RFC: use MPL cycler colors / "dark mode" [ci skip]#8217
RFC: use MPL cycler colors / "dark mode" [ci skip]#8217drammock wants to merge 1 commit intomne-tools:mainfrom
Conversation
|
I'd be OK with a manually specified set of dark colors. The traces look pretty good in your screenshot - are you planning on making the rest of the window dark as well? Also, EEG is normally black so I guess this will become white? |
Yes. I have that working locally
Yes. |
|
@drammock if you want to come back to this, we converged on a usable set of colors in mne-qt-browser / mne-tools/mne-qt-browser#108 . I've been using them for a while now with no (big) issues locally at least |
I see this as very low priority. To me it's only worth it if we make a global setting that applies to all of our 2D plots (similar to |
There's been some talk about making a "dark mode" for our 2D plotting. I did a bit of investigating with
raw.plot(), here's what I found:mne/defaults.py(which gets loaded when the package is imported), usingplt.style.use('dark_mode')orwith plt.style.context('dark_mode'):don't work as expected unless theimport mnestatement comes after the style invocation.C2), what is green in normal mode is lavender in dark mode. This is... not great.See below for what is the best compromise I was able to come up with so far while using the default cycler colors. Obviously this is an incomplete attempt, just wanted to show the blue-and-teal-on-black look:
alternatives
we could roll-our-own dark mode by adding a couple keys to the default color dict (i.e., "foreground" and "background"), and then having a separate pre-defined dict for a dark-mode version of the default colors.
cc @cbrnr, who has expressed interest in this.