FIX: read_arm_mmcr fails when using a read-only input file
#888
Conversation
isilber
requested review from
AdamTheisen,
zssherman and
mgrover1
as code owners
act/io/arm.py
Outdated
| if nc is not None: | ||
| ds = xr.open_dataset(xr.backends.NetCDF4DataStore(nc)) | ||
| # Change heights name to range to read appropriately to xarray | ||
| if 'heights' in ds.dims: | ||
| ds = ds.rename_dims({"heights": "range"}) |
There was a problem hiding this comment.
We are using heights later on for some calculations, this might break
There was a problem hiding this comment.
@zssherman Well, it doesn't, and I've tested it. Note that ds also has the 'heights' dim name changed prior to this PR.
As it stands, ARM currently holds only the b1 MMCR files, which have that dimension. The issue was with read-only files. Other than that, everything remained the same.
You can try the few lines of code in the issue I opened and see that it works.
There was a problem hiding this comment.
Yeah i was curious on why we are checking for nans later on etc for height and if that would break if we had range, but I see we check for range as well. Thanks for the clarification.
There was a problem hiding this comment.
I might need to test this out. There were issues with opening the mmcr dataset I was working with without renaming that dimension ahead of opening up with xarray. This week is hectic but I'll try and fit it in.
|
Tests failing is unrelated to this PR |
…did not work on my mac after downloading data.
|
@isilber I was running into errors when using this change on my local machine but I've tested this out on read only files and seems to work. Would you be able to test this out first before we merge? |
There was a problem hiding this comment.
I tested this fix out on ARM's dev system and was able to read from /data/archive. I think we can go ahead and merge this in @zssherman. Thanks all!
act/io/arm.py
Outdated
| if nc is not None: | ||
| ds = xr.open_dataset(xr.backends.NetCDF4DataStore(nc)) | ||
| # Change heights name to range to read appropriately to xarray | ||
| if 'heights' in ds.dims: | ||
| ds = ds.rename_dims({"heights": "range"}) |
There was a problem hiding this comment.
I might need to test this out. There were issues with opening the mmcr dataset I was working with without renaming that dimension ahead of opening up with xarray. This week is hectic but I'll try and fit it in.
read_arm_mmcrwhen using read-only files as input #887