CollectiveTransiton implementation

[ChristophHotter]

No description provided.

[github-actions]

Benchmark Results

Details

Benchmark suite	Current: f11b699	Previous: e7e7648	Ratio
Accumulation/Many-mode H/foldl M=16	10671 ns	10980 ns	0.97
Accumulation/Many-mode H/foldl M=24	23163 ns	24876 ns	0.93
Accumulation/Many-mode H/foldl M=8	2719.5555555555557 ns	2800.777777777778 ns	0.97
Accumulation/Many-mode H/sum M=16	1639.1 ns	1697.1 ns	0.97
Accumulation/Many-mode H/sum M=24	2260.8888888888887 ns	2305.4444444444443 ns	0.98
Accumulation/Many-mode H/sum M=8	706.1102941176471 ns	729.7297297297297 ns	0.97
Accumulation/Same-site/foldl	6342 ns	6275.6 ns	1.01
Accumulation/Same-site/sum	2287.6666666666665 ns	2295.3333333333335 ns	1.00
Commutator/Nested JC/depth=1	1677.2 ns	1705.1 ns	0.98
Commutator/Nested JC/depth=2	18114 ns	17583 ns	1.03
Commutator/Nested JC/depth=3	67427 ns	66503 ns	1.01
Commutator/Nested JC/depth=4	169008 ns	168372 ns	1.00
Commutator/Nested JC/depth=5	372422 ns	372361 ns	1.00
Commutator/Schrieffer-Wolff/[S, V]	195909 ns	198388 ns	0.99
Commutator/Schrieffer-Wolff/[S, [S, H0]]	361582 ns	367321 ns	0.98
Indexing/Diagonal collapse/[H_Dicke, S_j]	41959 ns	44253 ns	0.95
Indexing/Diagonal collapse/[H_JC, σ_j]	60214 ns	63227 ns	0.95
Indexing/Simplify/double-sum spin-spin	16000 ns	16551 ns	0.97
Indexing/Simplify/indexed JC H	1200.3 ns	1169.1 ns	1.03
Indexing/Sum construction/double Σ_ij(J_ij*S_i*S_j)	7980 ns	8766 ns	0.91
Indexing/Sum construction/single Σ_i(σ_i*σ_j)	10219 ns	10669 ns	0.96
Normal Order/Fock (c*c')^n/n=2	608.7865168539325 ns	611.7403314917127 ns	1.00
Normal Order/Fock (c*c')^n/n=3	877.290909090909 ns	892.1754385964912 ns	0.98
Normal Order/Fock (c*c')^n/n=4	1178.2 ns	1183.2 ns	1.00
Normal Order/Fock (c*c')^n/n=5	1562 ns	1565.9 ns	1.00
Normal Order/Ground state/3-level rewrite	826.0298507462686 ns	809.5487804878048 ns	1.02
Normal Order/Multi-mode/2-mode 6-op chain	967.5714285714286 ns	964.421052631579 ns	1.00
Simplify/Jaynes-Cummings/H	1082 ns	1050 ns	1.03
Simplify/Jaynes-Cummings/H²	2421.1111111111113 ns	2412.3333333333335 ns	1.00
Simplify/Two cavities/H	1129.1 ns	1103 ns	1.02
Simplify/Two cavities/H²	4399.714285714285 ns	4495.428571428572 ns	0.98
Simplify/Λ-system/H	1370.6 ns	1352.5 ns	1.01
Simplify/Λ-system/H²	2341.1111111111113 ns	2276.4444444444443 ns	1.03

This comment was automatically generated by workflow using github-action-benchmark.

[codecov]

Codecov Report

❌ Patch coverage is 94.87179% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 94.53%. Comparing base (e7e7648) to head (f11b699).

Files with missing lines	Patch %	Lines
src/operators/nlevel.jl	91.17%	3 Missing ⚠️
src/operators/operators.jl	94.44%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #211      +/-   ##
==========================================
+ Coverage   94.51%   94.53%   +0.01%     
==========================================
  Files          27       27              
  Lines        2937     3001      +64     
==========================================
+ Hits         2776     2837      +61     
- Misses        161      164       +3     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[oameye]

pre-exisiting bug:
commutator(Spin(hs,:A,1), Spin(hs,:B,2)) on one SpinSpace returns im*Bz via the fast path, while simplify(A*B − B*A) returns 0 (the pipeline treats different-named spins as commuting).

[oameye]

_commute_ops_at now computes _mul_cnum(c, residual2_coeff) on every non-commuting swap even for roles whose second slot is always zero; with a symbolic c that's a wasted poly scale. A cheap _iszero_cnum(residual2_coeff) || guard before the multiply would restore the old cost exactly.

[oameye]

We need to bump to 0.9.4