bootstrap module¶

A bootstrap tool for resampling and surrogate modeling.

class gemseo.mlearning.resampling.bootstrap.Bootstrap(sample_indices, n_replicates=100, seed=0)[source]¶

Bases: Resampler

A bootstrap tool for resampling and surrogate modeling.

Parameters:

sample_indices (NDArray[int]) – The original indices of the samples.
n_replicates (int) –
The number of bootstrap replicates.

By default it is set to 100.
seed (int | None) –
The seed to initialize the random generator. If None, then fresh, unpredictable entropy will be pulled from the OS.

By default it is set to 0.

execute(model, return_models, predict, stack_predictions, fit_transformers, store_sampling_result, input_data, output_data_shape)¶

Apply the resampling technique to a machine learning model.

Parameters:

model (MLAlgo) – The machine learning model.
return_models (bool) – Whether the sub-models resulting from resampling are returned.
predict (bool) – Whether the sub-models resulting from sampling do prediction on their corresponding learning data.
stack_predictions (bool) – Whether the sub-predictions are stacked.
fit_transformers (bool) – Whether to re-fit the transformers.
store_sampling_result (bool) – Whether to store the sampling results in the attribute resampling_results of the original model.
input_data (ndarray) – The input data.
output_data_shape (tuple[int, ...]) – The shape of the output data array.

Returns:

First the sub-models resulting from resampling if return_models is True then the predictions, either per fold or stacked.

Raises:

ValueError – When the model is neither a supervised algorithm nor a clustering one.

Return type:

tuple[list[MLAlgo], list[ndarray] | ndarray]

property n_replicates: int¶: The number of bootstrap replicates.

name: str¶

The name of the resampler.

Use the class name by default.

property sample_indices: NDArray[int]¶: The indices of the samples after shuffling.

property seed: int¶: The seed to initialize the random generator.

property splits: Splits¶

The train-test splits resulting from the splitting of the samples.

A train-test split is a partition whose first component contains the indices of the learning samples and the second one the indices of the test samples.