Remove redundant uplo argument in chol family. #17909
Conversation
kshyatt
added
linear algebra
|
Would be nice to update the docstring about this as well. |
|
new deprecations should not be backported. can you separate the bugfix from the deprecation into separate commits? |
|
They should have been removed in #16799 but slipped through. Since they are wrong, it's unlikely that anybody relies on them so the right thing is to remove them (and they have only existed on master since June). My thinking was that the deprecation could be helpful if somebody accidentally called the methods but I can also just remove the methods without deprecation. |
| @@ -187,7 +197,7 @@ end | |||
|
|
|||
| ### for StridedMatrices, check that matrix is symmetric/Hermitian | |||
| """ | |||
| cholfact!(A, uplo::Symbol, Val{false}) -> Cholesky | |||
| cholfact!(A, [uplo::Symbol,] Val{false}) -> Cholesky | |||
There was a problem hiding this comment.
change this in the rst so the signatures match
|
Updated the doc string. Please comment. |
|
I guess I'd prefer still doing a deprecation rather than a hard removal of the method, but if it's only been on master for a few months and is wrong, then we can put the deprecation in the 0.5 section and backport. |
and Symmetric, it is redundant to also have an uplo argument and occasionally, it also gave the wrong result. This can therefore be considered a bugfix.
…uliaLang#17909) and Symmetric, it is redundant to also have an uplo argument and occasionally, it also gave the wrong result. This can therefore be considered a bugfix.
…mitian (JuliaLang#17909)" This reverts commit d992f3d.
When using
HermitianandSymmetric, it is redundant to also have anuploargument andoccasionally, it also gave the wrong result. This can therefore be
considered a bugfix.