Project

General

Profile

Actions
Copy link

Bug #4500

closed

gautomatch fails

Added by Saikat Chowdhury about 8 years ago. Updated almost 7 years ago.

Status:
Closed
Priority:
High
Assignee:
Saikat Chowdhury
Category:
Image Processing
Target version:
Appion/Leginon 3.3
Start date:
10/04/2016
Due date:
% Done:

0%

Estimated time:
Affected Version:
Appion/Leginon 3.3
Show in known bugs:
No
Workaround:

Description

We were trying to run gautomatch through appion on a gpu node and it didn't work.

Following is the command we were trying to run to test on a micrograph:

$ gautomatch.py --overlapmult=1 --lp=30 --hp=1000 --pre_lp=8 --pre_hp=1000 --speed=1 --lsigma_D=200 --lsigma_cutoff=1.3 --lave_D=400 --runname=gauto1 --rundir=/gpfs/group/em/appion/saikat/16sep29g/extract/gauto1 --projectid=328 --session=16sep29g --no-rejects --no-commit --no-continue --mrclist=16sep29g_10a_00019gr_00007sq_00026hl_00003esup-b.mrc

This is the error we were getting.

  • ... Time stamp: 16oct03t51
    ... Function name: gautomatch
    ... Appion directory: /gpfs/home/glander/myami/appion
    ... Processing hostname: nodeb1203
    !!! WARNING: There is more than one appion directory in your PYTHONPATH
    ['/gpfs/home/glander/myami/appion/appionlib', '/gpfs/home/glander/myami/appion']
    Traceback (most recent call last):
    File "/gpfs/home/glander/myami/appion/bin/gautomatch.py", line 9, in <module>
    imgLoop = apGautomatch.GautomatchLoop()
    File "/gpfs/home/glander/myami/appion/appionlib/particleLoop2.py", line 92, in init
    appionLoop2.AppionLoop.__init__(self)
    File "/gpfs/home/glander/myami/appion/appionlib/appionLoop2.py", line 27, in init
    appionScript.AppionScript.__init__(self)
    File "/gpfs/home/glander/myami/appion/appionlib/appionScript.py", line 67, in init
    self.setParams(optargs,useglobalparams)
    File "/gpfs/home/glander/myami/appion/appionlib/appionScript.py", line 340, in setParams
    self.setupParserOptions()
    File "/gpfs/home/glander/myami/appion/appionlib/apGautomatch.py", line 98, in setupParserOptions
    self.parser.add_option("--minthresh",dest="minthresh",help="Cross-correlation cutoff, 0.2~0.4 normally; Try to select several typical micrographs to optimize this value.")
    File "/usr/lib64/python2.6/optparse.py", line 1020, in add_option
    self._check_conflict(option)
    File "/usr/lib64/python2.6/optparse.py", line 995, in _check_conflict
    option)
    optparse.OptionConflictError: option --minthresh: conflicting option string(s): --minthresh *
    We were also using python/2.6.6.

Seems like an issue with the appion wrapper

Actions
Copy link
 #1

Updated by Carl Negro about 8 years ago

Hi Saikat,

I had only committed some of my changes to the trunk, so the version of Gautomatch you tried to run is completely broken. I recently committed a cleaned-up version that should minimally work, just make sure you have the correct CUDA version of Gautomatch listed in the variable "exename" in trunk/appion/appionlib/apGautomatch.py. There are still some issues that I have to sort out with pikfiles, so you won't be able view the particles picked by individual templates from the image viewer after a successful run.

Also, Gautomatch can run in template correlator or DoG mode, but the Appion wrapper only lets you run it as a template correlator for now.

Carl

Actions
Copy link
 #2

Updated by Anchi Cheng about 8 years ago

  • Status changed from New to In Test
  • Assignee changed from Anchi Cheng to Saikat Chowdhury
  • Target version changed from Appion/Leginon 3.2 to Appion/Leginon 3.3

Does this work now ? I check Carl's updated code, it should be get pass this bug you reported.

Actions
Copy link
 #3

Updated by Saikat Chowdhury about 8 years ago

Dear Carl,

There is still a minor bug.
When we generate the command for running gautomatch from the appion web browser the "--pdiam" option shows up twice in the same command. Removing the duplicate "--pdiam" option from the generated command lets gautomatch run.

Thanks for the fixes.
-Saikat

Carl Negro wrote:

Hi Saikat,

I had only committed some of my changes to the trunk, so the version of Gautomatch you tried to run is completely broken. I recently committed a cleaned-up version that should minimally work, just make sure you have the correct CUDA version of Gautomatch listed in the variable "exename" in trunk/appion/appionlib/apGautomatch.py. There are still some issues that I have to sort out with pikfiles, so you won't be able view the particles picked by individual templates from the image viewer after a successful run.

Also, Gautomatch can run in template correlator or DoG mode, but the Appion wrapper only lets you run it as a template correlator for now.

Carl

Actions
Copy link
 #4

Updated by Anchi Cheng almost 7 years ago

  • Status changed from In Test to Closed

--pdiam not duplicated any more

Actions
Copy link

Also available in: Atom PDF