issue_comments

Table actions

11 rows where issue = 186680248 and user = 1217238 sorted by updated_at descending

✎ View and edit SQL

This data as json, CSV (advanced)

Suggested facets: updated_at (date)

user 1

  • shoyer · 11 ✖

issue 1

  • Allow concat() to drop/replace duplicate index labels? · 11 ✖

author_association 1

id html_url issue_url node_id user created_at updated_at ▲ author_association body reactions performed_via_github_app issue
274641329 https://github.com/pydata/xarray/issues/1072#issuecomment-274641329 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI3NDY0MTMyOQ== shoyer 1217238 2017-01-23T22:41:22Z 2017-01-23T22:41:22Z MEMBER

Fixed by #1204

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
271462549 https://github.com/pydata/xarray/issues/1072#issuecomment-271462549 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI3MTQ2MjU0OQ== shoyer 1217238 2017-01-10T01:42:26Z 2017-01-10T01:42:49Z MEMBER

@chunweiyuan import at the top of the method/ function, def fillna(self, other, ...): from .computation import apply_ufunc return apply_ufunc(...)

This is the pretty standard way to get around circular imports. There's basically no way to avoid this issue with APIs designed to have lots of logic in method calls. (Other than gigantic files, which isn't a real fix.)

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
270799884 https://github.com/pydata/xarray/issues/1072#issuecomment-270799884 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI3MDc5OTg4NA== shoyer 1217238 2017-01-06T00:39:32Z 2017-01-06T00:39:32Z MEMBER

I just merged #964.

If you're up for it, I would suggest rewriting fillna using apply_ufunc, instead of the current hack which relies on the injected _fillna method created with the _binary_op staticmethod (ugh).

Let me know if you need any guidance on using apply_ufunc or if the documentation is sufficient. In this case, at least you would not need to use confusing signature argument.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
261144196 https://github.com/pydata/xarray/issues/1072#issuecomment-261144196 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI2MTE0NDE5Ng== shoyer 1217238 2016-11-17T03:19:43Z 2016-11-17T03:19:43Z MEMBER

I would rather wait a little while here -- I think #964 will handle the logic we need to make add a join argument to fillna very easy, which I would prefer to adding a new method.

merge works a little differently than combine_first -- it raises an exception if any overlapping values conflict.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
260158359 https://github.com/pydata/xarray/issues/1072#issuecomment-260158359 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI2MDE1ODM1OQ== shoyer 1217238 2016-11-13T00:36:27Z 2016-11-13T00:36:27Z MEMBER

What about simply adding a keyword argument join to fillna? Then you could you write join='outer' to accomplish combine_first.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
258718785 https://github.com/pydata/xarray/issues/1072#issuecomment-258718785 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1ODcxODc4NQ== shoyer 1217238 2016-11-06T23:11:27Z 2016-11-06T23:11:27Z MEMBER

Use xarray.broadcast to do xarray style broadcasting (it also aligns). This will be easier after xarray.apply is merged

For now, maybe use Dataset._binary_op - take look at how fillna is implemented. This is actually almost exactly the same as fillna except using an outer join (instead of a left join).

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
258688677 https://github.com/pydata/xarray/issues/1072#issuecomment-258688677 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1ODY4ODY3Nw== shoyer 1217238 2016-11-06T15:35:43Z 2016-11-06T15:35:43Z MEMBER

I think we should still call it combine_first -- the only difference is the alignment behavior.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
258682131 https://github.com/pydata/xarray/issues/1072#issuecomment-258682131 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1ODY4MjEzMQ== shoyer 1217238 2016-11-06T13:50:54Z 2016-11-06T13:50:54Z MEMBER

Pandas implements this like: result = where(notnull(left), left, right). That's probably how we want want to do it -- much simpler than my proposal in the other issue.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
258237589 https://github.com/pydata/xarray/issues/1072#issuecomment-258237589 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1ODIzNzU4OQ== shoyer 1217238 2016-11-03T18:45:19Z 2016-11-03T18:45:19Z MEMBER

Indeed, I think we definitely want to do an outer align first. That behavior for pandas looks very strange.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
258026397 https://github.com/pydata/xarray/issues/1072#issuecomment-258026397 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1ODAyNjM5Nw== shoyer 1217238 2016-11-02T23:06:47Z 2016-11-02T23:06:47Z MEMBER

Sorry, nope not implemented yet. ops.choose also is not implemented yet, but it would be in the style of stack from core/ops.py.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248
257745190 https://github.com/pydata/xarray/issues/1072#issuecomment-257745190 https://api.github.com/repos/pydata/xarray/issues/1072 MDEyOklzc3VlQ29tbWVudDI1Nzc0NTE5MA== shoyer 1217238 2016-11-02T01:09:45Z 2016-11-02T01:09:45Z MEMBER

I don't think this is the best fit for concat. Rather, I would suggest a new method, something like combine_first. See these issues for more discussion: https://github.com/pydata/xarray/issues/835 https://github.com/pydata/xarray/issues/742

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Allow concat() to drop/replace duplicate index labels? 186680248

Advanced export

JSON shape: default, array, newline-delimited, object

CSV options: 

CREATE TABLE [issue_comments] (
   [html_url] TEXT,
   [issue_url] TEXT,
   [id] INTEGER PRIMARY KEY,
   [node_id] TEXT,
   [user] INTEGER REFERENCES [users]([id]),
   [created_at] TEXT,
   [updated_at] TEXT,
   [author_association] TEXT,
   [body] TEXT,
   [reactions] TEXT,
   [performed_via_github_app] TEXT,
   [issue] INTEGER REFERENCES [issues]([id])
);
CREATE INDEX [idx_issue_comments_issue]
    ON [issue_comments] ([issue]);
CREATE INDEX [idx_issue_comments_user]
    ON [issue_comments] ([user]);