Decide what to do with eddy-related functionality in Calc

[spencerahill]

2.5 years ago I implemented support for "eddy" fields, meaning the anomalies of a field with respect to it's monthly mean. And likewise for the zonally asymmetric component of fields, i.e. the difference between a field's full value and its zonal mean value. All of the related code is in Calc.

This code is quite convoluted and hackish. It also is not even documented at all in our official docs, nor can I remember the last time I used it...probably not since right after I implemented it! It's also now complicating our efforts to switch to the newly available CFTimeIndex (see #273 (comment)). So we should decide what to do with it.

In my opinion, the best course of action would be to simply strip it all out for now and then re-implement it in a non-cringe-worthy way if/when the use case arises. But if others (@spencerkclark, @chuaxr) are using it, we could leave it for now.

[chuaxr]

I don't use the eddy feature, but thanks for the heads-up!

[spencerahill]

Thanks @chuaxr.

So the plan moving forward is to just remove this stuff for now and re-implement it later if/when it arises as a compelling feature to have.

That being said, I would say removing this is neither a high priority nor exactly straightforward to do, since the relevant code is so convoluted (mea culpa). Could be wrong on the second part though.