Refinement errors (index is out of bounds)

[JuhaHuiskonen]

I have picked particles from visions using templates. This seemed to work. However, when running refinement, I am getting this error for some, but not all, tiltseries. Any clues what is causing this?

81	2024-02-14 21:47:06 [INFO] /opt/pyp/bin/run/pyp:3923 | PYP (cspswarm) finished successfully
82	2024-02-14 21:47:10 [INFO] /opt/pyp/bin/run/pyp:3273 | Job 3218860_10 (v0.6.2) launching on c5152.mahti.csc.fi using 128 task(s) and 200 GB of RAM
83	2024-02-14 21:47:10 [INFO] pyp/refine/csp/particle_cspt.py:57 | Merging particle stack took: 00h 00m 00s
84	2024-02-14 21:47:10 [INFO] pyp/refine/csp/particle_cspt.py:427 | Merging parameter files
85	2024-02-14 21:47:10 [INFO] pyp/refine/csp/particle_cspt.py:488 | Running reconstruction
86	2024-02-14 21:47:10 [INFO] pyp/refine/csp/particle_cspt.py:57 | Total time elapsed (csp_local_merge): 00h 00m 00s
87	Traceback (most recent call last):
88	  File "/opt/pyp/bin/run/pyp", line 3992, in <module>
89	    particle_cspt.merge_movie_files_in_job_arr(
90	  File "/opt/pyp/src/pyp/utils/timer.py", line 78, in wrapper_timer
91	    return func(*args, **kwargs)
92	  File "/opt/pyp/src/pyp/refine/csp/particle_cspt.py", line 522, in merge_movie_files_in_job_arr
93	    generate_ministar(current_block_par, movie_list, z_thicknes, binning, cls=current_class, output_path=star_output)
94	  File "/opt/pyp/src/pyp/inout/metadata/core.py", line 2957, in generate_ministar
95	    coords = coordinats_all[pid, :]
96	IndexError: index 315 is out of bounds for axis 0 with size 315
97	2024-02-14 21:47:10 [ERROR] /opt/pyp/bin/run/pyp:4012 | PYP (csp_local_merge) failed
98	2024-02-14 21:47:11 [INFO] /opt/pyp/bin/run/pyp:4018 | Deleted temporary files from /scratch/project_2009485/nextPYP/tmp/huiskone/3218860_10

Edit: I checked and all the tilt series that fail, have picked particles. Those that have none, are skipped correctly it seems. About half of those that have picked particles work.

[yezhou867]

I know a situation when re-running the preprocessing job after curating the virions or there may be some failed jobs at the first run then rerunning the job after merging will create this kind of bug. We fix it and will include the change in the next version.

[JuhaHuiskonen]

I confirm that if I run refinement directly after pre-processing, the refinement completes without errors. The errors occurred when I had first removed some bad virion segmentations.

[abartesaghi]

This issue is addressed in v0.6.3.