ENH: Add option to smooth to nearest

[larsoner]

Allows doing smoothing_steps='nearest' so you can see the patches of cortex like (not the most relevant example but gives the idea):

[mwaskom]

Maybe a better example use-case would be doing some kind of discrete labeling in low-res space and then propagating the labels to the high-res surface?

[larsoner]

some kind of discrete labeling in low-res space and then propagating the labels to the high-res surface?

That's basically what a MNE inverse solution is, a low(er)-res subsampling of the high res surface and then when we plot we upsample.

Ultimately this might not even need an example, so I can revert it if need be. I'd rather not write a separate new one just for this.

[codecov[bot]]

Codecov Report

Merging #275 into master will increase coverage by 0.35%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #275      +/-   ##
==========================================
+ Coverage   73.96%   74.31%   +0.35%     
==========================================
  Files           7        7              
  Lines        2535     2570      +35     
  Branches      511      514       +3     
==========================================
+ Hits         1875     1910      +35     
  Misses        482      482              
  Partials      178      178

[larsoner]

Got the little indexing demons worked out, this is what the MEG one looks like with 'nearest'

https://281-1551431-gh.circle-artifacts.com/0/html/auto_examples/plot_meg_inverse_solution.html

[mwaskom]

That's right — just the changes file that is usually the basis for a release announcement email.

[larsoner]

That file only goes up to 0.8, and we've released 0.9 already. I guess we're a bit behind there. I can add a 0.9 and dev section here sometime today or tomorrow

[larsoner]

Feast your eyes on the beautiful changelog:

https://282-1551431-gh.circle-artifacts.com/0/html/changes.html

[agramfort]

I have rarely been so pleased...