Return the AutoML estimator including the model specific preprocessor and learner.

Returns

The AutoML estimator including the model specific preprocessor and learner.

automl_featurizer

Return the AutoML featurizer which does transformations from raw features to engineered features.

Returns

The AutoML featurizer which does transformations from raw features to engineered features.

automl_pipeline

Return the entire fitted AutoML model.

Returns

The entire fitted AutoML model.

classes

Return the list of classes discovered in the labeled column in case of classification problem.

Returns

The list of classes discovered in the labeled column in case of classification problem.

engineered_feature_names

Return the list of names for the features generated by the AutoML featurizers.

Returns

The list of names for the features generated by the AutoML featurizers.

feature_map

Return the mapping of which raw features generated which engineered features.

Returns

The mapping of which raw features generated which engineered features.

raw_feature_names

Return the list of names for the raw features to be explained.

Returns

The list of names for the raw features to be explained.