differential_geometry.dependence.TensorDependenceMixin.adjust_tensor_signature#

TensorDependenceMixin.adjust_tensor_signature(variance_in: Sequence[int], variance_out: Sequence[int], /, *, as_field: bool = False)[source]#

Adjust the tensor variance (covariant/contravariant) of each index.

Parameters:

  • variance_in (Sequence[int]) – The current variance of each index (0 = covariant, 1 = contravariant).

  • variance_out (Sequence[int]) – The target variance signature to convert to.

  • as_field (bool, optional) – If True, return the symbolic expression. If False, return a new dependence object.

Returns:

Symbolic tensor with adjusted signature or a new dependence object.

Return type:

_GenericFieldType or _DepT

Raises:

ValueError – If the variance vector lengths do not match the tensor rank.