Streamlining of autoindexer workflow
This merge request introduces a significant streamlining of the autoindexer part of the workflow. Previously, the script would run PeakFinder
on a small number of frames, then attempt to autoindex. If this failed, it would add a frame, find peaks again, and so on, which was pretty inefficient.
I have introduced two main changes:
- Ability to filter out peaks outside a specific frame range (this just uses the third coordinate of
peak->shape().center()
, the non-integer (possibly interpolated) frame number. - Autoindexing functions in the
Experiment
class. These functions usePeakFilter
to get aPeakCollection
containing only peaks from a limited contiguous subset (of around 10, to start) frames, and attempt to autoindex. If this fails, the filter will add 1 frame to the range and retry.
I have updated the script workflow.py
to reflect these changes. Note that this means that the GUI workflow will need far less modification.
Edited by Raza, Zamaan