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/pull/1347#issuecomment-292832731,https://api.github.com/repos/pydata/xarray/issues/1347,292832731,MDEyOklzc3VlQ29tbWVudDI5MjgzMjczMQ==,6815844,2017-04-10T02:05:38Z,2017-04-10T02:05:38Z,MEMBER,Oops. Found a typo in reshaping.rst. I'll send another PR.,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-292826966,https://api.github.com/repos/pydata/xarray/issues/1347,292826966,MDEyOklzc3VlQ29tbWVudDI5MjgyNjk2Ng==,1217238,2017-04-10T01:02:06Z,2017-04-10T01:02:18Z,MEMBER,This looks great. Thank you @fujiisoup!,"{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-292469365,https://api.github.com/repos/pydata/xarray/issues/1347,292469365,MDEyOklzc3VlQ29tbWVudDI5MjQ2OTM2NQ==,6815844,2017-04-07T08:01:02Z,2017-04-07T08:01:02Z,MEMBER,"I added a section to `reshaping.rst`, which shows an example (default) usage of `expand_dims` and `squeeze`.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291812719,https://api.github.com/repos/pydata/xarray/issues/1347,291812719,MDEyOklzc3VlQ29tbWVudDI5MTgxMjcxOQ==,6815844,2017-04-05T09:54:59Z,2017-04-05T09:54:59Z,MEMBER,"Thanks for the careful review.

+ Now it raises `IndexError` if any axis is out of bounds.
   `ValueError` is OK for the other part?
+ Removed unnecessary `print` line.
+ Removed the `unzip` -> `rezip` sequences.
+ Divide `test_dataarray.test_expand_dims` into small 2 tests.
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291493199,https://api.github.com/repos/pydata/xarray/issues/1347,291493199,MDEyOklzc3VlQ29tbWVudDI5MTQ5MzE5OQ==,6815844,2017-04-04T13:05:22Z,2017-04-04T13:05:22Z,MEMBER,"Thanks for the review.

I completely forgot the negative axis. Thanks.
It is supported by my last commit.

I also changed `Variable.expand_dims` to `Variable.set_dims`.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291189965,https://api.github.com/repos/pydata/xarray/issues/1347,291189965,MDEyOklzc3VlQ29tbWVudDI5MTE4OTk2NQ==,1217238,2017-04-03T16:06:10Z,2017-04-03T16:06:10Z,MEMBER,"> How can I add API docs?
> Only to add a single line to api.rst?

Yes, that's correct.

> I changed my mind. reshape sounds too general. How about append_dims or attach_dims ?

Indeed, I would rather not do reshape, because this method does not support arbitrary reshaping (e.g., length 100 vector-> 10x10 matrix). Instead, it assigns dimensions using array broadcasting and transpose, which unlike reshape does not ever copy data. Maybe `set_dims` would be better? I also would be OK with `attach_dims`. `append_dims` makes it sound like dimensions are always appended (they aren't).

Actually, `reshape` could indeed be a nice method to add to `Dataset`/`DataArray`, as an alternative to `stack`/`unstack`.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291128527,https://api.github.com/repos/pydata/xarray/issues/1347,291128527,MDEyOklzc3VlQ29tbWVudDI5MTEyODUyNw==,6815844,2017-04-03T12:29:17Z,2017-04-03T12:29:17Z,MEMBER,"I moved `expand_dims` to `Dataset`.

How can I add API docs? 
Only to add a single line to api.rst?

> How about `reshape`?

I changed my mind. `reshape` sounds too general. How about `append_dims` or `attach_dims` ?
","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291078104,https://api.github.com/repos/pydata/xarray/issues/1347,291078104,MDEyOklzc3VlQ29tbWVudDI5MTA3ODEwNA==,6815844,2017-04-03T08:21:58Z,2017-04-03T08:22:12Z,MEMBER,"> Maybe with_dims would be a clearer name.

How about `reshape`? 
I think verb-name gives more intuitition about the action of the method.
The current implementation of `Variable.expand_dims` actually does this when `Ordered_dict` is passed.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734
https://github.com/pydata/xarray/pull/1347#issuecomment-291009605,https://api.github.com/repos/pydata/xarray/issues/1347,291009605,MDEyOklzc3VlQ29tbWVudDI5MTAwOTYwNQ==,1217238,2017-04-02T19:39:59Z,2017-04-02T19:39:59Z,MEMBER,"> The proposed API is similar to numpy.expand_dims and slightly different from Variables.expand_dims, which requires whole sequences of dims of the result array.

This looks great, but to avoid confusion, can we rename the existing `Variable.expand_dims` method? Maybe `with_dims` would be a clearer name.

> My concern is that I do not yet fully understand the lazy data manipulation in xarray.
Does Variable.expand_dims do it?

Yes, `Variable.expand_dims` is already lazy on dask arrays.","{""total_count"": 0, ""+1"": 0, ""-1"": 0, ""laugh"": 0, ""hooray"": 0, ""confused"": 0, ""heart"": 0, ""rocket"": 0, ""eyes"": 0}",,218745734