issue_comments

Table actions

8 rows where issue = 535703663 sorted by updated_at descending

✎ View and edit SQL

This data as json, CSV (advanced)

Suggested facets: reactions, created_at (date), updated_at (date)

author_association 3

issue 1

  • Feature Request: Efficient rolling with strides · 8 ✖
id html_url issue_url node_id user created_at updated_at ▲ author_association body reactions performed_via_github_app issue
888250482 https://github.com/pydata/xarray/issues/3608#issuecomment-888250482 https://api.github.com/repos/pydata/xarray/issues/3608 IC_kwDOAMm_X8408aBy niowniow 5802846 2021-07-28T11:58:28Z 2021-07-28T11:58:28Z CONTRIBUTOR

You are right. It's quite confusing. I've already added a stride parameter in my PR #3607 I didn't follow through with it and at the moment the checks are not successful anymore. Maybe someone else could give an opinion on the pro/cons of a stride parameter in rolling? 

{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
884377683 https://github.com/pydata/xarray/issues/3608#issuecomment-884377683 https://api.github.com/repos/pydata/xarray/issues/3608 IC_kwDOAMm_X840tohT kmsquire 223250 2021-07-21T17:53:59Z 2021-07-21T17:53:59Z CONTRIBUTOR

Question: instead of adding stride to reduce and _reduce_method, why not add it as a member of DataArrayRolling directly? This would allow, e.g., __iter__ to use it as well, and seems like a cleaner interface.

I've been confused why some parameters are available only in construct (stride, fill_value), some are available both in construct and in the DataArrayRolling constructor (keep_attrs), and some are only available in the constructor (min_periods, center, and soon pad).

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
851641056 https://github.com/pydata/xarray/issues/3608#issuecomment-851641056 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDg1MTY0MTA1Ng== dcherian 2448579 2021-05-31T19:09:00Z 2021-05-31T19:09:00Z MEMBER

sliding_window_view is a numpy function (see npcompat.py), so you need not transform to dask.

For reduce I think we just have to pass stride to _construct as in https://github.com/pydata/xarray/issues/3608#issuecomment-811156323 and for bottleneck insert the .isel call (copied from the end of _construct) after the DataArray is constructed.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
851469496 https://github.com/pydata/xarray/issues/3608#issuecomment-851469496 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDg1MTQ2OTQ5Ng== niowniow 5802846 2021-05-31T12:50:37Z 2021-05-31T12:50:37Z CONTRIBUTOR

Quickly glancing over sliding_window_view I didn't immediately understand how to use it with stride. Would I need to 1. transform the DataArray to dask array using chunk (which may involve an overhead!?), 2. then use rolling which itself uses sliding_window_view because its a dask array!? 3. Then use isel with stride on the new dimension?

reduce can easily support stride by passing it on here:

I think that's what I did in #3607. It's been a while

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
811156323 https://github.com/pydata/xarray/issues/3608#issuecomment-811156323 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDgxMTE1NjMyMw== dcherian 2448579 2021-03-31T15:27:47Z 2021-03-31T15:27:47Z MEMBER

No. but this should be really easy to fix.

construct already supports stride.

reduce can easily support stride by passing it on here: https://github.com/pydata/xarray/blob/57a4479fcd3ebc579cf00e0d6bf85007eda44b56/xarray/core/rolling.py#L441-L443

We should also add it to _reduce_method

bottleneck does not support stride so we can only use a .isel call at the end of _bottleneck_reduce https://github.com/pydata/xarray/blob/57a4479fcd3ebc579cf00e0d6bf85007eda44b56/xarray/core/rolling.py#L518

Note: sliding_window_view does not support stride because it's easy to stride after constructing that view (i saw this on some numpy issue)

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
810918988 https://github.com/pydata/xarray/issues/3608#issuecomment-810918988 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDgxMDkxODk4OA== keewis 14808389 2021-03-31T09:26:03Z 2021-03-31T09:26:03Z MEMBER

@dcherian, should this have been closed by #4977?

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
810663633 https://github.com/pydata/xarray/issues/3608#issuecomment-810663633 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDgxMDY2MzYzMw== GenevieveBuckley 30920819 2021-03-31T00:24:39Z 2021-03-31T00:24:39Z NONE

Is this useful/relevant for you? https://github.com/dask/dask/pull/7234

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   Feature Request: Efficient rolling with strides 535703663
564023352 https://github.com/pydata/xarray/issues/3608#issuecomment-564023352 https://api.github.com/repos/pydata/xarray/issues/3608 MDEyOklzc3VlQ29tbWVudDU2NDAyMzM1Mg== niowniow 5802846 2019-12-10T13:05:10Z 2019-12-10T13:05:10Z CONTRIBUTOR 
{
    "total_count": 4,
    "+1": 1,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 3
}
   Feature Request: Efficient rolling with strides 535703663

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]);