Bug #1359
closed
stack remake from a stack may have different particle order
Added by Anchi Cheng almost 13 years ago.
Updated over 8 years ago.
Category:
Python scripting
Affected Version:
Appion/Leginon 2.2.0
Description
Makestack loop through images and box particles that are found in the stack particle database record. If the stack is a combined stack or a junk-sorted stack, the resulting stack would not have the same order as the stack it is meant to be duplicated from.
This is not a problem if the new stack is committed and query is used to match particles but would be a problem if not committed as I intend to do for frealign refinestack.
Files
We have tiltStackSync and I thought we had a defocalStackSync (but I cannot find it) that sorts you stacks after you have made them. Forcing makestack to systemically go to each image in the same order could break other functions, so I would suggest writing a work around that would re-sort the particles in the stack.
Second you have to worry that if increased the boxsize, more particles will be excluded because they are too close to the edge, so you will never get a perfect sync.
- Category set to Python scripting
- Status changed from New to Assigned
- Assignee set to Anchi Cheng
- Target version set to Appion/Leginon 2.2.0
- Affected Version changed from Appion/Leginon 2.1.0 to Appion/Leginon 2.2.0 (trunk)
I will start working on this refinestack sync now that frealign is working and I want to test it on combined stacks.
Anchi, shall we move this to 3.0?
Wish we don't have to. This one need to be listed as known bug. I am sure people will try this in 2.2 but find it not working right.
- Target version changed from Appion/Leginon 2.2.0 to Appion/Leginon 2.2.1
- Show in known bugs changed from No to Yes
Putting this in known bugs.
This is the issue that is holding back frealign usage.
I've got a bit of a fix for now...but just attaching the files to this issue since I have not had a chance to do complete testing. I'll be out for the next 5 weeks, so hopefully Anchi can pick up where I left off. Need to fix an issue with losing the stack particle data just before callProcessParticles() in apParticleExtractor.py, or find a more efficient way of getting the particle number.
Because Frealign can not take particles that are not grouped by film number, I have to abandon the original idea of resorting the preped stack to the previous stack order. Instead, the corresponding particle number are saved in a file and matched during the upload of the refinement.
r16897 contains these changes.
- Status changed from Assigned to In Test
- Affected Version changed from Appion/Leginon 2.2.0 (trunk) to Appion/Leginon 2.2.0
- Status changed from In Test to Closed
Also available in: Atom
PDF