Example of new colormaps is misleading

[8one6]

Hooray for the new colormaps!

Boo for examples with the new colormaps that obfuscate their nature!

The example here (http://matplotlib.org/1.5.0/users/whats_new.html#new-colormaps) is attempting to show two things at once: first the new colormaps themselves and second the ability to use the new cycler functionality to change things from one place to another.

I think this is very misleading. I think it makes much more sense to show the four new colormaps by plotting the same data four times, with only the colormaps changing from one axes to the next.

Thoughts?

[tacaswell]

Fair point, but I think a bit strongly worded. I don't think anyone would look at those images and think they were the same data so using 'misleading' in the issue title is misleading as it were.

I would not be opposed to a PR that switched that data out for the 'standard' test pattern we use for all of the contour examples.

I am in favor of closing this issue independent of a future PR to change the example, but given that I wrote that example I will leave that for someone else.

[efiring]

Agreed. I will close this, but a simple PR as @tacaswell suggests is welcome.