issue_comments: 786759897
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/2799#issuecomment-786759897 | https://api.github.com/repos/pydata/xarray/issues/2799 | 786759897 | MDEyOklzc3VlQ29tbWVudDc4Njc1OTg5Nw== | 17484729 | 2021-02-26T16:43:23Z | 2021-02-26T16:43:23Z | NONE | Hi, I'm working on a machine learning application where I want to stream data and use xarray containers to store them in a buffer (with an additional "lag" dimension) and guaranty good alignement of the coordinates on various dimensions of the streamed data. Doing so, I noticed that the version of my code working with xarray is very slow when compared to a pure numpy implementation (with no coordinate alignement) or even an implementation with deque+pandas. I think the performance issue that I noticed is basically the same observation than the ones of this issue. I have the impression that for this kind of applications or more generally for intensive algorithmic usages, also as stated at the begining of this issue, a light (with less functionalities and checks) and fast version of xarray DataArray and Dataset containers could be developped. Do you think this could be something doable in the scope of xarray? Would it be preferable to create a dedicated library? |
{
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} |
416962458 |