search

LIMO-EEG-Toolbox / limo_tools

Hierarchical Linear Modelling for MEEG data
https://limo-eeg-toolbox.github.io/limo_meeg/
Other
58 stars 27 forks source link

2% trimmed means? #68

Closed JohnAtl closed 3 years ago

JohnAtl commented 3 years ago

Hello, It seems in limo_robust_rep_anova, the argument '2' to limo_trimmed_mean() would yield 2% trimmed means, rather than the customary 20%. Is this correct? Thanks

https://github.com/LIMO-EEG-Toolbox/limo_tools/blob/bba57e3ac74380720188d8fc0f3df1a6f18f578f/limo_robust_rep_anova.m#L124

CPernet commented 3 years ago

oops yes, indeed thx for spotting it please be very careful with this function, it's not validated yet (you can see it not called by any other functions, not is in the wiki) -- one factor works that's pretty much it

CPernet commented 3 years ago

pushed the changes in both master and hotfixes, thx

JohnAtl commented 3 years ago

Hi Cyril,

limo_trimmed_mean is called from limo_robust_rep_anova, which is what I was really interested in. (Although I can't figure out the structure for Data, since it is time x subject x measures, but represents more than one condition. Any suggestions?)

Has limo_rep_anova not been validated?

Do you have suggestions for robust anova and post-hoc tests on related/dependent data? (It doesn't fit in study since each sample is taken at a different time point, depending on subjects' movement.)

Struggling grad student with non-normal EEG data would be most grateful for suggestions.

Thanks, John

On May 2, 2021, at 1:06 AM, Cyril Pernet @.***> wrote:

oops yes, indeed thx for spotting it please be very careful with this function, it's not validated yet (you can see it not called by any other functions, not is in the wiki) -- one factor works that's pretty much it

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/LIMO-EEG-Toolbox/limo_tools/issues/68#issuecomment-830741511, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABLUQCVC4HMCZ4LHZX5UUQ3TLTMT7ANCNFSM434U6CPQ.