HKUDS/Vibe-Trading · error · RegistryError

{alpha_id}: output shape {result.shape} != close shape {ref.

Error message

{alpha_id}: output shape {result.shape} != close shape {ref.shape}

What it means

_validate_output compares the result's shape to the panel's 'close' DataFrame; every factor must produce a matrix with identical (rows, cols) so outputs stack across alphas. Mismatched shapes are rejected.

Source

Thrown at agent/src/factors/registry.py:388

            spec.loader.exec_module(module)
        except Exception:
            sys.modules.pop(alpha.module_path, None)
            raise
        return module

    @staticmethod
    def _validate_output(
        alpha_id: str,
        result: Any,
        panel: dict[str, pd.DataFrame],
    ) -> pd.DataFrame:
        if not isinstance(result, pd.DataFrame):
            raise RegistryError(
                f"{alpha_id}: compute() returned {type(result).__name__}, expected DataFrame"
            )
        ref = panel.get("close")
        if ref is not None and result.shape != ref.shape:
            raise RegistryError(
                f"{alpha_id}: output shape {result.shape} != close shape {ref.shape}"
            )
        arr = result.to_numpy(dtype=np.float64, na_value=np.nan)
        if np.isinf(arr).any():
            raise RegistryError(f"{alpha_id}: output contains +/- inf")
        nan_ratio = float(np.isnan(arr).mean()) if arr.size > 0 else 1.0
        if nan_ratio > 0.95:
            raise RegistryError(f"{alpha_id}: output >95% NaN (nan_ratio={nan_ratio:.3f})")
        return result

    def export_manifest(self) -> dict[str, Any]:
        """Return a JSON-serialisable snapshot for wiki rendering."""
        from datetime import datetime, timezone

        zoos: dict[str, list[dict[str, Any]]] = {}
        for a in self._alphas.values():
            zoos.setdefault(a.zoo, []).append(
                {

View on GitHub (pinned to 80ffdda44c)

Solutions

1. Reindex the result to close's index and columns before returning: `result = result.reindex(index=close.index, columns=close.columns)`
2. Avoid dropna on axes; use fillna/leave NaNs (up to 95% allowed)
3. Compute on full-dimension intermediates

Example fix

# before
return df.dropna(axis=1, how='all')
# after
return df.reindex(index=close.index, columns=close.columns)

ref = panel['close']
assert result.shape == ref.shape, (result.shape, ref.shape)

try:
    out = registry.compute(aid, panel)
except RegistryError as e:
    if 'output shape' in str(e): skip(aid)
    else: raise

When it happens

Trigger: compute returns a subset of columns, a resampled/reindexed frame, or drops NaN rows — anything changing shape relative to panel['close'].

Common situations: Factor drops symbols with missing data (dropna), computes only on filtered tickers, or reindexes dates; panels with duplicated symbols causing shape drift.

Related errors

• {alpha_id}: compute() returned {type(result).__name__}, expe
• unknown universe {v!r}; expected one of {sorted(_BENCH_UNIVE
• ts_max window must be >= 1, got {n}
• ts_min window must be >= 1, got {n}
• {alpha_id}: module has no compute() function

AI-assisted analysis of HKUDS/Vibe-Trading@80ffdda44c (2026-08-28). Data as JSON: /api/errors/a9c6c030cb5ac580. Report an issue: GitHub.