issue_comments: 264742277

This data as json

html_url issue_url id node_id user created_at updated_at author_association body reactions performed_via_github_app issue
https://github.com/pydata/xarray/issues/1084#issuecomment-264742277 https://api.github.com/repos/pydata/xarray/issues/1084 264742277 MDEyOklzc3VlQ29tbWVudDI2NDc0MjI3Nw== 6628425 2016-12-04T23:58:12Z 2016-12-04T23:58:12Z MEMBER

@shoyer @jhamman this is my first time digging in to the internals of pandas / xarray, so please forgive me if this is off-track, but I've started experimenting with a subclass of pandas.Index (which I've called NetCDFTimeIndex) that could be a start toward addressing this issue. Here's a link to a Jupyter notebook that sketches it out, and includes a few examples of its use.

As a proof of concept, it currently enables groupby operations through a crude version of the _field_accessor / get_date_field approach implemented in pandas.DatetimeIndex (which in pandas seems to leverage some cython). Any datetime object implemented in netcdftime can be used (so this supports a number of custom calendars out of the box). By overriding get_loc and slice_locs and leveraging an internal function of netcdftime, I've also experimented with enabling ISO 8601 string-based selection of times / time slices.

I have yet to look into serialization or how / if this could be integrated into xarray, but does this seem like an approach worth pursuing to address this issue?

cc @spencerahill

 
{
    "total_count": 1,
    "+1": 1,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   187591179