pyphi.campaign.CampaignTaskOutput

class pyphi.campaign.CampaignTaskOutput(task_id, pyphi_version, entries, metrics=None)

Bases: object

A task’s per-cell outcomes, aligned 1:1 with the task’s cells.

metrics records what the task cost to run: wall_s, cpu_s, cache_hits, cache_misses, cache_evictions, and for a shard task the payload_kind, units, memory_bytes, and n_mechanisms it was packed against. These are the observed side of what pyphi.cost predicts, so a campaign’s own outputs recalibrate pyphi.cost.SECONDS_PER_UNIT for the hardware it ran on.

Parameters:

• task_id (int)

• pyphi_version (str)

• entries (tuple[CellOutput, ...])

• metrics (dict[str, Any] | None)