HKUDS/Vibe-Trading · error · ValueError

delay requires n >= 1 (lookahead ban)

Error message

delay requires n >= 1 (lookahead ban)

What it means

alpha_048.py raises this ValueError from _delay when n < 1, upholding the no-lookahead invariant before calling df.shift. The file also neutralizes by industry (_ind_neutralize), but the failure is purely in the lag argument.

Source

Thrown at agent/src/factors/zoo/alpha101/alpha_048.py:62

    'extras_required': [],
    'requires_sector': True,
    'universe': ['equity_us', 'equity_in', 'equity_kr'],
    'frequency': ['1D'],
    'decay_horizon': 5,
    'min_warmup_bars': 251,
    'notes': "Industry neutralization implemented via per-row sector group demean (panel['sector'] required). When sector tag is absent the registry rejects via SkipAlpha; the compute() also has a degraded global demean fallback. This is a partial approximation of the paper's IndClass.industry/subindustry/sector neutralization.",
}


def _rolling_sum(df: pd.DataFrame, n: int) -> pd.DataFrame:
    """Rolling window sum; warmup -> NaN."""
    return df.rolling(window=n, min_periods=n).sum()


def _delay(df: pd.DataFrame, n: int) -> pd.DataFrame:
    """Backward shift by n (lookahead-safe; n>=1 required)."""
    if n < 1:
        raise ValueError("delay requires n >= 1 (lookahead ban)")
    return df.shift(n)


def _ind_neutralize(x: pd.DataFrame, panel: dict) -> pd.DataFrame:
    """Industry/sector neutralize: subtract the row-wise sector group mean.

    If panel has a 'sector' DataFrame (same shape as close), subtract the
    per-sector cross-sectional mean per row. If absent, degrade to global
    cross-sectional demean (subtract row mean). This is a degraded fallback
    relative to the paper's industry/subindustry neutralization; see notes.
    """
    sector_df = panel.get("sector")
    if sector_df is None:
        row_mean = x.mean(axis=1, skipna=True)
        return x.sub(row_mean, axis=0)
    # Per-row group demean. Iterate rows; numpy-fast enough for small panels.
    arr = x.to_numpy(dtype=np.float64, na_value=np.nan).copy()
    sec_arr = sector_df.to_numpy()

View on GitHub (pinned to 80ffdda44c)

Solutions

1. Inspect every _delay call site in alpha_048.py; fix lags < 1
2. Pass the frame directly when 'today' is meant
3. Reject lag parameters < 1 at the config/boundary layer
4. Add a unit test for compute() with a toy panel

Example fix

// before
raw = _delay(close, 0) / _delay(close, 2)
// after
raw = close / _delay(close, 2)

if lag < 1:
    raise ValueError(f"delay lag must be >= 1, got {lag}")
neutral = _ind_neutralize(_delay(close, lag), panel)

def is_valid_delay(n: int) -> bool:
    return isinstance(n, int) and not isinstance(n, bool) and n >= 1

When it happens

Trigger: compute() for alpha #48 passes a non-positive lag to _delay while assembling the inputs to _ind_neutralize — e.g. `_delay(x, n)` with n == 0 from a mistyped constant.

Common situations: Porting Alpha#48 (industry-neutral formulation) and miswriting one lag; refactors that rewrite lag arithmetic; config-driven lags permitting 0.

Related errors

• delay requires n >= 1 (lookahead ban)
• delay requires n >= 1 (lookahead ban)
• delay requires n >= 1 (lookahead ban)
• delay requires n >= 1 (lookahead ban)
• delay requires n >= 1 (lookahead ban)

AI-assisted analysis of HKUDS/Vibe-Trading@80ffdda44c (2026-08-28). Data as JSON: /api/errors/bf1e7d0917f4da02. Report an issue: GitHub.