Coordinates operations

[keewis]

What is your issue?

At the moment, the Coordinates objects are not as useful as I'd like them to be, so in this issue I'd like to collect all the methods it should support. Some of these already exist, but listing them again helps with defining a coherent API:

• coordinate variables
  • assignment of new coordinates (from Variable / DataArray / Coordinates / any array): Coordinates.assign()
  • deletion of coordinates (similar to Dataset.drop_vars): only via del coords["x"] but a Coordinates.drop_vars() might be welcome I guess
  • renaming coordinates: no specific method yet, Coordinates.rename_vars()?
  • renaming dimensions on coordinates: no specific method yet, Coordinates.rename_dims()?
• indexes
  • assignment of index instances: could use Coordinates.assign(Coordinates.from_xindex(idx_obj)) but one index at a time
  • dropping indexes: Coordinates.drop_indexes() (Add Coordinates set_xindex() and drop_indexes() methods #8118)
  • creation of an index based on variables, a given class, and additional options: Coordinates.set_xindex() (Add Coordinates set_xindex() and drop_indexes() methods #8118)
• combination
  • merging (also using |)

cc @benbovy, @scottyhq

[keewis]

Some examples of how I'd imagine to use these:

Coordinates:

# assigning variables
time = xr.DataArray(..., dims="time")
coords.assign(lat=xr.Variable("lat", lat), lon=("lon", lon), time=time)
# dropping variables (and their associated indexes)
coords.drop_coords(["lat", "lon"])
# renaming variables
coords.rename_coords({"lat": "latitude", "lon": "longitude"})
# renaming dims
coords.rename_dims({"lat": "latitude"})

Indexes:

# assign an existing index
coords.assign_indexes([index1, index2])
# drop indexes (without dropping the associated variables)
coords.drop_indexes([Frozen(["lon", "lat"]), "time"])
# create new index
coords.create_indexes({Frozen(["lon", "lat"]): (CustomIndex, {"option1": "value1", ...})})

Combination:

# merging two `Coordinates` objects
coords1.merge(coords2)
# or
coords1 | coords2