fix(normalisation): retry on ValueError when sampling null ensemble

[cbueth]

The @normalise_against_random decorator generates directed $G(n,m)$ random graphs and feeds them to the metric function to build the null distribution. For small graphs, a random graph can coincidentally end up with a perfectly symmetric adjacency matrix (all edges reciprocated). The reciprocity metric rejects such matrices with ValueError, causing a flaky test failure.

Fix: catch ValueError from the metric function during the sampling loop and retry with a fresh random graph. A for...else guard raises RuntimeError if valid samples cannot be collected within n_rand_val x 10 attempts, ensuring the null distribution always has exactly the requested number of samples.

[codspeed-hq]

Merging this PR will improve performance by 11.74%

⚡ 1 improved benchmark
✅ 27 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	test_detrend_method[z_score-p10]	2.1 ms	1.8 ms	+11.74%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing fix/flaky-reciprocity-test (3780481) with main (875adcd)}

[codecov]

Codecov Report

❌ Patch coverage is 55.55556% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 99.52%. Comparing base (875adcd) to head (3780481).
⚠️ Report is 5 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff             @@
##              main      #52      +/-   ##
===========================================
- Coverage   100.00%   99.52%   -0.48%     
===========================================
  Files           29       29              
  Lines          842      849       +7     
  Branches       151      152       +1     
===========================================
+ Hits           842      845       +3     
- Misses           0        3       +3     
- Partials         0        1       +1     

🚀 New features to boost your workflow:

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