Feature #1342
open
prepare stack and model for 3D refinement
Added by Anchi Cheng over 13 years ago.
Updated over 13 years ago.
Description
Move all refineparameter settings to job class called from refineJob.py which will sit on the cluster and work with agent and cluster class to launch the job
Therefore prepRefine only need to prepare stack and initial model to match scaling and write files required to be copied.
- Assignee changed from Anchi Cheng to Amber Herold
r15722 has the base class apPrepRefine.py and the first working subclass prepRefineEman.py
I made an example with this:
prepRefineEman.py --jobtype=preprefine --expid=8556 --projectid=303 --stackid=26 --modelid=1 --runname=eman_recon1 --bin=2
The database tables should help producing the runRefineJob step.
ApPrepRefineData (query session and method should give preped jobs prepared for a particular refinement jobtype for runRefineJob GUI.
ApRefineStackData references ApPrepRefineData, normally one to one correspondance but may be two to one for frealign when reconstack is different from refinestack.
ApRefineInitModelData references ApPrepRefineData, normally one to one unless for multimodel refinement
files_to_remote_host in rundir contains absolute path to files that need to be send to remote processing host
r15723 and r15724 added description field in ApPrepRefineData table
r15744 add xmipp prep.
prepRefineXmipp.py is for single model projection matching
prepRefineXmippML3D.py is for multi-model ML3D classification
- Assignee changed from Amber Herold to Anchi Cheng
r15812 and r15813 add prepRefineFrealign
This only works in cases where the particles in the stack has not being sorted or combined. See issue #1365
Also available in: Atom
PDF