tess

I just wanted to note at some point in the future implementing this a dictionary/dispatch table might be worthwhile. Something like ```python attach_dispatch = { pdarray.objType.lower(): create_pdarray, Strings.objType.lower(): Strings.from_return_msg, Datetime.special_objType.lower():...

`read_hdf` for dict has different column order for > 10 rows

2

I'm not sure if this matters but `read_hdf` for dict has different column order for `> 10` rows. I found this while working #2602 because `pd.testing.assert_frame_equal` failed ```python >>> df_dict...

One issue that I think is a problem with all the registration code is that when an object is registered the `user_defined_name` is set as `self.name` If the object is...

should we add `mink` and `maxk` aggregations to groupby

1

should we add `mink` and `maxk` aggregations to groupby, which return a segarray. Most of the logic needed for this was implemented for `find` to return all occurrences in issue...

Fixes #3001: nan broadcast bug

1

This PR fixes #3001 by using transmute to interpret the floats as `uint64`s before broadcasting and transmuting back afterwards

Closes #3009: `indexof1d` to handle null values

3

This PR (closes #3009) refactors `indexof1d` to use `find` since they have similar functionality and `find` is fairly optimized and correctly handles null values (once we `dropna=False` to the `Groupby`)....

This PR fixes #3156, `Categorical.sort` bug and adds testing

Add `normal` to our generators, we already have a `fillRandom` helper in `Random.chpl` so this shouldn't be a big lift

Our IO functions used to return a dictionary if multiple columns or datasets were provided and if only was provided to just return the object. It has been requested by...

Users have requested we have `read_parquet` always return a dictionary. Right now passing it a single column will return that column type (for example pdarray) but it would be nice...