BlockArray to Sparse quite slow about blockarrays.jl HOT 7 OPEN

sparse(::BlockArray) is just calling the default, which spends a lot of time in findall(x -> x != 0, M). It's possible a faster findall would accomplish the same feat, otherwise, can you file a PR with your createSP implemented via overloading SparseMatrixCSC{Tv,Ti}(M::AbstractBlockArray) where {Tv,Ti} which sparse eventually calls?

There's a question whether sparse(::AbstractBlockArray) should return a block array itself with sparse storage, so as to not lose the block information?

from blockarrays.jl.

We should limit the type to BlockArray{Tv,2,SparseMatrixCSC{Tv,Ti}}, don't you agree?

I am not sure I am so skilled as to do a pull request ;)

I wanted to check if I missed a point. Also, my function is geared toward blocks with same size.

from blockarrays.jl.

We should limit the type to BlockArray{Tv,2,SparseMatrixCSC{Tv,Ti}}, don't you agree?

No, as other types have nice sparse representations. Calling sparse(J[Block(i,j)]) should tackle the general case.

But it may be better to create the i, j, z directly, say by modifying sparse(::BandedBlockBandedMatrix): https://github.com/JuliaMatrices/BlockBandedMatrices.jl/blob/e19c22fc9c9e5399976c24aa06a6871bf0133841/src/BandedBlockBandedMatrix.jl#L254

I am not sure I am so skilled as to do a pull request ;)

In open-source software, people maintaining packages have limited time and so are unlikely to add new features that they don't directly benefit from. So if you really want this you'll probably have to do the PR yourself. Given the current state of Julia packages I'd consider being able to make a PR an essential skill to have.

from blockarrays.jl.

Which file should I put the function in?

from blockarrays.jl.

If it only works with BlockArray, then probably src/blockarray.jl, otherwise src.abstractblockarray.jl

from blockarrays.jl.

Did a PR eventually come out of this?

from blockarrays.jl.

I don't believe so, if you are keen to do one I'm sure it will be appreciated.

from blockarrays.jl.