issue_comments: 298253809

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/1388#issuecomment-298253809 https://api.github.com/repos/pydata/xarray/issues/1388 298253809 MDEyOklzc3VlQ29tbWVudDI5ODI1MzgwOQ== 23484003 2017-04-30T20:08:25Z 2017-04-30T20:08:25Z NONE

Well, xarray at least agrees with numpy's implementation of that function, but that's not to say it is 'correct.' It would be nice if numpy.argmin worked intuitively. That aside, it seems to me that applying min() to a xr.DataArray should return a reduced array with length 1 in each dimension; then you could just query this object and find the coordinate/dimension values. Perhaps then argmin() would just get a tuple of axis indices, such that arr[*arr.argmin()] == arr.min() would hold.

The next question is, what happens if you start supplying coordinate/dimension optional arguments to argmin? It doesn't make sense to minimize over a coordinate, so only dimensions should be accepted. This should result in a tuple of lists, the way numpy.where does.

Does that seem reasonable?

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   224878728