Andreas Noack
Andreas Noack
Bump. It would be good to get this one in and released such that we can continue the testing of Julia 1.9
Can the [downstream failure](https://github.com/SciML/NonlinearSolve.jl/actions/runs/3539148420/jobs/5940729733#step:6:808) safely be ignored?
See https://github.com/Homebrew/homebrew-core/issues/27751
This is a limitation in SuiteSparse. See https://github.com/PetterS/SuiteSparse/blob/27e5a8516464a6ac40bd3fa0e5b46e51b11f4765/CHOLMOD/Include/cholmod_internal.h#L175-L176. We'd have to either wait for CHOLMOD to be extended or to implemenet our own sparse Cholesky solver which would be a...
Are you proposing that we change it to `AbstractVecOrMat`?
Seems reasonable to me
I'm fine with changing the names but it would better if we could generally use getters as in https://github.com/JuliaLang/julia/blob/292a3b4c73b91bcb3dbe2d9d7c56f8d747ce5207/base/sparse/sparsematrix.jl#L48-L53 since it would allow many algorithms to work efficiently for views...
Are you sure this is about https://github.com/JuliaLang/julia/blob/master/stdlib/SparseArrays/src/sparsematrix.jl#L2701? Isnøt the issue just that `SparseMatrixCSC` doesn't really work for element types without a zero defined.
Why do think that? Both the current versions test for mismatch of the first dimension of the arrays.
I see. In `lmul!`, `Q` is square and that is the same for dense and sparse. ```julia julia> A = randn(4,2); julia> Asp = sparse(A); julia> F = qr(A); julia>...