Local variable 'headerdict' referenced before assignment

[asarnow]

When I re-run a preprocessing block, all of the tilt-series will fail with the following error. Could cancelling a re-run or another specific action lead to this? I'm not sure if the previous run data are OK or not - the previews in the dashboard are missing but you can still browse the reconstructions/tilt-series within the block.

1	2024-03-19 22:46:25 [INFO] /opt/pyp/bin/run/pyp:3232 | Job 1131694_1 (v0.6.3) launching on emnodec41 using 8 task(s) and 8 GB of RAM
2	2024-03-19 22:46:26 [INFO] /opt/pyp/bin/run/pyp:3178 | Filesystem      Size  Used Avail Use% Mounted on
3	2024-03-19 22:46:26 [INFO] /opt/pyp/bin/run/pyp:3178 | /dev/sdb1       7.3T  6.3G  7.3T   1% /scratch/pyp
4	2024-03-19 22:46:26 [INFO] /opt/pyp/bin/run/pyp:1347 | Working path: /scratch/pyp/nextpyp/1131694_1/23dec05a-u2os-spike_142
5	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving tomo-preprocessing-f2iO2oFOonwh3z8V.micrographs
6	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving .pyp_config.toml
7	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving tomo-preprocessing-f2iO2oFOonwh3z8V.films
8	2024-03-19 22:46:26 [INFO] pyp/system/mpi.py:180 | First function is: load_files
9	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142_raw.webp
10	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142.rec
11	2024-03-19 22:46:26 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142.mrc
12	2024-03-19 22:46:27 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142_sides.webp
13	2024-03-19 22:46:27 [INFO] pyp/inout/utils/core.py:17 | Retrieving virion_thresholds.next
14	2024-03-19 22:46:27 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142_rec.webp
15	2024-03-19 22:46:27 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142.pkl
16	2024-03-19 22:46:28 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142_ali.webp
17	2024-03-19 22:46:29 [INFO] pyp/inout/utils/core.py:17 | Retrieving 23dec05a-u2os-spike_142.webp
18	2024-03-19 22:46:51 [INFO] /opt/pyp/bin/run/pyp:57 | Loading results took: 00h 00m 25s
19	2024-03-19 22:46:52 [INFO] pyp/inout/image/core.py:245 | Using gain reference: CountCDSRef_DA27-3_51_0000_0.0.mrc, shape: (4092, 5760), min: 0.514263, max: 1.779763, mean: 1.002406
20	2024-03-19 22:46:52 [INFO] /opt/pyp/bin/run/pyp:57 | Total time in tomo_swarm 00h 00m 26s seconds
21	2024-03-19 22:46:52 [INFO] /opt/pyp/bin/run/pyp:3178 | Filesystem      Size  Used Avail Use% Mounted on
22	2024-03-19 22:46:52 [INFO] /opt/pyp/bin/run/pyp:3178 | /dev/sdb1       7.3T   20G  7.3T   1% /scratch/pyp
23	Traceback (most recent call last):
24	  File "/opt/pyp/bin/run/pyp", line 3448, in <module>
25	    tomo_swarm(args.path, args.file, args.debug, args.keep, args.skip)
26	  File "/opt/pyp/src/pyp/utils/timer.py", line 78, in wrapper_timer
27	    return func(*args, **kwargs)
28	  File "/opt/pyp/bin/run/pyp", line 1466, in tomo_swarm
29	    ] = preprocess.read_tilt_series(
30	  File "/opt/pyp/src/pyp/preprocess/core.py", line 370, in read_tilt_series
31	    x, y, z = imageio.readMoviefileandsave(name + ".mrc", parameters, binning)
32	  File "/opt/pyp/src/pyp/inout/image/core.py", line 990, in readMoviefileandsave
33	    total_header_size = 1024 + int(headerdict["nsymbt"])
34	UnboundLocalError: local variable 'headerdict' referenced before assignment
35	2024-03-19 22:46:52 [ERROR] /opt/pyp/bin/run/pyp:3453 | PYP (tomoswarm) failed

[asarnow]

Seems like a real bug. In pyp.inout.image.core.readMoviefileandsave(), if grouping >= 2 (else branch currently line 985) then headerdict is accessed but never instantiated.

As an aside: I've been fractionating finely ~0.15 e-/Å^2/fm for 20 frames, and grouping 4 frames. The cost is low as the movies are still small compared to SPA movies, it's useful for testing as you can make it coarser but not finer post facto, and it leads to better CTF estimation in Relion where averaged movie power spectra (not power spectrum of the average) can be used.

[yezhou867]

Yes. Thanks for reporting this problem. We will fix it in the next version.

[abartesaghi]

This was addressed in version 0.6.4.