Drop support for integer matrices

[amontoison]

• Drop support for integer matrices (Wrap automatically preconditioners in a LinearOperator #181)
• I also removed the variant for A and b with different type. If it's the case, we should not always convert b to a Vector. It could be a CuVector for example.
  -> The user should give A and b with the same type.

[codecov]

Codecov Report

Merging #182 into master will increase coverage by 0.02%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #182      +/-   ##
==========================================
+ Coverage   97.15%   97.17%   +0.02%     
==========================================
  Files          28       28              
  Lines        2495     2478      -17     
==========================================
- Hits         2424     2408      -16     
+ Misses         71       70       -1

Impacted Files	Coverage Δ
src/variants.jl	100% <ø> (ø)	⬆️
src/cg_lanczos.jl	100% <ø> (ø)	⬆️
src/diom.jl	93.58% <0%> (+1.28%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b487cbf...3c56bfa. Read the comment docs.

[coveralls]

Coverage increased (+0.02%) to 97.175% when pulling 3c56bfa on amontoison:adios_int_matrices into b487cbf on JuliaSmoothOptimizers:master.

[amontoison]

@dpo Can you merge this pull request? I want to rebase #181 and test it with CuArrays.

[dpo]

Yes. I was waiting yesterday for the Appveyor builds to finish, which was never happening... Thank you for this.