tempor.models.samplers module¶

Custom torch samplers.

class tempor.models.samplers.BaseSampler(data_source: Sized | None = None)[source]¶

Bases: Sampler

DataSampler samples the conditional vector and corresponding data.

get_dataset_conditionals() → ndarray | None[source]¶

sample_conditional(batch: int, **kwargs: Any) → tuple | None[source]¶

sample_conditional_for_class(batch: int, c: int) → ndarray | None[source]¶

conditional_dimension() → int[source]¶

Return the total number of categories.

conditional_probs() → ndarray | None[source]¶

Return the total number of categories.

train_test() → tuple[source]¶

class tempor.models.samplers.ImbalancedDatasetSampler(labels: list, train_size: float = 0.8)[source]¶

Bases: BaseSampler

Samples elements randomly from a given list of indices for imbalanced dataset.

train_test() → tuple[source]¶

Return the train and test indices.

Returns:¶

(self.train_idx, self.test_idx)

Return type:¶

Tuple