DataProviderLinked

class glotaran.optimization.data_provider.DataProviderLinked(scheme: Scheme, dataset_group: DatasetGroup)

Bases: DataProvider

A class to provide aligned data for optimization.

Initialize a linked data provider for a scheme and a dataset_group.

Parameters:

• scheme (Scheme) – The optimization scheme.

• dataset_group (DatasetGroup) – The dataset group.

Attributes Summary

aligned_global_axis	Get the aligned global axis for the dataset group.
group_definitions	Get the group definitions for the dataset group.

Methods Summary

add_model_weight	Add model weight to data.
align_data	Align the data in a dataset group.
align_dataset_indices	Align the global indices in a dataset group.
align_groups	Align the groups in a dataset group.
align_index	Align an index on a target axis.
align_weights	Align the weights in a dataset group.
create_aligned_global_axes	Create aligned global axes for the dataset group.
get_aligned_data	Get the aligned data for an index.
get_aligned_dataset_indices	Get the aligned dataset indices for an index.
get_aligned_group_label	Get the group label for an index.
get_aligned_weight	Get the aligned weight for an index.
get_axis_slice_from_interval	Get a slice of indices from a min max tuple and for an axis.
get_data	Get data for a dataset.
get_flattened_data	Get flattened data for a dataset.
get_flattened_weight	Get flattened weight for a dataset.
get_from_dataset	Get a copy of data from a dataset with dimensions (model, global).
get_global_axis	Get the global axis for a dataset.
get_global_dimension	Get the global dimension for a dataset.
get_model_axis	Get the model axis for a dataset.
get_model_dimension	Get the model dimension for a dataset.
get_weight	Get weight for a dataset.
infer_global_dimension	Infer the name of the global dimension from tuple of dimensions.

Methods Documentation

add_model_weight(model: Model, dataset_label: str, model_dimension: str, global_dimension: str)

Add model weight to data.

Parameters:

• model (Model) – The model.

• dataset_label (str) – The label of the data.

• model_dimension (str) – The model dimension.

• global_dimension (str) – The global dimension.

align_data(aligned_global_axes: dict[str, ArrayLike]) → tuple[ArrayLike, list[ArrayLike]]

Align the data in a dataset group.

Parameters:

aligned_global_axes (dict[str, ArrayLike]) – The aligned global axes.

Returns:

The aligned global axis and data.

Return type:

tuple[ArrayLike, list[ArrayLike]]

align_dataset_indices(aligned_global_axes: dict[str, ArrayLike]) → list[ArrayLike]

Align the global indices in a dataset group.

Parameters:

aligned_global_axes (dict[str, ArrayLike]) – The aligned global axes.

Returns:

• list[ArrayLike]

• The aligned dataset indices.

static align_groups(aligned_global_axes: dict[str, ArrayLike]) → tuple[ArrayLike, dict[str, list[str]]]

Align the groups in a dataset group.

Parameters:

aligned_global_axes (dict[str, ArrayLike]) – The aligned global axes.

Returns:

The aligned grouplabels and group definitions.

Return type:

tuple[ArrayLike, dict[str, list[str]]]

static align_index(index: int, target_axis: ArrayLike, tolerance: float, method: Literal['nearest', 'backward', 'forward']) → int

Align an index on a target axis.

Parameters:

• index (int) – The index to align.

• target_axis (ArrayLike) – The axis to align the index on.

• tolerance (float) – The alignment tolerance.

• method (Literal["nearest", "backward", "forward"]) – The alignment method.

Returns:

The aligned index.

Return type:

int

align_weights(aligned_global_axes: dict[str, ArrayLike]) → list[ArrayLike | None]

Align the weights in a dataset group.

Parameters:

aligned_global_axes (dict[str, ArrayLike]) – The aligned global axes.

Returns:

The aligned weights.

Return type:

list[ArrayLike | None]

property aligned_global_axis: ArrayLike

Get the aligned global axis for the dataset group.

Returns:

The aligned global axis.

Return type:

ArrayLike

create_aligned_global_axes(scheme: Scheme) → dict[str, ArrayLike]

Create aligned global axes for the dataset group.

Parameters:

scheme (Scheme) – The optimization scheme.

Returns:

The aligned global axes.

Return type:

dict[str, ArrayLike]

Raises:

AlignDatasetError – Raised when dataset alignment is ambiguous.

get_aligned_data(index: int) → ArrayLike

Get the aligned data for an index.

Parameters:

index (int) – The index on the aligned global axis.

Returns:

The aligned data.

Return type:

ArrayLike

get_aligned_dataset_indices(index: int) → ArrayLike

Get the aligned dataset indices for an index.

Parameters:

index (int) – The index on the aligned global axis.

Returns:

The aligned dataset indices.

Return type:

ArrayLike

get_aligned_group_label(index: int) → str

Get the group label for an index.

Parameters:

index (int) – The index on the aligned global axis.

Returns:

The aligned group label.

Return type:

str

get_aligned_weight(index: int) → ArrayLike | None

Get the aligned weight for an index.

Parameters:

index (int) – The index on the aligned global axis.

Returns:

The aligned weight.

Return type:

ArrayLike | None

static get_axis_slice_from_interval(interval: tuple[float, float], axis: ArrayLike) → slice

Get a slice of indices from a min max tuple and for an axis.

Parameters:

• interval (tuple[float, float]) – The min max tuple.

• axis (ArrayLike) – The axis to slice.

Returns:

The slice of indices.

Return type:

slice

get_data(dataset_label: str) → ArrayLike

Get data for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The data.

Return type:

ArrayLike

get_flattened_data(dataset_label: str) → ArrayLike

Get flattened data for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The flattened data.

Return type:

ArrayLike

get_flattened_weight(dataset_label: str) → ArrayLike | None

Get flattened weight for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The flattened weight.

Return type:

ArrayLike | None

static get_from_dataset(dataset: xr.Dataset, name: str, model_dimension: str, global_dimension: str) → ArrayLike | None

Get a copy of data from a dataset with dimensions (model, global).

Parameters:

• dataset (xr.Dataset) – The dataset to retrieve from.

• name (str) – The name of the data to retrieve.

• model_dimension (str) – The model dimension.

• global_dimension (str) – The global dimension.

Returns:

The copy of the data. None if name is not present in dataset.

Return type:

ArrayLike | None

get_global_axis(dataset_label: str) → ArrayLike

Get the global axis for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The global axis.

Return type:

ArrayLike

get_global_dimension(dataset_label: str) → str

Get the global dimension for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The global dimension.

Return type:

str

get_model_axis(dataset_label: str) → ArrayLike

Get the model axis for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The model axis.

Return type:

ArrayLike

get_model_dimension(dataset_label: str) → str

Get the model dimension for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The model dimension.

Return type:

str

get_weight(dataset_label: str) → ArrayLike | None

Get weight for a dataset.

Parameters:

dataset_label (str) – The label of the data.

Returns:

The weight.

Return type:

ArrayLike | None

property group_definitions: dict[str, list[str]]

Get the group definitions for the dataset group.

Returns:

The group definitions.

Return type:

dict[str, list[str]]

static infer_global_dimension(model_dimension: str, dimensions: tuple[str]) → str

Infer the name of the global dimension from tuple of dimensions.

Parameters:

• model_dimension (str) – The model dimension.

• dimensions (tuple[str]) – The dimensions tuple to infer from.

Returns:

The inferred name of the global dimension.

Return type:

str