SciML / LinearSolve.jl

LinearSolve.jl: High-Performance Unified Interface for Linear Solvers in Julia. Easily switch between factorization and Krylov methods, add preconditioners, and all in one interface.
https://docs.sciml.ai/LinearSolve/stable/
Other
244 stars 52 forks source link

Promote abstol and reltol type to match eltype(A) #397

Closed ma-sadeghi closed 11 months ago

ma-sadeghi commented 11 months ago

Fixes #396

codecov[bot] commented 11 months ago

Codecov Report

Merging #397 (4c77e2c) into main (f4f6940) will decrease coverage by 0.93%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #397      +/-   ##
==========================================
- Coverage   68.97%   68.04%   -0.93%     
==========================================
  Files          26       26              
  Lines        1905     1906       +1     
==========================================
- Hits         1314     1297      -17     
- Misses        591      609      +18     
Files Coverage Δ
src/common.jl 89.28% <100.00%> (-1.46%) :arrow_down:

... and 6 files with indirect coverage changes

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more

ma-sadeghi commented 11 months ago

@ChrisRackauckas Should we change here as well? (to prob.b)

https://github.com/SciML/LinearSolve.jl/blob/f4f69409250ee359bb26e1409d56d53ca3d53e29/src/common.jl#L118-L119

ChrisRackauckas commented 11 months ago

yes that would be good to change.

amontoison commented 10 months ago

It will not work if the linear systems have complex numbers. You need real(eltype(b)) here.