Skip to content

Index

cycles ¤

Cycles

Utilities to detect and process cycles in timeseries.

  • CycleExtractor: Build cycles from state/step/value changes.
  • process_persistent_cycle: A True stretch (rising to falling edge) is a cycle.
  • process_trigger_cycle: Each trigger (rising edge) runs to the next one.
  • process_separate_start_end_cycle: Separate start and end signals.
  • process_step_sequence: Entering a start step opens, an end step closes.
  • process_state_change_cycle: Every sample is a boundary.
  • process_value_change_cycle: Any significant value change is a boundary.

  • validate_cycles / detect_overlapping_cycles / suggest_method.

  • CycleDataProcessor: Split/merge/group by cycle windows.

  • split_by_cycle: Split values by cycle ranges.
  • merge_dataframes_by_cycle: Annotate values with cycle UUIDs.
  • group_by_cycle_uuid: Group values by cycle key.
  • split_dataframes_by_group: Further split by column groupings.
  • compute_cycle_statistics / compare_cycles / identify_golden_cycles. A sample belongs to a cycle when cycle_start <= systime < cycle_end.