hsliuping/TradingAgents-CN · error · ValueError

不支持的指标: {name}

Error message

不支持的指标: {name}

What it means

compute_indicator dispatches on the indicator name against the SUPPORTED set {'ma','ema','macd','rsi','boll','atr','kdj'}. An unrecognized name falls through all branches to this ValueError listing nothing but the name, so the caller can correct the spec. It is the single entry point also used by compute_many, so bad specs in batch lists surface here too.

Source

Thrown at tradingagents/tools/analysis/indicators.py:246

        return out

    if name == "atr":
        _require_cols(df, ["high", "low", "close"])
        n = int(params.get("n", 14))
        out[f"atr{n}"] = atr(df["high"], df["low"], df["close"], n=n)
        return out

    if name == "kdj":
        _require_cols(df, ["high", "low", "close"])
        n = int(params.get("n", 9))
        m1 = int(params.get("m1", 3))
        m2 = int(params.get("m2", 3))
        kdj_df = kdj(df["high"], df["low"], df["close"], n=n, m1=m1, m2=m2)
        for c in kdj_df.columns:
            out[c] = kdj_df[c]
        return out

    raise ValueError(f"不支持的指标: {name}")


def compute_many(df: pd.DataFrame, specs: List[IndicatorSpec]) -> pd.DataFrame:
    if not specs:
        return df.copy()
    # 粗略去重（按 name+sorted(params)）
    def key(s: IndicatorSpec):
        p = s.params or {}
        items = tuple(sorted(p.items()))
        return (s.name.lower(), items)

    unique_specs: List[IndicatorSpec] = []
    seen = set()
    for s in specs:
        k = key(s)
        if k not in seen:
            seen.add(k)
            unique_specs.append(s)

View on GitHub (pinned to 74783e8817)

Solutions

1. Check the SUPPORTED constant at the top of indicators.py and use one of: ma, ema, macd, rsi, boll, atr, kdj.
2. Validate spec names against indicators.SUPPORTED before calling compute_many with config-driven lists.
3. Compute unsupported indicators separately with custom pandas/TA-Lib code and merge the resulting columns.

Example fix

# before
out = compute_indicator(df, "bollinger", n=20)

# after
from tradingagents.tools.analysis.indicators import SUPPORTED
name = "bollinger" if "bollinger" in SUPPORTED else "boll"
out = compute_indicator(df, name, n=20)

from tradingagents.tools.analysis.indicators import SUPPORTED
name = name.strip().lower()
if name not in SUPPORTED:
    raise ConfigError(f"indicator {name!r} not supported; choose from {sorted(SUPPORTED)}")
out = compute_indicator(df, name, **params)

from tradingagents.tools.analysis.indicators import SUPPORTED

def is_supported_indicator(name: str) -> bool:
    """Type guard against the library's SUPPORTED indicator set."""
    return isinstance(name, str) and name in SUPPORTED

try:
    out = compute_indicator(df, name, **params)
except ValueError as e:
    if "不支持的指标" in str(e):
        log.warning("skipping unsupported indicator %s", name)
        out = df.copy()
    else:
        raise

When it happens

Trigger: Calling compute_indicator(df, 'stoch') or compute_indicator(df, 'cci'); passing a typo like 'macd2', 'MA' (if case-sensitive), or 'bollinger'; or a compute_many spec list containing an unsupported name.

Common situations: Porting indicator lists from other TA libraries (TA-Lib names like 'STOCH', 'CCI', 'WILLR'); assuming a longer indicator menu than the 7 supported ones; case/format mismatches between config-driven indicator lists and SUPPORTED.

Understand the failure class

Background: UnsupportedOperationException and "is not supported" errors: when a library deliberately refuses a call — this error's family across 30 libraries.

Related errors

• Indicator {indicator} is not supported. Please choose from:
• REDDIT FETCHING ERROR: max limit is less than the number of
• Trading Agents Graph Setup Error: no analysts selected!
• QIANFAN_API_KEY格式错误，应为: bce-v3/ALTAK-xxx/xxx
• 不支持的OpenAI兼容提供商: {provider}

AI-assisted analysis of hsliuping/TradingAgents-CN@74783e8817 (2026-08-28). Data as JSON: /api/errors/92577eee8e332565. Report an issue: GitHub.