HKUDS/Vibe-Trading · error · TypeError

run_bench_strict requires random_control to be passed explic

Error message

run_bench_strict requires random_control to be passed explicitly (True or False). This rail is borrowed from Soli22de/Bili_Stock's foundation engine after a 9-month audit where every accidental random_control=None call inflated alpha by 3-8 percentage points.

What it means

run_bench_strict takes random_control as a keyword-only parameter with no default; passing None (or omitting it) raises TypeError. This rail exists because accidental random_control=None calls silently inflated alpha by 3-8 percentage points in a past audit, so the API forces an explicit True/False choice.

Source

Thrown at agent/src/factors/bench_runner_strict.py:372

        registry: Optional pre-built registry for tests.

    Returns:
        Dict containing all the keys ``run_bench()`` returns, plus:

        - ``random_control`` (bool)
        - ``n_random_seeds`` (int)
        - ``oos_split`` (str | None)
        - ``alpha_t_threshold`` (float)
        - ``confirmed_alive`` / ``train_only`` / ``reversed_strict`` /
          ``noise`` count keys
        - Each row carries ``alpha_t_full``, ``alpha_t_train`` (when OOS),
          ``alpha_t_test`` (when OOS), ``random_ic_mean``.

    Raises:
        TypeError: If ``random_control`` is omitted (keyword-only, no default).
    """
    if random_control is None:  # pragma: no cover — guarded by signature
        raise TypeError(
            "run_bench_strict requires random_control to be passed explicitly "
            "(True or False). This rail is borrowed from "
            "Soli22de/Bili_Stock's foundation engine after a 9-month audit "
            "where every accidental random_control=None call inflated alpha "
            "by 3-8 percentage points."
        )

    start = time.monotonic()
    thresholds = thresholds or StrictThresholds()
    # Clamp n_random_seeds once and store the actual value used so the
    # wire response doesn't lie about the seed count when callers pass 0
    # (e.g. from a JSON-config import).
    effective_seeds = max(1, int(n_random_seeds))

    # Initialise the full schema up-front so even early-error returns
    # carry zeroed counters and empty lists — downstream consumers can
    # depend on the keys always being present.
    entry: dict[str, Any] = {

View on GitHub (pinned to 80ffdda44c)

Solutions

1. Pass random_control=True or random_control=False explicitly as a keyword
2. In wrappers, require the parameter yourself instead of defaulting to None
3. Read the docstring: the strictness is intentional (prevents silent benchmark inflation)

Example fix

# before
run_bench_strict(registry, **kwargs)  # random_control missing
# after
run_bench_strict(registry, random_control=True, **kwargs)

def has_explicit_random_control(kwargs: dict) -> bool:
    v = kwargs.get('random_control')
    return v is True or v is False

try:
    run_bench_strict(reg, random_control=True, **kw)
except TypeError as e:
    if 'random_control' in str(e): raise ValueError('caller must set random_control') from e
    raise

When it happens

Trigger: Calling run_bench_strict(...) without random_control=True/False, or explicitly passing random_control=None, or forwarding a None default from a wrapper function.

Common situations: Wrapping run_bench_strict in a convenience function whose own default is None and forwarding it; older call sites written before the parameter became mandatory; test helpers that omit kwargs.

Understand the failure class

Background: "missing required argument" and "the following required arguments were not provided": what required-argument errors mean and how to fix them — this error's family across 20 libraries.

Related errors

• index_levels is empty; a public market equivalent needs a be

AI-assisted analysis of HKUDS/Vibe-Trading@80ffdda44c (2026-08-28). Data as JSON: /api/errors/8e372fefbed23c69. Report an issue: GitHub.