Choose a better default colormap in plot_traj.

[danielballan]

Thanks to Kristen Thyng's talk at SciPy 2014 (not yet online) I now know that cm.winter is the very worst colormap in matplotlib, perceptually speaking. It should not be the default for coloring-by-time in trackpy.plots.plot_traj.

[danielballan]

@RebeccaWPerry, any interest in tackling this? Should be a one-line fix once the best alternative colormap is chosen.

[RebeccaWPerry]

Sure-- have you found her slides online? Someone asked her to commit it as documentation to matplotlib. Or maybe her video will be up soon?

[danielballan]

I'm hoping for a video too. Meanwhile, keep an eye on matplotlib/matplotlib#3223. The world needs your slides, @kthyng. :- D

[nkeim]

May I just say a word in defense of cm.winter? I've found it to be the best colormap in matplotlib when color conveys the least important information in the figure. But plot_traj is not such a case.

Just wanted to make sure no colormaps have their feelings hurt today.

[dmcdougall]

Videos are here and here.

[dmcdougall]

Slides are here and here.

[danielballan]

Thanks, @dmcdougall.

[kthyng]

Oh and I'm working on writing up a "colormaps guide" for that matplotlib
issue! Things are moving along... slowly. :)

On Wed, Jul 16, 2014 at 9:59 PM, Dan Allan notifications@github.com wrote:

Thanks, @dmcdougall https://github.com/dmcdougall.

—
Reply to this email directly or view it on GitHub
#124 (comment).

Kristen M. Thyng
Postdoctoral Research Associate
Department of Oceanography
Texas A&M University
http://kristenthyng.com

[RebeccaWPerry]

I hope to look at colormaps for this specific plot_traj case this weekend... Sunday maybe.

[RebeccaWPerry]

Dan, is your scipy2014 presentation notebook posted somewhere? I'd like to test colormaps using it.

[RebeccaWPerry]

So plot_traj optionally colors by frame or by particle. I think we might want different colormaps for these different applications. When coloring by frame, I like the idea of a monochrome colormap that appears to brighten as time progresses. For coloring by particle, I would like lots of very different colors. Thoughts?

[danielballan]

I haven't put my presentation notebook online because of its weird dependencies (a PR branch of IPython, mpld3 master, trackpy master). But it is basically a stripped down version of the walkthrough notebook

[danielballan]

I agree. The current behavior is the winter colormap for color-by-frame and the default color cycle (however it is set up) for color-by-particle.

[tacaswell]

One thing I found useful was a combination of the two, use the colored monochrome color maps with time being what sets the color and then cycling through the color maps on a per-particle basis.