DryRunReport¶

class DryRunReport(pipeline_name: str, stages: tuple[StageInfo, ...], total_circuits: int, env_artifacts: dict[str, Any] = {}, circuit_stats: dict[str, float] = {})[source]¶

Complete dry-run report for a single pipeline.

Create new instance of DryRunReport(pipeline_name, stages, total_circuits, env_artifacts, circuit_stats)

Attributes Summary

`circuit_stats`	Aggregate depth/width stats across the post-fan-out final batch's DAG bodies — the pre-execution analogue of `depth_history`.
`env_artifacts`	Stage-produced artifacts captured during the forward pass — e.g. `per_group_shots` (when a `shot_distribution` is configured on `MeasurementStage`), `ham_ops` (for expval-native backends).
`pipeline_name`	Alias for field number 0
`stages`	Alias for field number 1
`total_circuits`	Alias for field number 2

Attributes Documentation

circuit_stats: dict[str, float]¶: Aggregate depth/width stats across the post-fan-out final batch’s DAG bodies — the pre-execution analogue of depth_history. Empty when the final batch has no DAG bodies (e.g. probability-mode pipelines that only carry bound QASM strings). Populated keys: mean_depth, std_depth, min_depth, max_depth, mean_width, std_width, min_width, max_width, mean_2q_depth.

env_artifacts: dict[str, Any]¶: Stage-produced artifacts captured during the forward pass — e.g. per_group_shots (when a shot_distribution is configured on MeasurementStage), ham_ops (for expval-native backends). These are the same artifacts the pipeline would produce on a real run, so a dry-run report is the canonical surface for "what would my pipeline do?" introspection — no need to drop into private helpers or rerun the forward pass manually.