[Perf] Another round of performance improvements

[lkdvos]

This is another round of improvements for generic performances.

In particular, the renormalize_west_edge function now got the same treatment with optimizing intermediate permutations, and I altered some implementations with dtmap as these were completely type-unstable.
At least now, from what I can test, ctmrg_iteration is actually type-stable, and while I don't think this would lead to massive speedups, I do believe this might help with compilation times.

[codecov]

Codecov Report

Attention: Patch coverage is 88.52459% with 7 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/algorithms/contractions/ctmrg_contractions.jl	20.00%	4 Missing ⚠️
src/algorithms/ctmrg/sparse_environments.jl	0.00%	2 Missing ⚠️
src/utility/svd.jl	75.00%	1 Missing ⚠️

Files with missing lines	Coverage Δ
src/PEPSKit.jl	100.00% <ø> (ø)
src/algorithms/correlators.jl	100.00% <100.00%> (ø)
src/algorithms/ctmrg/ctmrg.jl	90.47% <100.00%> (ø)
src/algorithms/ctmrg/sequential.jl	98.41% <100.00%> (+0.05%)	⬆️
src/algorithms/ctmrg/simultaneous.jl	100.00% <100.00%> (+1.72%)	⬆️
src/utility/diffable_threads.jl	100.00% <100.00%> (ø)
src/utility/svd.jl	86.95% <75.00%> (-0.27%)	⬇️
src/algorithms/ctmrg/sparse_environments.jl	29.85% <0.00%> (-0.92%)	⬇️
src/algorithms/contractions/ctmrg_contractions.jl	55.62% <20.00%> (-0.48%)	⬇️

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[pbrehmer]

This is really great, thanks! I have high hopes that the improved type-stable dtmap and ctmrg_iteration will actually improve the compilation times (especially for the backwards pass with Zygote) since these have gotten really bad.

I was wondering, the other edge renormalization functions (renormalize_north_edge, etc.) should have the exact same intermediate permutations, right? The only difference should be which of the bra/ket bond indices will be connected.

Let me know if I can help out or when I should review!

[lkdvos]

For me this would be ready to review/merge, I'm actually 🤒 right now so I won't be able to continue this anytime soon.
Indeed the other contractions should get the same treatment, I focused on the sequential ctmrg to start with though. It's a bit manual in any case, I can't say I'm super confident I always have the best permutations so I'd say it's better to check explicitly the performances, and that takes quite a bit of effort.

[pbrehmer]

Looking good! Other than the import thing, this would be good to go for me.

Get well soon!