NetCDF utilities
Most of the time xarray
load_dataset(), open_dataset() or
open_mfdataset() functions will be enough to handle your NetCDf file
or set of files.Casys Toolbox provides two additional utility functions to explore and read NetCDF
files containing groups and sub-groups of data.
nc_content
The
nc_content() function allows you to display the
content of a netCDF file.List the content of netCDF files including its subgroups.
Parameters
----------
paths
NetCDF file to explore (use the first file if a pattern is provided).
show_vars
Whether to list variables or not.
show_dims
Whether to list dimensions or not.
show_attrs
Whether to list attributes or not.
Returns
-------
:
List of groups and subgroups associated with their content.
Example showing the content of some netCDF files.
from casys.utilities import nc_content
files = "/path/to/JA3_GPN_2PfP057_*.nc"
nc_content(paths=files, show_vars=False, show_dims=True, show_attrs=False)
{'/': {'variables': 0, 'dimensions': [], 'attrs': 41},
'/data_01': {'variables': 83, 'dimensions': ['time'], 'attrs': 0},
'/data_01/ku': {'variables': 75, 'dimensions': [], 'attrs': 0},
'/data_01/c': {'variables': 27, 'dimensions': [], 'attrs': 0},
'/data_20': {'variables': 15, 'dimensions': ['time'], 'attrs': 0},
'/data_20/ku': {'variables': 31, 'dimensions': [], 'attrs': 0},
'/data_20/c': {'variables': 11, 'dimensions': [], 'attrs': 0}
}
open_grdataset
The
open_grdataset() function allows you to open the
content of a single or a set of NetCDF files including groups and sub-groups as well
as dimensions which might be included in upper-groups.open_grdataset() is build using open_mfdataset()
and allows you to benefit from the delayed loading as well as Dask parallel loading.Alternative function to xarray.open_mfdataset allowing to open a group
with its dimensions even if included in a parent group.
Subgroups might be included with their variables renamed by prefixing them with
their group path.
Parameters
----------
paths
NetCDF file(s) to read.
group
Name/path of the group to open.
sub_groups
Whether to include subgroups or not.
separator
Separator to use when renaming group fields (default to ".").
pre_kwargs
Additional xarray.open_dataset parameters used during the preprocessing.
mf_kwargs
Additional xarray.open_mfdataset parameters.
Returns
-------
:
xarray.Dataset.
Examples illustrating
open_grdataset() possibilities
when working with a set of netCDF.from casys.utilities import open_grdataset
files = "/path/to/JA3_GPN_2PfP057_*.nc"
# Opening a single subgroup
# time dimension, even if not included in the subgroup itself, will be included
# in the resulting dataset.
ds_01_ku = open_grdataset(paths=files, group="data_01/ku")
# Opening a subgroup with its subgroups (ku/c)
ds_01 = open_grdataset(paths=files, group="data_01")
# Opening a subgroup with its subgroups using a Dask cluster (if one is active)
ds_20 = open_grdataset(paths=files, group="data_20", mf_kwargs={"parallel": True})
# Opening the whole netCDF structure including data_01, data_20 and their respective
# subgroups
ds = open_grdataset(paths=files, mf_kwargs={"parallel": True})