imod.util.where#

imod.util.where(condition, if_true, if_false, keep_nan: bool = True) xarray.core.dataarray.DataArray[source]#

Wrapped version of xarray’s .where.

This wrapped version does two differently:

Firstly, it prioritizes the dimensions as: if_true > if_false > condition. xarray.where(cond, a, b) will choose the dimension over a or b. This may result in unwanted dimension orders such as ("y", "x", "layer) rather than ("layer", "y', "x").

Secondly, it preserves the NaN values of if_true by default. If we wish to replace all values over 5 by 5, yet keep the NoData parts, this requires two operations with with xarray’s where.

Parameters
  • condition (DataArray, Dataset) – Locations at which to preserve this object’s values. dtype must be bool.

  • if_true (scalar, DataArray or Dataset, optional) – Value to use for locations where cond is True.

  • if_false (scalar, DataArray or Dataset, optional) – Value to use for locations where cond is False.

  • keep_nan (bool, default: True) – Whether to keep the NaN values in place of if_true.