Vector-valued outputs as fzd objectives (part 2/2)

[yannrichet-asnr]

Summary

Part 2 of vector-output support — part 1 is #75 (fzr/fzo). This PR is stacked on top of it (base branch: feat/vector-outputs-fzr-fzo) since the docs cross-reference each other; it'll auto-retarget to main once #75 merges.

fzd's algorithms (sampling, optimization, ...) always need a single scalar objective per case, but the underlying model output can now be a vector (e.g. a time series, per #75). This PR makes output_expression — the place where that reduction happens — actually usable for it.

Changes

• fz/algorithms.py: evaluate_output_expression() gains sum(), len(), sorted(), mean(), median(), stdev(), variance() in its safe eval namespace, alongside the pre-existing math functions and indexing/slicing (which already worked, e.g. series[-1], max(series)). Referencing a vector-valued output without reducing it used to fail with a bare float() argument must be a string or a real number, not 'list' TypeError; it now raises a ValueError naming the offending output(s) and suggesting a concrete reduction (e.g. mean(T_series)). fzd's per-case error handling already treats any evaluate_output_expression failure as a failed point (output value None) rather than aborting the run, so this is a message-quality fix, not a behavior change there.
• tests/test_fzd_vector_outputs.py (new, 14 tests): unit coverage of the new reduction helpers (including a realistic "RMS over a vector output" expression) and the improved error message (names the vector-valued key(s), leaves scalar outputs out of the blame list, keeps the old undefined-name error unchanged); plus two end-to-end fzd() tests with a real vector-output model (python://json_file time series) — one with a correctly reduced output_expression, verified against the manually computed mean for every sampled point, and one with a deliberately unreduced expression, verifying the run completes with every point reported as a failed evaluation rather than crashing.
• Docs: doc/core-functions.md and examples/fzd_example.md get a new "Vector-valued outputs as objectives" subsection (with a working RMS example: sqrt(sum(v**2 for v in T_series) / len(T_series))); README.md and skills/fz/reference.md get matching notes; doc/INDEX.md links added; NEWS.md entry.

Scope note

fzd's own objective is still a single scalar per case — this is about reducing a vector-valued model output to that scalar, not multi-objective/vector-objective optimization (out of scope here).

Testing

• New suite: pytest tests/test_fzd_vector_outputs.py — 14/14 passed.
• No regressions: full test_fzd.py (35/35, including the slower subprocess-based TestFzdIntegration class), test_algorithm_options.py, test_algorithm_resolution.py, test_algorithm_installation.py, test_algorithm_plugins.py, test_r_algorithms.py, test_examples_scripts.py, test_readme_snippets.py, test_skill_static.py, plus a broad sweep of the rest of the suite (same exclusions CI itself applies).

🤖 Generated with Claude in Cowork mode.

[yannrichet-asnr]

Pushed a follow-up commit checking (and confirming) that `output_expression` also supports objectives built from more than one vector output, not just a reduction of a single one:

• Concatenate then reduce (already worked, no code change — just verified and documented): plain `+` on two lists concatenates them, e.g. `mean(a + b)` pools both series before averaging.
• Combine two independent reductions: `mean(a) - mean(b)`, etc. — already worked.
• Element-wise combination via a new `zip()` helper: e.g. an RMSE/residual between a simulated and a reference series — `sqrt(sum((x - y) ** 2 for x, y in zip(sim, ref)) / len(sim))`.

While adding the `zip()` case I found a real, separate bug it exposed: `evaluate_output_expression()` called `eval(expr, {"builtins": {}}, safe_dict)` — a split globals/locals dict. Any generator-expression or comprehension body resolves names through the globals dict only (never the separately-passed locals dict), so calling a helper function like `abs()` inside a genexp body (e.g. `max(abs(x - y) for x, y in zip(a, b))`) raised a spurious `name 'abs' is not defined` — even though the exact same call works fine outside a genexp. Fixed by evaluating with a single combined globals dict.

Added TestEvaluateOutputExpressionCombiningDifferentOutputs (6 tests) plus an end-to-end fzd() RMSE test to tests/test_fzd_vector_outputs.py (20/20 passing, was 14/14), and updated all the doc surfaces accordingly.