apache/beam · error · ValueError

The number of elements in the provided pre-timestamped data…

Error message

The number of elements in the provided pre-timestamped data sequence is not enough to span the full impulse duration. Expected duration: %s, actual data duration: %s. Please either provide more data or decrease `stop_timestamp`.

What it means

PeriodicImpulse/PeriodicSequence with pre-timestamped elements validates that the provided data spans the requested impulse duration. For pre-timestamped data, if the data duration is shorter than stop_timestamp, a ValueError is raised (non-pre-timestamped data is just repeated with a warning).

Solutions

Provide more pre-timestamped elements covering the full duration.
Decrease the stop_timestamp to match the actual data duration.
Set pre_timestamped_data=False if repeating the data is acceptable.

Example fix

// before
PeriodicSequence(beam.Timestamp(0), data, stop_timestamp=100, pre_timestamped_data=True)  # data spans 10s
// after
PeriodicSequence(beam.Timestamp(0), data, stop_timestamp=10, pre_timestamped_data=True)

Defensive patterns

Strategy: validation

Validate before calling

data_duration = ts[-1] - ts[0]  # for pre-timestamped data
assert data_duration >= stop_timestamp, f"data spans {data_duration} < stop_timestamp {stop_timestamp}"

Prevention

Compute data span before choosing stop_timestamp
Prefer letting the transform repeat data unless exact timestamps matter

When it happens

Trigger: Constructing PeriodicSequence with pre_timestamped_data=True where the last element timestamp minus first is less than stop_timestamp.

Common situations: Generating test data streams with a fixed small dataset but a long stop_timestamp; converting notebook demos to longer runs without adding data.

Understand the failure class

Background: "value must be between 0 and 1" / "out of range" / "must not be negative" errors: fixing range-validation failures across open-source libraries — this error's family across 42 libraries.

Related errors

AI-assisted analysis of apache/beam@12126d8942 (2026-09-13). Data as JSON: /api/errors/cd1ca6025ec9004a. Report an issue: GitHub.

Appendix: source

Thrown at sdks/python/apache_beam/transforms/periodicsequence.py:285

      # When the stop timestamp is unbounded (MAX_TIMESTAMP), set it to the
      # data's actual end time plus an extra fire interval, because the
      # impulse duration's upper bound is exclusive.
      self.stop_ts = start_ts + data_duration + Duration(self.interval)
      stop_ts = self.stop_ts

    # The total time for the impulse signal which occurs in [start, end).
    impulse_duration = stop_ts - start_ts
    if data_duration + Duration(self.interval) < impulse_duration:
      # We don't have enough data for the impulse.
      # If we can fit at least one more data point in the impulse duration,
      # then we will be in the repeat mode.
      message = 'The number of elements in the provided pre-timestamped ' \
        'data sequence is not enough to span the full impulse duration. ' \
        f'Expected duration: {impulse_duration}, ' \
        f'actual data duration: {data_duration}.'

      if is_pre_timestamped:
        raise ValueError(
            f'{message} Please either provide more data or decrease '
            '`stop_timestamp`.')
      else:
        warnings.warn(
            f'{message} As a result, the data sequence will be repeated to '
            'generate elements for the entire duration.')

  def __init__(
      self,
      start_timestamp: TimestampTypes = Timestamp.now(),
      stop_timestamp: TimestampTypes = MAX_TIMESTAMP,
      fire_interval: float = 360.0,
      apply_windowing: bool = False,
      data: Optional[Sequence[Any]] = None,
      rebase: RebaseMode = RebaseMode.REBASE_NONE):
    '''
    :param start_timestamp: Timestamp for first element.
    :param stop_timestamp: Timestamp at or after which no elements will be

View on GitHub (pinned to 12126d8942)