polaris.cf.drop_inherited_attrs

polaris.cf.drop_inherited_attrs(ds: Dataset, names: Iterable[str]) → Dataset

Clear the attributes of each named variable whose attributes are identical to another variable’s. xarray propagates attributes through arithmetic, comparisons, where() and ones_like(), so a variable with exactly another’s attributes was derived from it and describes itself as that variable (a bottomDepth made from ones_like(xCell) calls itself the x coordinate of cell centers). A label someone wrote for a variable never duplicates another variable’s whole attribute set, so those are kept. Use before add_var_attrs() to give the cleared variables their own attributes.

Parameters:

• ds (xarray.Dataset) – A dataset about to be written

• names (iterable of str) – The variables to consider, typically the ones a table of attributes can describe. Variables not in the dataset are ignored.

Returns:

ds (xarray.Dataset) – A shallow copy of the dataset with the inherited attributes cleared