HKUDS/Vibe-Trading · error · ValueError

Unsupported model_type: {model_type}

Error message

Unsupported model_type: {model_type}

What it means

This walk-forward ML training loop in the ml-strategy skill instantiates a model by branching on model_type, supporting 'xgboost' (or a tree booster config) and 'ridge' (L2 logistic regression). Any other value falls through to else: raise ValueError(f"Unsupported model_type: {model_type}") — a typo or unsupported algorithm is rejected before fit().

Source

Thrown at agent/src/skills/ml-strategy/SKILL.md:159

            # Standardization: fit only on training set
            scaler = StandardScaler()
            X_train = scaler.fit_transform(X_train)

            # Build the model
            if model_type == "random_forest":
                model = RandomForestClassifier(
                    n_estimators=100, max_depth=5, random_state=42,
                )
            elif model_type == "gradient_boosting":
                model = GradientBoostingClassifier(
                    n_estimators=100, max_depth=3, learning_rate=0.05,
                    random_state=42,
                )
            elif model_type == "ridge":
                model = LogisticRegression(penalty="l2", C=1.0, random_state=42)
            else:
                raise ValueError(f"Unsupported model_type: {model_type}")

            model.fit(X_train, y_train)

        # Predict today
        X_today = features.iloc[i : i + 1].values
        if np.isnan(X_today).any():
            predictions.iloc[i] = 0.0
            continue

        X_today = scaler.transform(X_today)

        if hasattr(model, "predict_proba"):
            prob = model.predict_proba(X_today)[0, 1]
            predictions.iloc[i] = prob * 2 - 1  # [0,1] -> [-1,1]
        else:
            predictions.iloc[i] = float(model.predict(X_today)[0])

    # Output contract: no NaN, clipped to [-1, 1]

View on GitHub (pinned to 80ffdda44c)

Solutions

  1. Set model_type to exactly 'xgboost' or 'ridge' (lowercase, no whitespace)
  2. Strip/normalize the config value before use: model_type.strip().lower()
  3. If you need another estimator, extend the if/elif chain with a new branch that constructs and fits it

Example fix

# before
model = train_walkforward(features, target, model_type="RandomForest")

# after
model = train_walkforward(features, target, model_type="xgboost")  # or "ridge"
Defensive patterns

Strategy: type-guard

Validate before calling

SUPPORTED_MODELS = {"xgboost", "ridge"}

model_type = str(cfg.get("model_type", "")).strip().lower()
if model_type not in SUPPORTED_MODELS:
    raise ValueError(f"model_type must be one of {sorted(SUPPORTED_MODELS)}, got {model_type!r}")

Type guard

def is_supported_model_type(value: str) -> bool:
    return str(value or "").strip().lower() in {"xgboost", "ridge"}

Try / catch

try:
    model = train_walkforward(features, target, model_type=model_type)
except ValueError as e:
    if str(e).startswith("Unsupported model_type"):
        raise ConfigError(f"fix strategy config: {e}") from e
    raise

Prevention

When it happens

Trigger: Passing model_type strings such as 'random_forest', 'lstm', 'XGBoost' (wrong case), 'ridge ' (trailing space), or 'lightgbm' — typically from a strategy config YAML/CLI arg.

Common situations: Config copy-paste from other projects expecting scikit-learn estimator names, case sensitivity, whitespace from templated config files, or an attempt to use a model the skill simply doesn't implement.

Related errors

AI-assisted analysis of HKUDS/Vibe-Trading@80ffdda44c (2026-08-28). Data as JSON: /api/errors/b8b081b9192c8a9d. Report an issue: GitHub.