create a mi

In [13]: mi = pd.MultiIndex.from_product([[1,2,3], list('abc'), pd.date_range('20200101', periods=2, tz='UTC')], names=['int', 'string', 'dt'])                            

You can already get .dtypes, but is slightly cumbersome. I would propose adding Multidex.dtypes (we already have MultiIndex.dtype but its always object). I think this is worth the convenience api.

Like xarray, pandas supports attaching arbitrary metadata to DataFrames and persisting it across operations. https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.attrs.html

Dask could pretty easily implement this as well. We'd have a

• _Frame.attrs property. This would likely return self._meta.attrs.
• We'd verify that dd.from_pandas(data) correctly extracts attrs from `da

Improve readability of thread id based branches by giving them more descriptive names.

e.g.

if (!t) // is actually a t == 0

and

https://github.com/rapidsai/cudf/blob/57ef76927373d7260b6a0eda781e59a4c563d36e/cpp/src/io/statistics/column_stats.cu#L285
Is actually a lane_id == 0
As demonstrated in rapidsai/cudf#6241 (comment), pr

Support Series.median()

What happened:
xr.DataArray([1], coords=[('onecoord', [2])]).sel(onecoord=2).to_dataframe(name='name') raise an exception ValueError: no valid index for a 0-dimensional object

What you expected to happen:

the same behavior as: xr.DataArray([1], coords=[('onecoord', [2])]).to_dataframe(name='name')

Anything else we need to know?:

I see that the array after the select