Guido Imperiale

The whole distinction between UTF-8 strings and bytes objects isn't there either. str.encode() crashes.

@ev-br , 1. You can see the ripped-out t logic here, and it's very much not trivial! https://github.com/crusaderky/xarray_extras/blob/master/xarray_extras/kernels/interpolate.py#L11-L70 2. make_interp_spline is well designed for pure numpy users. What I'm suggesting...

I haven't looked at this problem for 4 years, so my memory is failing me. Private functions are better than having to reimplement everything from zero. I think they're missing...

I'm afraid I stopped working on this a long time ago; xarray_extras as a whole is in long term maintenance only from my side. Still I'll be happy to update...

I actually didn't work on interp. What I did is http://xarray-extras.readthedocs.io/en/latest/api/interpolate.html, which is based on scipy BSpline. The key feature I needed was to work with an n-dimensional y ;...

As discussed before, I am against splitting the class in two. I simply see no benefit. Could you explain these cases with pandas that you believe can't fit in the...

We should remain aligned to the decision of the core numpy developers where they chose to use the same class for scalar and vectorized ndarrays. Pandas should be changed to...

> This is still an issue. I noticed that the documentation of `map_blocks` states: **kwargs** ([mapping](https://docs.python.org/3/glossary.html#term-mapping)) – Passed verbatim to func after unpacking. xarray objects, if any, will not be...

Having said the above, your design is... contrived. There isn't, as of today, a straightforward way to scatter a local dask collection (`persist()` will push the whole thing through the...

```python new_data_future = xr.apply_ufunc( _copy_test, data, a_x, ... ) ``` *instead* of using kwargs. I've opened https://github.com/dask/distributed/issues/7140 to simplify this. With it implemented, my snippet ```python test = np.full((20,), 30)...