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/386#issuecomment-91679470,https://api.github.com/repos/pydata/xarray/issues/386,91679470,MDEyOklzc3VlQ29tbWVudDkxNjc5NDcw,1217238,2015-04-10T20:37:43Z,2015-04-10T20:37:43Z,MEMBER,"Resampling actually supports custom aggregation methods with the `how` parameter: http://xray.readthedocs.org/en/stable/generated/xray.Dataset.resample.html#xray.Dataset.resample
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,67332234
https://github.com/pydata/xarray/issues/386#issuecomment-91495357,https://api.github.com/repos/pydata/xarray/issues/386,91495357,MDEyOklzc3VlQ29tbWVudDkxNDk1MzU3,10194086,2015-04-10T09:40:36Z,2015-04-10T09:40:36Z,MEMBER,"Thanks, it works. I didn't think of this.

However, for anything else than the mean this would not work? So I would opt for the second option.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,67332234
https://github.com/pydata/xarray/issues/386#issuecomment-91355259,https://api.github.com/repos/pydata/xarray/issues/386,91355259,MDEyOklzc3VlQ29tbWVudDkxMzU1MjU5,1217238,2015-04-09T21:01:04Z,2015-04-09T21:01:04Z,MEMBER,"The better solution here is to use `resample` rather than `groupby('time.date')`:

```
ts_mean = ts.resample('1D', dim='time')
```

`time.date` and `time.time` (with is currently broken) are weird virtual variables, because, like you point out, they create object arrays. That makes them not so useful. It might be better to get rid of them entirely, or to replace them with some sort of shortcut that returns datetime64 instead.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,67332234
https://github.com/pydata/xarray/issues/386#issuecomment-91219325,https://api.github.com/repos/pydata/xarray/issues/386,91219325,MDEyOklzc3VlQ29tbWVudDkxMjE5MzI1,10194086,2015-04-09T12:44:16Z,2015-04-09T12:44:16Z,MEMBER,"Ok the problem seems to be that the datetime64 array is converted to an object array.

```
print(ts)
print(ts_mean)
```

Could you keep it as a datetime64 object? Or would that be a pandas problem?
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,67332234