bw2calc.single_value_diagonal_matrix

Module Contents

Classes

SingleValueDiagonalMatrix

A scipy sparse matrix handler which takes in bw_processing data packages. Row and column

class bw2calc.single_value_diagonal_matrix.SingleValueDiagonalMatrix(*, packages: Sequence[bw_processing.Datapackage], matrix: str, dimension: int, use_vectors: bool = True, use_arrays: bool = True, use_distributions: bool = False, seed_override: int | None = None, indexer_override: Any = None, custom_filter: Callable | None = None)

Bases: matrix_utils.MappedMatrix

A scipy sparse matrix handler which takes in bw_processing data packages. Row and column ids are mapped to matrix indices, and a matrix is constructed.

Use primarily in the weighting step of life cycle impact assessment.

indexer_override allows for custom indexer behaviour. Indexers should follow a simple API: they must support .__next__(), and have the attribute .index, which returns an integer.

custom_filter allows you to remove some data based on their indices. It is applied to all resource groups. If you need more fine-grained control, process the matrix after construction/iteration. custom_filter should take the indices array as an input, and return a Numpy boolean array with the same length as the indices array.

Args:

• packages: A list of Ddatapackage objects.

• matrix: The string identifying the matrix to be built.

• use_vectors: Flag to use vector data from datapackages

• use_arrays: Flag to use array data from datapackages

• use_distributions: Flag to use stats_arrays distribution data from datapackages

• row_mapper: Optional instance of ArrayMapper. Used when matrices must align.

• col_mapper: Optional instance of ArrayMapper. Used when matrices must align.

• seed_override: Optional integer. Overrides the RNG seed given in the datapackage, if any.

• indexer_override: Parameter for custom indexers. See above.

• diagonal: If True, only use the row indices to build a diagonal matrix.

• custom_filter: Callable for function to filter data based on indices values. See above.

• empty_ok: If False, raise AllArraysEmpty if the matrix would be empty

rebuild_matrix()