xarray_masks

File Purpose: tools related to masking xarray objects

for simply filling masked regions with nan, can use xarray_obj.where(…).

But, that can cause a lot of memory full of nan which is not desireable.

Methods in here help with stacking/unstacking mask dimensions,

so it can be possible to keep only the non-nan points.

Datasets are able to store the full mask info as a data_var,

so it is possible to ds.pc.mask() or ds.pc.unmask() if ‘_mask’ already in ds.

DataArrays can’t store extra dimensions,

so mask needs to be passed, e.g. arr.pc.mask(mask), arr.pc.unmask(mask)

Functions

xarray_demask_from_ds(ds, array[, stackdim, ...])	unmask (i.e., unstack) a masked (and stacked) xarray object, using mask=xarray_stored_mask(ds).
xarray_demask_from_mask(mask, array[, ...])	unmask (i.e., unstack) a masked (and stacked) xarray object, using this mask.
xarray_has_mask(array)	returns whether an xarray object has a mask.
xarray_mask(array[, mask, stackdim, stack, ...])	mask an xarray object; result has non-nan values only where mask=False.
xarray_popped_mask(ds)	returns (stored mask, copy of ds without '_mask')
xarray_store_mask(array, mask)	return Dataset like input but containing data_var '_mask'.
xarray_stored_mask(ds)	returns original mask, based on info stored in ds.
xarray_unmask(array[, mask, stackdim, ...])	unmask (i.e., unstack) a masked (and stacked) xarray object.
xarray_unmask_var(ds, var[, mask, stackdim, ...])	unmask a single variable from a Dataset.
xarray_unmask_vars(ds, vars[, mask, ...])	unmask multiple variables from a Dataset.