Local motion correction (multi) error

This is reproducible, tried with 2 or 3 GPUs. Runs OK with single GPU version (not multi).

Launching job on lane default target machinename …
License is valid.
Running job on master node hostname machinename
Project P4 Job J12 Started
Master running v2.0.27, worker running v2.0.27
Running on lane default
Resources allocated:
Worker: machinename
CPU : [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]
GPU : [0, 1]
RAM : [0, 1, 2, 3]
SSD : False

Importing job module for job type local_motion_correction_multi…
Job ready to run

Particles already have CTF estimates: these will not be modified.
num_movies 2325
parent process is 3163
Calling CUDA init from 3241
– 0.0: processing J1/imported/20180425_1116_A001_G000_H001_D000.mrcs
Loading raw movie data from J1/imported/20180425_1116_A001_G000_H001_D000.mrcs …
Done in 14.62s
Processing …
Done in 4.74s
Curvature 2043723.250000 Smooth lambda cal 20.000000
Computed scale constant of 0.06674
Computed overall normalizer of 1.01788
Writing out particles…
Done in 0.04s
Traceback (most recent call last):
File “cryosparc2_worker/cryosparc2_compute/run.py”, line 78, in cryosparc2_compute.run.main
File “cryosparc2_worker/cryosparc2_compute/jobs/motioncorrection/run_local.py”, line 317, in cryosparc2_compute.jobs.motioncorrection.run_local.run_local_motion_correction_multi
KeyError: 12564213794753284476


Is there any chance that one of the movies has no particles in it? That could cause this error.

Please let me know if that is the case and I can look further into this issue. Thank you for reporting!


Yes, it’s possible. The dataset had 2300 movies.


That was indeed the case - we will be patching this in an upcoming release. Thank you for your help!