maubin/sbmy_control
0
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

small fixes

Browse source
This commit is contained in:
Mayeul Aubin 2025-05-19 11:48:11 +02:00
parent d0a8439be7
commit a485db9465
3 changed files with 16 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

0
analysis/slices.py Normal file → Executable file
View file

0
scripts/__init__.py Normal file
View file

29
scripts/scola_submit.py
View file

@ -152,9 +152,15 @@ def check_previous_jobs(args,job_ids_array,box,resubmit_count,error_count):
if job_id is None:
print(f"Error submitting job for box {box}")
error_count+=1
# Check if the error count exceeds the maximum
if error_count >= MAX_ERRORS:
raise RuntimeError(f"Error count exceeded {MAX_ERRORS}. Stopping job submission.")
else:
job_ids_array[prev_box-1] = int(job_id)
resubmit_count += 1
# Check if the resubmit count exceeds the maximum
if resubmit_count >= MAX_RESUBMIT:
raise RuntimeError(f"Resubmit count exceeded {MAX_RESUBMIT}. Stopping job submission.")
job_status_categories[status].append(prev_box) # Classify as failed
# Sleep for a while before resubmitting the next job
time.sleep(args.sleep)
@ -199,11 +205,11 @@ def print_summary_job_status(job_status_categories, box, resubmit_count, error_c
# Print summary of job statuses
print(f"Job statuses after box {box}:")
# Print a table with columns for each status and below the % of jobs in that status
row0 = f"{'Status':<10}"
row0 = f"{'Status':<14}"
for status in job_status_categories.keys():
row0 += f"{status:>10}"
row0 += f"{status:>9} "
print(row0)
row1 = f"{'Percentage':<10}"
row1 = f"{'Percentage':<14}"
for status in job_status_categories.keys():
row1 += f"{len(job_status_categories[status])/box*100:>9.1f}%"
print(row1)
@ -220,13 +226,13 @@ if __name__ == "__main__":
parser.add_argument("-sd", "--slurmdir", type=str, default=None, help="Directory where the slurm scripts are saved (default is -d/slurm_scripts).")
parser.add_argument("-wd", "--workdir", type=str, default=None, help="Directory where the tiles are saved (default is -d/work).")
parser.add_argument("-sf","--slurmfile", type=str, default="scola_sCOLA.sh", help="Slurm script file (located in slurmdir).")
parser.add_argument("-tf","--tilefile", type=str, default="scola_tile", help="Tile file name (located in workdir).")
parser.add_argument("--jobname", type=str, default="sCOLA_", help="Job name for the slurm jobs.")
parser.add_argument("-sf","--slurmfile", type=str, default="scola_sCOLA.sh", help="Slurm script file (located in slurmdir, default is scola_sCOLA.sh).")
parser.add_argument("-tf","--tilefile", type=str, default="scola_tile", help="Tile file name (located in workdir, default is scola_tile).")
parser.add_argument("--jobname", type=str, default="sCOLA_", help="Job name for the slurm jobs (default is sCOLA_).")
parser.add_argument("-Nt","--N_tiles", type=int, default=4, help="Number of tiles per dimension.")
parser.add_argument("--sleep", type=float, default=1.0, help="Sleep time between each job submission (in s).")
parser.add_argument("--sleep", type=float, default=1.5, help="Sleep time between each job submission (in s).")
args=parser.parse_args()
@ -268,7 +274,7 @@ if __name__ == "__main__":
print(f"Max jobs at once: {MAX_JOBS_AT_ONCE} jobs")
print(f"Check every: {CHECK_EVERY} jobs")
print("---------------------------------------------------")
print(f"ETA: {convert_seconds_to_time(args.N_tiles**3*args.sleep*1.05)}")
print(f"ETA: {convert_seconds_to_time(args.N_tiles**3*args.sleep*1.2)}")
print("Starting job submission...")
@ -329,12 +335,6 @@ if __name__ == "__main__":
# Now wait for all jobs to finish
while len(job_status_categories['CP'])<args.N_tiles**3:
time.sleep(10*args.sleep)
# Check if the error count exceeds the maximum
if error_count >= MAX_ERRORS:
raise RuntimeError(f"Error count exceeded {MAX_ERRORS}. Stopping job submission.")
# Check if the resubmit count exceeds the maximum
if resubmit_count >= MAX_RESUBMIT:
raise RuntimeError(f"Resubmit count exceeded {MAX_RESUBMIT}. Stopping job submission.")
job_status_categories, resubmit_count, error_count = check_previous_jobs(args,job_ids_array,args.N_tiles**3+1,resubmit_count,error_count)
print_summary_job_status(job_status_categories, args.N_tiles**3+1, resubmit_count, error_count)
job_status_categories = cap_number_of_jobs(job_status_categories,job_ids_array,MAX_JOBS_AT_ONCE,args.sleep)
@ -343,5 +343,6 @@ if __name__ == "__main__":
print("All jobs finished.")
# Remove the slurm scripts
for box in range(1,args.N_tiles**3+1):
if os.path.exists(args.slurmdir+args.slurmfile+"."+str(box)):
os.remove(args.slurmdir+args.slurmfile+"."+str(box))