act.transform.Transform

class act.transform.Transform(ds)

Xarray Dataset accessor exposing the act.transform kernels.

Every method here forwards straight through to the corresponding module-level function (act.transform.bin_average(), act.transform.interpolate(), act.transform.subsample(), act.transform.transform_dataset()) using self._ds[var] in place of an explicit DataArray argument. No transform logic is duplicated here.

Examples

The accessor takes variable names and looks them up in the Dataset, where the module-level functions take xarray.DataArray objects. Note that the QC argument changes name accordingly: qc_var_name here versus qc on the functions.

import act

ds = act.io.arm.read_arm_netcdf(filename, cleanup_qc=True)
target = act.transform.make_coord('2023-03-01', '2023-03-02', '30min')

result, result_qc = ds.transform.bin_average(
    'temp_mean', target, dim='time', qc_var_name='qc_temp_mean', qc_mask=4
)

These two calls are equivalent; the accessor forwards straight to the function, so neither form is more capable than the other.

result, result_qc = ds.transform.interpolate('temp_mean', target, dim='time')

result, result_qc = act.transform.interpolate(ds['temp_mean'], target, dim='time')

Any extra keyword arguments are passed through to the underlying function, and transform_dataset takes every argument except ds itself.

result, result_qc = ds.transform.bin_average(
    'wspd_arith_mean', target, dim='time', std_ind_max=0.8
)

new_ds = ds.transform.transform_dataset(
    target=target, dim='time', transform='bin_average', qc_mask=4
)

bin_average(var_name, target, dim[, ...])	Bin-average self._ds[var_name] onto target.
interpolate(var_name, target, dim[, ...])	Interpolate self._ds[var_name] onto target.
subsample(var_name, target, dim[, ...])	Subsample self._ds[var_name] onto target.
transform_dataset(**kwargs)	Apply a transform to every matching variable in this Dataset.