pyoe package

Subpackages

Submodules

pyoe.preprocessors module

class pyoe.preprocessors.Preprocessor(missing_fill: Literal['knn', 'regression', 'avg', 'zero'] = 'zero')

Bases: object

fill(x: Tensor) → Tensor

This function fills missing values in a dataset

Parameters:

x (torch.Tensor) – the input tensor with missing values.

Returns:

the input tensor with missing values filled.

Return type:

out (torch.Tensor)

pyoe.preprocessors.fill_missing_value(window_x: Tensor, missing_fill: Literal['knn', 'regression', 'avg', 'zero']) → Tensor

this function provides a few methods to fill missing values in a dataset

Parameters:

  • window_x (torch.Tensor) – the input tensor with missing values.

  • missing_fill (Literal["knn", "regression", "avg", "zero"]) – the method to fill missing values.

pyoe.utils module

pyoe.utils.shingle(series: array, dim: int) → array

Takes a one dimensional series and shingles it into dim dimensions.

Parameters:

  • series (np.array) – the input series.

  • dim (int) – the dimension of the shingled array.

Returns:

the shingled array.

Return type:

shingled (np.array)

Module contents