home / github / issue_comments

Menu
  • Search all tables
  • GraphQL API

issue_comments: 254676127

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/158#issuecomment-254676127 https://api.github.com/repos/pydata/xarray/issues/158 254676127 MDEyOklzc3VlQ29tbWVudDI1NDY3NjEyNw== 12852539 2016-10-19T00:18:53Z 2016-10-19T00:25:38Z NONE

Why won't this be fixed?

I think some clarification in the documentation would be useful. Currently they say:

xarray supports “group by” operations with the same API as pandas [and that the required parameter for Dataset/DataArray.groupby is an] Array whose unique values should be used to group this array.

However, pandas supports grouping by a function or by any array (e.g. it can be a pandas object or a numpy array). The xarray API is narrower than pandas, and has an undocumented requirement of a named DataArray (contrasting xarray behaviour of creating default names like "dim_0" elsewhere).

``` python import numpy as np data = np.arange(10) + 10 # test data f = lambda x: np.floor_divide(x,2) # grouping key

import pandas as pd for key in f, f(data), pd.Series(f(data)): print pd.Series(data).groupby(key).mean().values print pd.DataFrame({'thing':data}).groupby(key).mean().thing.values

these pandas examples are all equivalent

import xarray as xr da = xr.DataArray(data) key = xr.DataArray(f(data)) key2 = xr.DataArray(f(data), name='key') print da.groupby(key2).mean().values # this line works print da.groupby(key).mean().values # broken: ValueError: group must have a name ```

{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
  35682274
Powered by Datasette · Queries took 316.181ms · About: xarray-datasette