Skip to content

Thresholding and rebuild updates. #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
apluff wants to merge 59 commits into
base: master
Choose a base branch
from
Open

Thresholding and rebuild updates. #2

apluff wants to merge 59 commits into from

Conversation

@apluff
Copy link

@apluff apluff commented Jul 30, 2025

Updates seas.ica.rebuild()

  • New options as described in docstrings, of note:
  • New options for 'filter_method', incl. 'constant'
  • New option 'apply_component_filter', which applies a butterworth lowpass filter to IC timecourses before rebuilding video. lpf threshold determined using 'chigh' option.
  • New option 'apply_masked_mean' for use with thresholded ICs.

New function seas.domains.threshold_by_domains()

  • Optional function to spatially threshold ICs in various ways. Use alongside 'apply_masked_mean' above to selectively rebuild IC videos. Docstrings cover options.

apluff added 30 commits July 21, 2025 16:00
@apluff
Added draft percentile filter in threshold_by_domains.
c577223
@apluff
Debug flipped assignment.
33868fd
@apluff
Added mask case to thresh_by_type to yield binary activity masks.
08959ac
@apluff
Merge branch 'master' of https://github.com/apluff/pySEAS
d6f1688
@apluff
Added alternative mean addition routine, incl. option and masks impor…
8d57966 
…t from dict.
@apluff
Changed masks output to dictionary rather than encoding in eig_vec.
8c1ac66
@apluff
Broadcasting debug on apply_masked_mean.
26ad526
@apluff
Add temp variable for eig_vec to do blurring for threshold (rather th…
cd7d913 
…an data).
@apluff
Added small mask ROI filtering before blurring tp threshold_by_domains.
c29e041
@apluff
Debug eigenbrain mask dtype.
1ce6a14
@apluff
Debug dtype changes between morph filtering and blurring.
c23ba85
@apluff
Added draft of IC timecourse lpf as part of threshold_by_domains.
a6c7e8f
@apluff
Debug blurring.
3c4ce61
@apluff
Debug timecourse indexing.
a22b796
@apluff
Commented out lpf code from threshold_by_domains, need to rewrite.
0c37375
@apluff
Debug indexing issue.
0d2b19a
@apluff
Revert indexing issue and debug blurring dtype.
2658d3f
@apluff
Debug dtype assignment.
8c8c9aa
@apluff
Try filtering routine without blurring (which seems to be blanking wh…
1841894 
…ole video).
@apluff
Add option apply_component_filter to ica.rebuild and domains.threshol…
f860922 
…d_by_domains for lpf on IC timecourses.
@apluff
Make default component lpf 0.5Hz.
592ae07
@apluff
Debug masks assignment.
a0e246f
@apluff
Debug and readd combined morph filtering + blur for threshold_by_doma…
fdaa51f 
…ins.
@apluff
Added cfilt and mfilt parameter passthrough for seas.ica.rebuild()
b2b7ad5
@apluff
Testing constant as filter_mean method.
6a438a0
@apluff
Added z-score threshold to threshold_by_domains.
dc9f00e
@apluff
Debug z-score for threshold_by_domains
d52373b
@apluff
Cleaned and consolidated apply_component_filter option for rebuilding…
3ab58bf 
…. Consolidated new options in docstrings.
@apluff
Consolidated docstrings for threshold_by_domains, and removed compone…
0ade191 
…nt filtering (now in seas.ica.rebuild).
@apluff
Refactor filtering ICs into its own function seas.ica.filter_componen…
8b5113e 
…ts().
apluff added 29 commits August 7, 2025 11:42
@apluff
dtype debug
43af207
@apluff
Debug eig_mix assignment order.
86daec8
@apluff
Merge branch 'master' of https://github.com/apluff/pySEAS
a0bc538
@apluff
Update all instances of np.NAN to np.nan for numpy 2.x compatability.
3a65829
@apluff
Debug include_noise logic in ica.rebuild()
2ae6341
@apluff
Debug use of thresh_masks in ica rebuild.
9743982
@apluff
Added method to threshold component timeseries' and associated option…
2faadf8 
… in rebuild.
@apluff
Added argument ica.project to optionally retain excess noise.
c3e956b
@apluff
Debug calc_residuals fork in ica.project.
93a38bd
@apluff
Replace missing artifact_components argument in calc_residuals fork.
190114c
@apluff
Add new functions to export event mask compilation tifs and threshold…
0cc2a40 
…ed video.
@apluff
Added experiment.sort_components() to handle unsorted outputs in Balb…
cbd2936 
…i data. Added experiment.flip_negative_components() to convert older dictionaries to all-positive. Corrected deprecated flip adjustment in seas.gui.
@apluff
Replace np.bool with bool for new numpy compatability.
97388e7
@apluff
Added extra thresholding step to ica.threshold_by_domains as 'schemat…
8fb4703 
…ic' mode and implementation in experiment.export_event_masks. This is to give poster activity maps a more contracted/schematic-like appearance.
@apluff
Debug schematic thresholding.
0debc34
@apluff
Further debug of schematic option.
06f6608
@apluff
Further debug on schematic option.
009e11f
@apluff
Further debug of schematic option.
3abf1df
@apluff
Added schematic fork for ica.threshold_by_domains.
63ec4c1
@apluff
Debug schematic fork.
74b0840
@apluff
Further debug on schematic fork.
5383a36
@apluff
Further debug schematic fork.
ec15638
@apluff
Further debug schematic fork.
eaec6f4
@apluff
Further debug schematic fork.
a737a68
@apluff
Brittle but working schematic fork. Use only blur=1 to avoid labellin…
35c9c69 
…g padded boundary from blur operation.
@apluff
Refactor and update experiment.sort_components to use lag1 by default.
f3ce117
@apluff
Refine sort_components for Balbi data. Sort artifact_components and d…
624651e 
…on't run domain_map.
@apluff
Make sorting of artifact_components conditional to generalise to newl…
c738e41 
…y decomposed pyseas dicts.
@apluff
Tidying sort_components.
027b49f
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@apluff