benchmark.logparser Module¶
- class paraview.benchmark.logparser.BasicStats(samples=[])[source]¶
Bases:
object
- property Mean¶
- property StdDev¶
- class paraview.benchmark.logparser.FrameLog(parent=None, indent=0)[source]¶
Bases:
object
Hold the log entries for an entire (sub-)frame
- class paraview.benchmark.logparser.FrameLogEntry(log_msg)[source]¶
Bases:
object
A basic container for holding timing information for a given filter
- re = <module 're' from '/usr/lib64/python3.10/re.py'>¶
- paraview.benchmark.logparser.collect_stats(frame_logs, stat_summary='Mean')[source]¶
Collect statistics on the specified key in a set of logs
- paraview.benchmark.logparser.process_logs(merge_before_nframes=0)[source]¶
Collect and parse logs for all ranks
Keyword arguments: merge_before_nframes – All entries before this many frames will be merged
- paraview.benchmark.logparser.process_stats_across_ranks(rank_frame_logs)[source]¶
Calculate stats across all ranks for each frame
- paraview.benchmark.logparser.summarize_results(num_frames, num_seconds_m0, items_per_frame, item_label, save_logs=False, output_basename=None)[source]¶
Process the timing logs to display, save, and gather stats
Keyword arguments: num_frames – Number of frames to process num_seconds_m0 – Total number of seconds, excluding the first frame items_per_frame – Number of items per frame getting processed item_label – Output label for associated items_per_frame save_logs – Whether or not to write the logs to a file output_basename – Basename to use for output files