Changing units in dataset

This is an example of how to change units in the xarray dataset.

Variables in read data
first_cbh: mean=677.3226928710938 units=m
second_cbh: mean=872.345703125 units=m
alt: mean=318.0 units=m

Variables with one changed to km
first_cbh: mean=0.6773227453231812 units=km
second_cbh: mean=872.345703125 units=m
alt: mean=318.0 units=m

Variables with both changed to km
first_cbh: mean=0.6773227453231812 units=km
second_cbh: mean=0.8723457455635071 units=km
alt: mean=0.318 units=km

Variables changed back to m by looping over all variables in dataset
Orginal dataset is same as retured dataset: True
first_cbh: mean=677.3226928710938 units=m
second_cbh: mean=872.345703125 units=m
alt: mean=318.0 units=m

Variables and coordinate variable values changed to km
Orginal dataset is same as retured dataset: False
first_cbh: mean=0.6773227453231812 units=km
second_cbh: mean=0.8723457455635071 units=km
alt: mean=0.318 units=km
range: mean=3.7800002098083496 units=km

from arm_test_data import DATASETS
import numpy as np

import act


def print_summary(ds, variables):
    for var_name in variables:
        print(
            f'{var_name}: mean={np.nanmean(ds[var_name].values)} '
            f"units={ds[var_name].attrs['units']}"
        )
    print()


variables = ['first_cbh', 'second_cbh', 'alt']

# Read in some example data
filename_ceil = DATASETS.fetch('sgpceilC1.b1.20190101.000000.nc')
ds = act.io.arm.read_arm_netcdf(filename_ceil)

# Print the variable name, mean of values and units
print('Variables in read data')
print_summary(ds, variables)

# Change units of one varible from m to km
ds.utils.change_units(variables='first_cbh', desired_unit='km')
print('Variables with one changed to km')
print_summary(ds, variables)

# Change units of more than one varible from to km
ds.utils.change_units(variables=variables, desired_unit='km')
print('Variables with both changed to km')
print_summary(ds, variables)

# Can change all data variables in the dataset that are units of length by not providing
# a list of variables. Here we are changing back to orginal meters.
# Also, because it needs to loop over all variables and try to convert, will take
# longer if we keep the QC variables. Faseter if we exclude them.
# The method will return a dataset. In this case the dataset returned is the same
# dataset.
skip_variables = [ii for ii in ds.data_vars if ii.startswith('qc_')]
new_ds = ds.utils.change_units(variables=None, desired_unit='m', skip_variables=skip_variables)
print('Variables changed back to m by looping over all variables in dataset')
print('Orginal dataset is same as retured dataset:', ds is new_ds)
print_summary(new_ds, variables)

# For coordinate variables need to explicitly give coordinage variable name and use
# the returned dataset. The xarray method used to change values on coordinate
# values requries returning a new updated dataet.
var_name = 'range'
variables.append(var_name)
new_ds = ds.utils.change_units(variables=variables, desired_unit='km')
print('Variables and coordinate variable values changed to km')
print('Orginal dataset is same as retured dataset:', ds is new_ds)
print_summary(new_ds, variables)

previous

Parse the ARM datastream filename

next

Working with TAR and gunzip files