Color Cycle Survey

Dear all,

For the past year or so, I've been working on techniques for generating
visually pleasing colorblind-friendly color cycles, since I find the
current default difficult to interpret (see [1] for details). The
colorblind-friendly aspect can be solved through technical means, by
enforcing minimum perceptual distances in combination with color vision
deficiency simulations, but the aesthetic preference aspect is subjective
and requires a different approach. To this end, I've created a survey for
collecting the data necessary for creating a model of general aesthetic
preference for color cycles:

https://colorcyclesurvey.mpetroff.net/

As the usefulness of this approach hinges on collecting enough data [2],
I'd appreciate folks taking the survey. Feedback on the survey or the
general approach taken would also be appreciated (there's an issue tracker
[3]). Thanks.

-Matthew Petroff

[1] https://github.com/matplotlib/matplotlib/issues/9460
[2] What constitutes "enough" is unfortunately difficult to define a priori.
[3] https://github.com/mpetroff/color-cycle-survey/issues
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-users/attachments/20181215/36c5f227/attachment.html>

Thanks for this work, Matthew! I'm excited about this. For what it's worth, though (probably not much), in the 11 tries I had, I don't think I liked any of the cycles as much as the one in this comment [1], which I think is close to perfect:

https://github.com/matplotlib/matplotlib/issues/9460#issuecomment-377312565

But I am not colorblind nor have I worked in this space so I defer to the data. =)

Thanks again,

Juan.

···

On Sun, Dec 16, 2018, at 8:51 AM, Matthew Petroff wrote:

Dear all,

For the past year or so, I've been working on techniques for generating visually pleasing colorblind-friendly color cycles, since I find the current default difficult to interpret (see [1] for details). The colorblind-friendly aspect can be solved through technical means, by enforcing minimum perceptual distances in combination with color vision deficiency simulations, but the aesthetic preference aspect is subjective and requires a different approach. To this end, I've created a survey for collecting the data necessary for creating a model of general aesthetic preference for color cycles:

https://colorcyclesurvey.mpetroff.net/

As the usefulness of this approach hinges on collecting enough data [2], I'd appreciate folks taking the survey. Feedback on the survey or the general approach taken would also be appreciated (there's an issue tracker [3]). Thanks.

-Matthew Petroff

[1] https://github.com/matplotlib/matplotlib/issues/9460
[2] What constitutes "enough" is unfortunately difficult to define a priori.
[3] https://github.com/mpetroff/color-cycle-survey/issues
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users at python.org
https://mail.python.org/mailman/listinfo/matplotlib-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-users/attachments/20181217/e47d0572/attachment.html>

I spent something like two minutes creating that color cycle for the sole
purpose of creating the screenshot and don't consider myself much of a
designer, so I'm glad you like it.

The color vision deficiency simulation strength used for that cycle (50%)
was less than what was used to generate the sets used in the survey (100%),
so that may be a contributing factor (a minimum perceptual distance of 18
was used for both). However, I think it's mostly just that of all the
possible color cycles, many more look bad than good. If all goes well, the
survey data can be used to create a model of what looks good, although it's
still an experiment, so there are no guarantees.

-Matthew Petroff

···

On Mon, Dec 17, 2018 at 1:34 AM Juan Nunez-Iglesias <jni.soma at gmail.com> wrote:

Thanks for this work, Matthew! I'm excited about this. For what it's
worth, though (probably not much), in the 11 tries I had, I don't think I
liked any of the cycles as much as the one in this comment [1], which I
think is close to perfect:

https://github.com/matplotlib/matplotlib/issues/9460#issuecomment-377312565

But I am not colorblind nor have I worked in this space so I defer to the
data. =)

Thanks again,

Juan.

On Sun, Dec 16, 2018, at 8:51 AM, Matthew Petroff wrote:

Dear all,

For the past year or so, I've been working on techniques for generating
visually pleasing colorblind-friendly color cycles, since I find the
current default difficult to interpret (see [1] for details). The
colorblind-friendly aspect can be solved through technical means, by
enforcing minimum perceptual distances in combination with color vision
deficiency simulations, but the aesthetic preference aspect is subjective
and requires a different approach. To this end, I've created a survey for
collecting the data necessary for creating a model of general aesthetic
preference for color cycles:

https://colorcyclesurvey.mpetroff.net/

As the usefulness of this approach hinges on collecting enough data [2],
I'd appreciate folks taking the survey. Feedback on the survey or the
general approach taken would also be appreciated (there's an issue tracker
[3]). Thanks.

-Matthew Petroff

[1] https://github.com/matplotlib/matplotlib/issues/9460
[2] What constitutes "enough" is unfortunately difficult to define a
priori.
[3] https://github.com/mpetroff/color-cycle-survey/issues
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users at python.org
https://mail.python.org/mailman/listinfo/matplotlib-users

_______________________________________________
Matplotlib-users mailing list
Matplotlib-users at python.org
https://mail.python.org/mailman/listinfo/matplotlib-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/matplotlib-users/attachments/20181217/1b5992b0/attachment.html>