Currently the PointWiseDownscaler seems to fail if the model data time dimension does not have a pandas Datetime Index, and it fails with KeyError: 'time' even when the time attribute is present. This occurs in the _da_to_df function when transposing the data array and extracting the time index for the resulting dataframe.
Currently the PointWiseDownscaler seems to fail if the model data time dimension does not have a pandas Datetime Index, and it fails with
KeyError: 'time'
even when the time attribute is present. This occurs in the_da_to_df
function when transposing the data array and extracting the time index for the resulting dataframe.