change to SciMLOperators

[vpuri3]

https://github.com/SciML/SciMLOperators.jl/issues/55

[codecov[bot]]

Codecov Report

Merging #171 (b09d09b) into main (bb1f9f4) will decrease coverage by 9.25%. The diff coverage is 57.00%.

@@            Coverage Diff             @@
##             main     #171      +/-   ##
==========================================
- Coverage   66.87%   57.62%   -9.26%     
==========================================
  Files           9        8       -1     
  Lines         631      623       -8     
==========================================
- Hits          422      359      -63     
- Misses        209      264      +55     

Impacted Files	Coverage Δ
src/simplelu.jl	0.00% <0.00%> (ø)
src/LinearSolve.jl	12.50% <25.00%> (-62.50%)	:arrow_down:
src/default.jl	33.33% <38.59%> (-36.06%)	:arrow_down:
src/factorization.jl	71.14% <58.33%> (-10.45%)	:arrow_down:
src/common.jl	75.43% <87.50%> (-8.57%)	:arrow_down:
src/iterative_wrappers.jl	70.81% <93.33%> (+2.27%)	:arrow_up:
src/solve_function.jl	100.00% <100.00%> (ø)
... and 3 more

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

[vpuri3]

vector diagonal preconditioner should be updated as follows:

            s = rand(n)
            #Pl, Pr = DiagonalOperator(s), LinearSolve.InvPreconditioner(Diagonal(s))
            Pl = DiagonalOperator(s)
            Pr = inv(Pl)

            Pl = cache_operator(Pl, x)
            Pr = cache_operator(Pr, x)```

[vpuri3]

this is good

[ChrisRackauckas]

Tests fail (needs formatting, and downstream tests)

[vpuri3]

might have to keep ComposePreconditioner since OrdinaryDiffEq uses that in dolinsolve function

[ChrisRackauckas]

It's fine to remove if the deprecation is correct.

[vpuri3]

reminder to make https://github.com/JuliaSmoothOptimizers/Krylov.jl/issues/612#issuecomment-1227391841 part of this

[vpuri3]

yes lotsa rebase issues ://. I have half a mind of abandoning this branch and starting anew