apache/beam · error · ValueError

subspace for not found.

Error message

subspace for {spec_type} not found.

What it means

_spec_type_to_subspace maps a Spec type string to the subspace (e.g. detector, transformation, scalers) where its registered Specifiable class lives. If the type is not registered in any known subspace in _KNOWN_SPECIFIABLE, a ValueError naming the type is raised. from_spec calls this before looking up the concrete subclass.

Solutions

1. Check the exact registered type string via Specifiable classes / _KNOWN_SPECIFIABLE and fix the Spec.type spelling.
2. Decorate your custom class with @specifiable so it registers in a subspace.
3. Register the type before from_spec runs (imports must execute the registration code).
4. If migrating from an older Beam release, update spec type names to the current registry.

Example fix

// before
Spec(type="RobustZ" , config=None)
// after
Spec(type="RobustZScore", config=None)

from apache_beam.ml.anomaly import specifiable
assert any(spec.type in known for known in specifiable._KNOWN_SPECIFIABLE.values()), f"unregistered type: {spec.type}"

def is_registered_type(t): return any(t in known for known in _KNOWN_SPECIFIABLE.values())

try:
    obj = Specifiable.from_spec(spec)
except ValueError as e:
    logger.error("bad spec type %r: %s", spec.type, e)
    raise

When it happens

Trigger: Calling Specifiable.from_spec with a Spec whose .type string is not registered, is misspelled, or refers to a class that was never registered via the @specifiable decorator; also triggered by test_default_inference_fn with unknown types.

Common situations: Typos in spec type strings ('zscore' vs 'ZScore'); custom detectors used in a spec without registering them; Beam version changes renaming built-in spec types; deserialized specs from older pipeline definitions.

Understand the failure class

Background: Invalid enum value errors: "Unknown type", "Invalid scope", "must be one of" — when a string is not on the library's allowed list — this error's family across 23 libraries.

Related errors

• Unknown spec type ' ' in
• Spec type not found in
• Coder registry has no fallback coder. This can happen if…
• "IQR.learn_one expected univariate input, but got
• "IQR.score_one expected univariate input, but got

AI-assisted analysis of apache/beam@12126d8942 (2026-09-13). Data as JSON: /api/errors/0259135c9dcebbdb. Report an issue: GitHub.

Appendix: source

Thrown at sdks/python/apache_beam/ml/anomaly/specifiable.py:89

  if hasattr(cls, "mro"):
    # some classes do not have "mro", such as functions.
    for c in cls.mro():
      if c.__name__ in _ACCEPTED_SUBSPACES:
        return c.__name__

  return _FALLBACK_SUBSPACE


def _spec_type_to_subspace(spec_type: str) -> str:
  """
  Look for the subspace for a spec type. This is usually called to retrieve
  the subspace of a registered specifiable class.
  """
  for subspace in _ACCEPTED_SUBSPACES:
    if spec_type in _KNOWN_SPECIFIABLE[subspace]:
      return subspace

  raise ValueError(f"subspace for {spec_type} not found.")


@dataclasses.dataclass(frozen=True)
class Spec():
  """
  Dataclass for storing specifications of specifiable objects.
  Objects can be initialized using the data in their corresponding spec.
  """
  #: A string indicating the concrete `Specifiable` class
  type: str
  #: An optional dictionary of keyword arguments for the `__init__` method of
  #: the class. If None, when we materialize this Spec, we only return the
  #: class without instantiate any objects from it.
  config: Optional[dict[str, Any]] = dataclasses.field(default_factory=dict)


def _specifiable_from_spec_helper(v, _run_init):
  if isinstance(v, Spec):

View on GitHub (pinned to 12126d8942)