Update tests

[borisdevos]

TKS 0.3.6 improved on the test suite. Sadly we can't just blindly import that because of the use of sparse arrays here, so multiple tests are actually not brought over.

I'm also not sure of how much is being doubly tested. I wrote the tests assuming every fusion category is not in the braided list and vice versa.

Edit: This requires a new version release of TKS (QuantumKitHub/TensorKitSectors.jl#70)

[github-actions]

Your PR no longer requires formatting changes. Thank you for your contribution!

[lkdvos]

Would it be possible to refactor the test suite so we can actually reuse it? It's really not the goal to have to copy that code, the point of the suite was to be able to reuse it

[lkdvos]

I think you are missing a slight formatting issue in the categories.jl file (newline at the end), and I left some minor comments. Otherwise good to go for me as soon as the TensorKitSectors changes are released

[codecov]

Codecov Report

❌ Patch coverage is 90.24390% with 4 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/artifacts.jl	90.00%	4 Missing ⚠️

Files with missing lines	Coverage Δ
src/objects.jl	86.20% <100.00%> (+3.44%)	⬆️
src/artifacts.jl	91.70% <90.00%> (+2.96%)	⬆️

... and 3 files with indirect coverage changes

🚀 New features to boost your workflow:

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

[borisdevos]

I'm really confused with the test fail, I thought the point of QuantumKitHub/TensorKitSectors.jl#77 was to capture this? Either way, it seems the error for Object{ZVecD4} is in the Artin braid tests, the two paths seem to differ by a global minus sign.

[lkdvos]

Well, it wasn't running the testsuite yet since you are adding that in this PR, the downstream tests are only attempted on the latest release

[borisdevos]

I seem to have already noticed in QuantumKitHub/TensorKitSectors.jl#59 that certain sectors weren't passing the Artin braid test, and kind of forgot about that 🫠 I've locally tested which braided categories fail the test, of which there are 9 which can be grouped into 4 with different F and R solutions. I'm trying to guess which categories these are, and comparing with the anyonwiki to perhaps make sense of these.

Edit: Let me mention the 4 grouped up categories and what I think they represent here. I'll drop the Object{} for brevity.

• PMFC{4, 1, 0, 1, 0, 2} and PMFC{4, 1, 0, 1, 1, 2}: I think this is $Z_2 \times Z_2$ with the non-trivial 2-cocycle (I think).
• PMFC{6, 1, 0, 1, D1, D2} for $(D1, D2) \in (0,0), (1,1), (2,0)$: Based on fusion rules and dimensions, I think these are $Z_2 \times Ising$. Not sure which $Z_2$, but the Artin braid test with either Z2Irrep or both Z2Elements of TensorKitSectors passes.
• PMFC{6, 1, 0, 8, 0, D} for $D = 0, 2, 3$: Similarly, I think this is $Rep(D_9)$. This sector type exists in TensorKitSectors, and doesn't raise issues.

So yeah, not sure really where to go with this at this point. Maybe just wait till we read in the new data?