tighten up prose and code#1
Conversation
| # As phase-coherence variable is a continuous variable, we'll need to | ||
| # split the values of face coherence to percentiles to match roughly their | ||
| # original scale (i.e, 0 to 100 % signal-to-noise ratio in 10% steps). | ||
| # **Note:** We'll only do this for visualization purposes, |
There was a problem hiding this comment.
What is the rationale for doing this? The data has 18 levels of standardized phase-coherence, and you're forcing it into 11 bins.
It's also somewhat distracting from the main point of the tutorial, which is regression.
There was a problem hiding this comment.
So as described in the main PR, it's actually 18 levels of phase-coherence, (0 to 85% with increments of 5%) not 11. This is because the data was (probably) taken from a sub-sample of the original paper, which (for whatever reasons) saw a slightly different version of the paradigm. Sorry that I didn't noticed that earlier on. I just assumed everybody saw the version of the paradigm which was cited on the original paper.
There was a problem hiding this comment.
Also I though, I might be useful for people to see what the effect of phase-coherence look like, at least on a descriptive level, and thus know what to expect from the linear regression analysis.
* add info / refs on limo dataset add sklearn regression add class linear regression update limo doc fix limo fetcher docstring fix merging issues * FIX: Indentation * FIX: A couple more * STY: Nitpicks [ci skip] * ENH: Links * Apply suggestions from code review add jonas suggestions Co-Authored-By: jona-sassenhagen <jona.sassenhagen@gmail.com> * fix syntax + remaining issues * tighten up prose and code (#1) * drop class approach / add LIMO_PATH to config * Apply suggestions from code review Co-Authored-By: Daniel McCloy <dan.mccloy@gmail.com>
No description provided.