Bug #3660
open
Mean/StdDev filtering of stack not working with virtual stack
Added by Priyamvada Acharya almost 9 years ago.
Updated almost 9 years ago.
Affected Version:
Appion/Leginon 3.2
Workaround:
click the "*generate some missing information*" in stack summary table to generate the real stack before running this
Description
Error message:
subStack.start()
File "/opt/myamisnap/bin/stackFilter.py", line 88, in start
apStack.makeNewStack(oldstack, newstack, self.params['keepfile'], bad=True)
File "/opt/myamisnap/lib/appionlib/apStack.py", line 55, in makeNewStack
a.run()
File "/opt/myamisnap/lib/appionlib/proc2dLib.py", line 45, in run
self.approc2d.checkConflicts()
File "/opt/myamisnap/lib/appionlib/proc2dLib.py", line 112, in checkConflicts
self.inputNumParticles = self.inheader['nz']
TypeError: 'NoneType' object is unsubscriptable
Session ID: 15oct05d
Path: /gpfs/appion/pacharya/15oct05d/stacks/alignsub53/
- Project changed from 138 to Appion
- Assignee set to Priyamvada Acharya
- Affected Version set to Appion/Leginon 3.2
The error in /gpfs/appion/pacharya/15oct05d/stacks/meanfilt103
!!! WARNING: could not find old stack: /gpfs/appion/pacharya/15oct05d/stac
ks/alignsub53/start.hed
alignsub53 Ended at Wed, 14 Oct 2015 11:17:50 but
the stack modified time was Oct 14 15:00
It seems that the stack in question was not generated until 15:00. The stackFilter.py ran after that worked fine.
Any comment on why the stack was modified after the run ended ? Is this a virtual stack ? Could this be related to virtual stack function ?
I had the same issue... Actually some times the stack created shows like "*generate some missing information*" in the place of size of the stack in MB.
So, when I tried to click the "*generate some missing information*" it shows a command which as I remember starts with something like "generate information". I copied the command and run over the terminal in cluster and then size of the stack showed up.
After this any stack operation worked. Don't know what causing this issue... But there is work around for now. I hope this info helps to debug.
the Mean/STDdev function hasn't been fixed to work with virtual stacks yet. Nor has xmipp projection matching. Unfortunately I haven't had the time to dedicate to this, in the meantime continue with the workaround of generating the stack.
Alternatively, when you make a substack, there's a checkbox to make the actual file instead of a virtual stack.
- Subject changed from Mean/StdDev filtering of stack not working to Mean/StdDev filtering of stack not working with virtual stack
- Status changed from New to Assigned
- Assignee changed from Priyamvada Acharya to Gabriel Lander
- Workaround updated (diff)
Also available in: Atom
PDF