Some simple broadcast_dim method?

[Hoeze]

I've already found xr.broadcast(arrays).
However, I'd like to just add a new dimension with a specific size to one DataArray.
I could not find any simple option to do this.

If there is no such option:

• add a size parameter to DataArray.expand_dims?
• DataArray.broadcast_dims({"a": M, "b": N})?

[fmaussion]

Is exand dims not what you are looking for?

[Hoeze]

No, I'd need something like np.tile.
expand_dims inserts only a dimension of length '1'

[Hoeze]

As an explanation:
I'd like to change my program to only use lazy / chunked calculations in order to save RAM.

I recognized that np.broadcast_to converts the DataArray into a numpy one.
Therefore I needed some xarray way to solve this.

I tried:

DataArray.expand_dims("new_dim").isel("new_dim", np.repeat(0, target_dim_size))

but this really looks ugly and I'm not sure about the performance implications of this.

[shoyer]

We could add an optional size argument to expand_dims, with a default value of 1.

[Hoeze]

@shoyer so there is no direct xarray equivalent to np.broadcast_to?

[shoyer]

We have xarray.Variable.set_dims, but that's only in the low-level variable API.

[Hoeze]

Maybe related:

Consider the following example to calculate pairwise distances:

x = np.array([[1,2,3,4]])
dist = x.T - x

numpy automatically broadcasts the one-dimensions to get 4x4 matrices and substracts them.

As far as I can see, this example is really hard to recreate with xarray, since there is nearly no possibility to add a new dimension to x and broadcast it properly.

[shoyer]

let's continue the repeated dimension discussion over in #1378

[dcherian]

This should be fixed by the latest expand_dims