Seed the two unseeded RNG uniformity tests so they cannot fail by chance

[ciaranra]

Summary

test_randint_uniformity and test_choice_uniformity in python/pecos-rslib/tests/test_numpy_random_comparison.py run a chi-square test at p > 0.01 on unseeded draws. By construction that fails about one CI run in a hundred; it took down the pr-core-python (rest) shard on #711 (p=0.0093) for a change that does not touch the RNG.

Every other statistical test in the file already calls pc.random.seed(...) "for deterministic test behavior". This applies the same convention to the two that were missed, with seeds not used elsewhere in the file.

Verification

uv run --frozen pytest python/pecos-rslib/tests/test_numpy_random_comparison.py -k uniformity

4 passed. pre-commit run --files on the changed file: all hooks pass.