34902152587
60%

Ran 14 Sep 2026 10:07PM UTC

Jobs 1

Files 41

Run time 1min

Badge

Embed ▾

Committed 14 Sep 2026 10:03PM UTC coverage: 60.133% (+0.07%) from 60.067%

Build # 34902152587

Build Type

push

github

Committed by

web-flow

Commit Message

Support per-feature-group fac in TV denoising (#112)

* Support per-feature-group fac in TV denoising

EphysDenoiser/denoise_dataframe/denoise_raw_features_data accept fac as
a dict keyed by raw_ap/raw_lf/raw_lf_csd/waveforms instead of a single
scalar, so each feature group can use its own TV-denoise weight. Groups
absent from the dict keep the previous default of 1. Scalar fac still
works unchanged.

* Set the bracketed per-group fac as the default

DEFAULT_FAC = {raw_ap: 0.1, raw_lf: 0.1, raw_lf_csd: 0.1, waveforms: 5},
selected by the bracketing sweep and confirmed with 5-fold CV, replaces
the old scalar fac=1 default in EphysDenoiser/denoise_dataframe/
denoise_raw_features_data. test_denoise_features now pins fac=1
explicitly since its expected values were computed under the old
default.

Coverage Stats

2703 of 4495 relevant lines covered (60.13%)

0.6 hits per line

Coverage Regressions

Lines	Coverage	∆	File
43	88.44	0.04%	features.py
3	93.1	0.0%	aggregation.py

Jobs

ID	Job ID	Ran	Files	Coverage
1	34902152587.1	14 Sep 2026 10:07PM UTC	41	60.13	GitHub Action Run

int-brain-lab / ibleatools / 34902152587
60%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Coverage Regressions

Jobs

Source Files on build 34902152587

int-brain-lab / ibleatools / 34902152587 60%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Coverage Regressions

Jobs

Source Files on build 34902152587

int-brain-lab / ibleatools / 34902152587
60%

README BADGES
x