issues: 60766810
This data as json
| id | node_id | number | title | user | state | locked | assignee | milestone | comments | created_at | updated_at | closed_at | author_association | active_lock_reason | draft | pull_request | body | reactions | performed_via_github_app | state_reason | repo | type |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 60766810 | MDU6SXNzdWU2MDc2NjgxMA== | 368 | Default reading netCDF3 files with scipy.io instead of netCDF4? | 1217238 | closed | 0 | 4 | 2015-03-12T03:44:41Z | 2019-01-15T20:10:10Z | 2019-01-15T20:10:10Z | MEMBER | In my microbenchmarks, scipy.io appears to be ~3x faster than netCDF4 for reading netCDF3 files: ``` python ds = xray.Dataset({'foo': (['x', 'y'], np.random.randn(10000, 10000).astype(np.float32))}) ds.to_netcdf('test.nc', engine='scipy') ds_scipy = xray.open_dataset('test.nc', engine='scipy') ds_nc4 = xray.open_dataset('test.nc', engine='netcdf4') %timeit ds_scipy.isel(x=slice(5000)).load_data() 10 loops, best of 3: 123 ms per loop%timeit ds_nc4.isel(x=slice(5000)).load_data() 1 loops, best of 3: 319 ms per loop``` We might want to switch the default engine to use scipy for reading netCDF3 files. Note that netCDF4 does seem to be a bit faster for writing. |
{
"url": "https://api.github.com/repos/pydata/xarray/issues/368/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} |
completed | 13221727 | issue |