cardea.modeling.Modeler

class cardea.modeling.Modeler(pipeline, problem_type)[source]

A class responsible for executing various Machine Learning Pipelines using MLBlocks.

__init__(pipeline, problem_type)[source]

Initialize self. See help(type(self)) for accurate signature.

Methods

__init__(pipeline, problem_type)

Initialize self.

evaluate(X, y[, test_size, shuffle, tune, …])

Evaluate the pipelines.

fit(X, y[, tune, max_evals, scoring, verbose])

Fit and select the pipelines.

fit_predict(X, y[, tune, max_evals, …])

Fit the pipeline and make predictions

k_fold_validation(hyperparameters, X, y[, …])

Score the pipeline through k-fold validation with the given scoring function.

load(path)

Load a Modeler object from a pickle file

predict(X)

Predict the input data

save(path)

Save the object in a pickle file.

test(X, y[, scoring])

Test the trained pipeline.

train_test_split(X, y[, test_size, shuffle])

Split the training dataset and the testing dataset.

tune(X, y[, max_evals, scoring, verbose])

Tune the pipeline hyper-parameters and select the optimized model.

Attributes

classification_metrics

Supported classification metrics functions.

pipeline

Pipeline.

regression_metrics

Supported regression metrics functions.

target_metrics

Supported metrics functions for the given problem type.