Optimize some copying

[headtr1ck]

• Potentially closesxarray 2022.10.0 much slower then 2022.6.0 #7181
• Tests added
• User visible changes (including notable bug fixes) are documented in whats-new.rst

I have passed along some more memo dicts, which could prevent some double deep-copying of the same data (don't know how exactly, but who knows :P)
Also, I have found some copy calls that did not pass along the deep argument (I am not sure if that breaks things, lets find out).
And finally I have found some places where shallow copies are enough.

All together it should improve the performance a lot when copying things around.

[dcherian]

Thanks @headtr1ck do we have a benchmark for this, if not can we add one please?

[headtr1ck]

Thanks @headtr1ck do we have a benchmark for this, if not can we add one please?

Since the benchmark didn't change we either don't have one or my change doesn't matter much, haha.

I think the most important change is the shallow copy for Variable.to_index_variable, but I will have to check how to test this in a useful way, probably better testing some functions that use it indirectly.

[dschwoerer]

The change does matter - but deep copies are still much more expensive than they used to be (as to be expected, I guess)

[headtr1ck]

The change does matter - but deep copies are still much more expensive than they used to be (as to be expected, I guess)

Do you by any chance know which parts have improved, so we can add them as a benchmark here?

[headtr1ck]

I added a benchmark for swap_dims which uses Variable.to_index_variable when swapping into an existing coord, but I can not see any significant improvement...

Any ideas what else to test?
Maybe Indexes.copy_indexes, but I have not found a more high-level method that can take advantage of the memo dict...

[benbovy]

The change in Variable.to_index_variable seems sensible (not sure when one wants a deep copy of an IndexVariable or an Xarray / Pandas index).

to_index_variable may be called in some core functions of Xarray internals (e.g., in as_variable()) so it might be tricky to benchmark its effect Xarray-wise. Perhaps it would be good to track it down in the original issue #7181?

[headtr1ck]

Are we merging this anyway, or should we try harder to find a benchmark that shows some improvement?

[dcherian]

I don't think we need a benchmark to merge. Sorry that wasn't clear, it mostly for information purposes.

The change in Variable.to_index_variable seems sensible (not sure when one wants a deep copy of an IndexVariable or an Xarray / Pandas index).

Great! let's merge