Closed wbhart closed 6 years ago
Same question applies to similarity!, which applies a similarity transform. The code seems to only work for square matrices, and the docs seem to suggest we assume that. I'm wondering if there is a reason we don't test for squareness there, or whether I should just add that in.
I think returning false
in ishessenberg
sounds like a good option.
Regarding similarity!
, I don't know what it is doing, but usually we don't check arguments for unsafe functions (exception is setindex!
).
Ok, thanks.
Currently, we can only put a matrix in Hessenberg form if it is square. But we allow non-square matrices in ishessenberg. I'm just wondering if it was intentional, and whether we should return false or an exception if it is not square.