You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At the moment FitFactories need to be in core PYME to be discovered, yet many in the SMLM community have reasons to develop new/use old fitters, etc., and some have expressed interest in adding their algorithms (e.g. phasor-based localization) to PYME. This would also be a great opportunity to clean things up a bit, tidy up/unify units/offsets/etc. while making the then-legacy fitters available in plugins so people can still easily rerun old routines for reproducibility.
Lessons learned when shimming in the warp-drive fit is that there are quite some units/reference frames which need to be documented well to get everything to play nice/not duplicate computation doing/undoing conversions. This should be considerably smoother now, but would be worth touching up again (I think specifically with the multiview metadata stuff I've used with the AstigGaussGPUFitFR the camera ROI offset is dropped, which is only a problem is you want to go back post facto and re-map localizations to the raw camera frames (outside of the test frame functionality).
The text was updated successfully, but these errors were encountered:
At the moment FitFactories need to be in core PYME to be discovered, yet many in the SMLM community have reasons to develop new/use old fitters, etc., and some have expressed interest in adding their algorithms (e.g. phasor-based localization) to PYME. This would also be a great opportunity to clean things up a bit, tidy up/unify units/offsets/etc. while making the then-legacy fitters available in plugins so people can still easily rerun old routines for reproducibility.
Lessons learned when shimming in the warp-drive fit is that there are quite some units/reference frames which need to be documented well to get everything to play nice/not duplicate computation doing/undoing conversions. This should be considerably smoother now, but would be worth touching up again (I think specifically with the multiview metadata stuff I've used with the
AstigGaussGPUFitFR
the camera ROI offset is dropped, which is only a problem is you want to go back post facto and re-map localizations to the raw camera frames (outside of thetest frame
functionality).The text was updated successfully, but these errors were encountered: