issue_comments: 892477231

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/5655#issuecomment-892477231 https://api.github.com/repos/pydata/xarray/issues/5655 892477231 IC_kwDOAMm_X841Mh8v 14371165 2021-08-04T08:39:53Z 2021-08-04T08:39:53Z MEMBER

I'm not so sure it simplifies that considerably. The linked PR is the minimal changes I had to do to get it working for my use cases and most of the changes were just removing unneccessary dict(x). I admittedly haven't checked every part of the code yet though.

My files have 2000+ variables with each variable having like 8 attributes. It starts taking a while when you have to read each one of those. At the moment, reading from file to Dataset takes about 2s, 600ms of those were reading attributes. With the PR I got it down to 200ms. Not as much as I'd hoped but I think I can get my LazyDict implementation much faster.

Changing file formats is too large of a change. We have used hdf5-files for many years and just switching to a different file format is just not something you do in painless way without (fast) backwards compatible alternative. It's hard to motivate a switch to xarray if the old alternative reads in files faster.

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   957201551