_transforms

The old transforms.py; it is deprecated

Attributes

read_scaler
logger

Classes

NormalizeState_Quantile	Class to use the Quantile scaler functionality.
NormalizeState	Class to normalize state.
NormalizeTendency	Normalize tendency.
ToTensor	Convert variables from xr.Datasets to Pytorch Tensors.

Module Contents

_transforms.read_scaler = None

_transforms.logger

class _transforms.NormalizeState_Quantile(conf)

Class to use the Quantile scaler functionality.

scaler_file

variables

surface_variables

levels

scaler_df

scaler_3ds

scaler_surfs

scaler_3d

scaler_surf

__call__(sample: credit.data.Sample, inverse: bool = False) → credit.data.Sample

Normalize via quantile transform.

Normalize via provided scaler file/s.

Parameters:

• sample – batch.

• inverse – if true, will inverse the transform.

Returns:

transformed type.

Return type:

torch.tensor

inverse_transform(x: torch.Tensor) → torch.Tensor

Inverse transform.

Inverse transform.

Parameters:

x – batch.

Returns:

inverse transformed x.

transform(sample: Dict[str, numpy.ndarray]) → Dict[str, numpy.ndarray]

Transform.

Transform.

Parameters:

sample – batch.

Returns:

transformed batch.

class _transforms.NormalizeState(conf)

Class to normalize state.

mean_ds

std_ds

variables

surface_variables

levels

__call__(sample: credit.data.Sample, inverse: bool = False) → credit.data.Sample

Normalize via quantile transform.

Normalize via provided scaler file/s.

Parameters:

• sample – batch.

• inverse – if true, will inverse the transform.

Returns:

transformed type.

Return type:

torch.tensor

transform_dataset(DS: xarray.Dataset) → xarray.Dataset

transform_array(x: torch.Tensor) → torch.Tensor

Transform from unscaled to scaled values.

Transform.

Parameters:

x – batch.

Returns:

transformed x.

transform(sample: Dict[str, numpy.ndarray]) → Dict[str, numpy.ndarray]

Transform from unscaled to scaled values.

Transform.

Parameters:

sample – batch.

Returns:

transformed sample.

inverse_transform(x: torch.Tensor) → torch.Tensor

Inverse transform between tensor forms.

Inverse transform.

Parameters:

x – batch.

Returns:

inverse transformed x.

class _transforms.NormalizeTendency(variables, surface_variables, base_path)

Normalize tendency.

variables

surface_variables

base_path

mean

std

transform(tensor, surface_tensor)

Transform.

Transform input tensor/s.

Parameters:

• tensor (torch tensor) – batch.

• surface_tensor (torch tensor) – surface batch.

Returns:

transformed torch tensors.

Return type:

torch.Tensor

inverse_transform(tensor, surface_tensor)

Inverse transform.

Inverse transform input tensor/s.

Parameters:

• tensor (torch tensor) – batch.

• surface_tensor (torch tensor) – surface batch.

Returns:

inverse transformed torch tensors.

Return type:

torch.Tensor

class _transforms.ToTensor(conf)

Convert variables from xr.Datasets to Pytorch Tensors.

conf

hist_len

for_len

variables

surface_variables

allvars

static_variables

__call__(sample: credit.data.Sample) → credit.data.Sample

Convert to reshaped tensor.

Reshape and convert to torch tensor.

Parameters:

sample (interator) – batch.

Returns:

reshaped torch tensor.

Return type:

torch.tensor