Add script to pair slurm job runtimes with number of sources

tools/analyze_logs.py

@@ -0,0 +1,110 @@
+#!/usr/bin/env python
+import pathlib
+import argparse
+import pandas as pd
+import warnings
+from datetime import timedelta
+import json
+
+BASE_DIR = pathlib.Path(__file__).parent.parent.absolute()
+
+
+def get_parser():
+
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument(
+        "--job-ids-path",
+        type=str,
+        help="path (from base_dir) to file containing slurm job ids",
+    )
+    parser.add_argument(
+        "--logs-path",
+        type=str,
+        help="path (from base_dir) to slurm logs",
+    )
+    parser.add_argument(
+        "--logs-name-pattern",
+        type=str,
+        default="",
+        help="common naming convention for slurm logs (e.g. generate_features)",
+    )
+    parser.add_argument(
+        "--logs-suffix",
+        type=str,
+        default="out",
+        help="suffix for log files",
+    )
+    parser.add_argument(
+        "--output-path",
+        type=str,
+        default="runtime_output.json",
+        help="path (from base_dir) to output file",
+    )
+
+    return parser
+
+
+def main(
+    job_ids_path,
+    logs_path,
+    logs_name_pattern="",
+    logs_suffix="out",
+    output_path="runtime_output.json",
+):
+    job_ids = pd.read_table(BASE_DIR / job_ids_path, header=None)
+
+    logs_path = BASE_DIR / logs_path
+
+    results_dct = {}
+    for id_runtime_pair in job_ids.values:
+        job_id, runtime = id_runtime_pair[0].split(',')
+
+        log_path = logs_path.glob(f"{logs_name_pattern}_{job_id}_*.{logs_suffix}")
+        log_path = [x for x in log_path]
+
+        if len(log_path) > 0:
+            log_output = pd.read_table(log_path[0], header=None)
+
+            n_sources_start = int(log_output.iloc[2].values[0].split()[1])
+            n_sources_end = int(log_output.iloc[-1].values[0].split()[3])
+
+            runtime_split = runtime.split("-")
+            if len(runtime_split) == 1:
+                runtime_days = 0
+                runtime_hms = runtime_split[0].split(":")
+            else:
+                runtime_days = int(runtime_split[0])
+                runtime_hms = runtime_split[1].split(":")
+
+            runtime_hours = int(runtime_hms[0])
+            runtime_minutes = int(runtime_hms[1])
+            runtime_seconds = int(runtime_hms[2])
+
+            delta = timedelta(
+                days=runtime_days,
+                hours=runtime_hours,
+                minutes=runtime_minutes,
+                seconds=runtime_seconds,
+            )
+            total_seconds = delta.total_seconds()
+
+            results_dct[int(job_id)] = {
+                "n_sources_start": n_sources_start,
+                "n_sources_end": n_sources_end,
+                "runtime_seconds": total_seconds,
+                "seconds_per_source_start": total_seconds / n_sources_start,
+            }
+
+        else:
+            warnings.warn(f"Could not find log for job ID {job_id}")
+
+    with open(BASE_DIR / output_path, "w") as f:
+        json.dump(results_dct, f)
+    print(f"Wrote results to {BASE_DIR / output_path}")
+
+
+if __name__ == "__main__":
+    parser = get_parser()
+    args = parser.parse_args()
+    main(**vars(args))