Open miakramer opened 2 years ago
Almost all operations that you can do with Tensor(Map)
objects, such as contractions and linear algebra factorizations, are currently only supported for DenseMatrix
. Actually, in practice, typically only an implementation for Matrix
is available; the parametric type is there to support CuArray
in the (hopefully near) future. I thus don't quite see what you hope to gain by widening this to AbstractMatrix
?
Is there any particular reason the interface for
Tensor
requiresA <: DenseMatrix
for the storage type? Does this package assume column-major indexing, for example? If not, would a PR removing that requirement and instead requiringA <: AbstractMatrix
be welcome?