v2.0.0

LinearSolve v2.0.0

Diff since v1.42.0

Closed issues:

• Factorization runs twice for sparse matrices (#179)
• Cache issues with sequential calls to solve (#199)
• QRFactorization causes change in the cacheval after solve (#299)
• solving the same problem twice is broken (#302)

Merged pull requests:

• Migrate from SnoopPrecompile to PrecompileTools (#296) (@timholy)
• Mutable cache form (#303) (@ChrisRackauckas)
• Make all extra solvers into extension packages (#304) (@ChrisRackauckas)
• Deprecate the extra libs (#305) (@ChrisRackauckas)
• Always pivot in RFLUFactorization (#306) (@ChrisRackauckas)
• Rework default algorithm to be fully type stable (#307) (@ChrisRackauckas)
• Update documentation regarding alias behavior (#309) (@fredrikekre)
• Turn IterativeSolvers and KrylovKit into extensions (#310) (@ChrisRackauckas)
• Transform all solver docs to use docstrings (#312) (@ChrisRackauckas)
• add release notes (#314) (@ChrisRackauckas)
• Re-enable some precompilation (#316) (@ChrisRackauckas)

v1.42.0

LinearSolve v1.42.0

Diff since v1.41.1

Closed issues:

• Better initialization for most factorizations (#67)
• Phantom allocations: need to define effects? (#245)
• precompilation warning (#293)

Merged pull requests:

• Fix some typos (#294) (@goggle)
• Attempt to close #293 (#295) (@thomvet)

v1.41.1

LinearSolve v1.41.1

Diff since v1.41.0

v1.41.0

LinearSolve v1.41.0

Diff since v1.40.0

Closed issues:

• Solving tall, sparse linear systems is insanely slow (#283)
• Sparse QR doesn't work with 'view' (#287)

Merged pull requests:

• Fix non-square sparse views (#288) (@ChrisRackauckas)
• Expand OperatorAssumptions to allow for choosing conditioning (#289) (@ChrisRackauckas)
• Add NormalCholeskyFactorization and WellConditioned defaults (#290) (@ChrisRackauckas)
• clean up sparse ldiv for GPL-free builds (#291) (@ChrisRackauckas)
• CompatHelper: add new compat entry for EnumX at version 1, (keep existing compat) (#292) (@github-actions[bot])

v1.40.0

LinearSolve v1.40.0

Diff since v1.39.0

Merged pull requests:

• Don't use qr! on SparseMatrixCSC (#285) (@ChrisRackauckas)
• Fix HYPRE Solution Building (#286) (@ChrisRackauckas)

v1.39.0

LinearSolve v1.39.0

Diff since v1.38.0

Merged pull requests:

• enable dependabot for GitHub actions (#279) (@ranocha)
• Bump actions/checkout from 1 to 3 (#280) (@dependabot[bot])
• Bump codecov/codecov-action from 1 to 3 (#281) (@dependabot[bot])
• Bump actions/cache from 1 to 3 (#282) (@dependabot[bot])
• Fix QR and SVD initializations (#284) (@ChrisRackauckas)

v1.38.0

LinearSolve v1.38.0

Diff since v1.37.0

Closed issues:

• Use AbstractSparseMatrixCSC in the API (#190)

Merged pull requests:

• Bump SciMLOperators (#276) (@ChrisRackauckas)

v1.37.0

LinearSolve v1.37.0

Diff since v1.36.0

Merged pull requests:

• Use AbstractSparseMatrixCSC for sparse factorizations (#273) (@j-fu)
• strict links (#274) (@ArnoStrouwen)

v1.36.0

LinearSolve v1.36.0

Diff since v1.35.0

Merged pull requests:

• fix typo when setting Pr to default (#272) (@j-fu)
• Update to ArrayInterface v7 (#275) (@ChrisRackauckas)

v1.35.0

LinearSolve v1.35.0

Diff since v1.34.1

Closed issues:

• LinearSolve fails to precompile on julia 1.8.5 (#266)
• Make issquare extend SciMLOperators.issquare (#268)

Merged pull requests:

• CompatHelper: bump compat for CUDA to 4 for package LinearSolveCUDA, (keep existing compat) (#267) (@github-actions[bot])
• Support SciMLOperators in LinearSolve (#270) (@vpuri3)