Flag Variables

cf_xarray has some support for flag variables.

import cf_xarray
import xarray as xr

da = xr.DataArray(
    [1, 2, 1, 1, 2, 2, 3, 3, 3, 3],
    dims=("time",),
    attrs={
        "flag_values": [1, 2, 3],
        "flag_meanings": "atlantic_ocean pacific_ocean indian_ocean",
        "standard_name": "region",
    },
  name="region",
)
da.cf

Flag Variable:
       Flag Meanings:   1: atlantic_ocean
                        2: pacific_ocean
                        3: indian_ocean

Coordinates:
             CF Axes:   X, Y, Z, T: n/a

      CF Coordinates:   longitude, latitude, vertical, time: n/a

       Cell Measures:   area, volume: n/a

      Standard Names:   n/a

              Bounds:   n/a

       Grid Mappings:   n/a

Now you can perform meaningful boolean comparisons that take advantage of the flag_meanings attribute:

# compare to da == 1
da.cf == "atlantic_ocean"

<xarray.DataArray 'region' (time: 10)>
array([ True, False,  True,  True, False, False, False, False, False,
       False])
Dimensions without coordinates: time

xarray.DataArray

'region'

• time: 10

• True False True True False False False False False False

  array([ True, False,  True,  True, False, False, False, False, False,
         False])

• Coordinates: (0)
• Indexes: (0)
• Attributes: (0)

Similarly with membership tests using DataArray.cf.isin()

# compare to da.isin([2, 3])
da.cf.isin(["indian_ocean", "pacific_ocean"])

<xarray.DataArray 'region' (time: 10)>
array([False,  True, False, False,  True,  True,  True,  True,  True,
        True])
Dimensions without coordinates: time

xarray.DataArray

'region'

• time: 10

• False True False False True True True True True True

  array([False,  True, False, False,  True,  True,  True,  True,  True,
          True])

• Coordinates: (0)
• Indexes: (0)
• Attributes: (0)

You can also check whether a DataArray has the appropriate attributes to be recognized as a flag variable using DataArray.cf.is_flag_variable()

da.cf.is_flag_variable

True

Tip

cf_xarray does not support flag masks yet but a Pull Request to add this functionality is very welcome!