search

ME-ICA / tedana

TE-dependent analysis of multi-echo fMRI
https://tedana.readthedocs.io
GNU Lesser General Public License v2.1
158 stars 94 forks source link

Include intercept in `get_coeffs` calls #1089

Open tsalo opened 2 months ago

tsalo commented 2 months ago

Closes none. I was just thinking that we should include an intercept term when calculating echo-wise betas.

Changes proposed in this pull request:

-

codecov[bot] commented 2 months ago

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 89.79%. Comparing base (af5e99a) to head (04278f1).

Additional details and impacted files ```diff @@ Coverage Diff @@ ## main #1089 +/- ## ========================================== - Coverage 89.80% 89.79% -0.01% ========================================== Files 26 26 Lines 3540 3537 -3 Branches 621 621 ========================================== - Hits 3179 3176 -3 Misses 212 212 Partials 149 149 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

tsalo commented 2 months ago

Weirdly, including the intercept didn't change anything. I'm not sure why. All of the other components should have mean values of zero, right? So how could raw echo-wise time series (i.e., with different scales) have the same OLS fit with and without an intercept term?