issue_comments: 134279075

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/548#issuecomment-134279075 https://api.github.com/repos/pydata/xarray/issues/548 134279075 MDEyOklzc3VlQ29tbWVudDEzNDI3OTA3NQ== 1217238 2015-08-24T16:18:00Z 2015-08-24T16:18:00Z MEMBER

This is actually already supported, though poorly documented (so it's basically unknown).

We seem to have some sort of bug in our documentation generation for recent versions, but in the v0.5.1 IO docs, you can see the encoding attribute at the end of the section on writing netCDFs: http://xray.readthedocs.org/en/v0.5.1/io.html#netcdf

The way this works is that encoding on each data array stores a dictionary of options that is used when serializing that array to disk. It support most of the options in netCDF4-python's createVariable method, including chunksizes, zlib, scale_factor, add_offset, _FillValue and dtype. This metadata is automatically filled in when reading a file from disk, which means that in principle xray should roundtrip the encoding faithfully.

Because encoding is read in when files are opened, invalid encoding options are currently ignored when saving a file to disk. This means that the current API is not very user friendly.

So I'd like to extend this into a keyword argument encoding for the to_netcdf method. The keyword argument would expect a dictionary where the keys are variable names and the values are encoding parameters, and errors would be raised for invalid encoding options. Here's my branch for that feature: https://github.com/shoyer/xray/tree/encoding-error-handling

 
{
    "total_count": 0,
    "+1": 0,
    "-1": 0,
    "laugh": 0,
    "hooray": 0,
    "confused": 0,
    "heart": 0,
    "rocket": 0,
    "eyes": 0
}
   102703065