nip.utils.nested_array_dict.CompositeSpec

class nip.utils.nested_array_dict.CompositeSpec(shape: tuple[int], dim_names: tuple[str] | str | None = None, **specs: dict[str, NumpySpec])

Specification for a nested dictionary of numpy arrays.

Methods Summary

__eq__(other)	Return self==value.
__getitem__(key)
__init__(shape[, dim_names])
__post_init__()
__repr__()	Return repr(self).
__setitem__(key, value)
keys([recurse])	Iterate over the keys of CompositeSpec, optionally recursing to sub-specs.
zero()	Return a dictionary of zero arrays with the specified shape.

Attributes

dim_names
shape

Methods

__eq__(other)

Return self==value.

__getitem__(key: str) → NumpySpec

__init__(shape: tuple[int], dim_names: tuple[str] | str | None = None, **specs: dict[str, NumpySpec])

__post_init__()

__repr__()

Return repr(self).

__setitem__(key: str, value: NumpySpec)

keys(recurse=False)

Iterate over the keys of CompositeSpec, optionally recursing to sub-specs.

Parameters:

recurse (bool, default=False) – Whether to recurse to sub-specs. In this case, the keys are tuples of strings representing the path to the data in the nested dictionary.

Yields:

key (str | tuple[str, …]) – The key of the spec. If recurse is True, the key is a tuple of strings representing the path to the data in the nested dictionary. Otherwise, the key is a string.

zero() → NestedArrayDict

Return a dictionary of zero arrays with the specified shape.

Returns:

zero_dict (NestedArrayDict) – The dictionary of zero arrays with the specified shape.