Silent truncation of values is no longer supported in the newest numpy versions (see NEP 50 ).
numpy.uint8(-1) has therefore been replaced with numpy.iinfo(numpy.uint8).max in get_default_fill_value, and further related modifications were made. The constraint on netcdf4 was there to constrain the numpy version to <2.0.0 and is therefore no longer required and has been removed.
Silent truncation of values is no longer supported in the newest numpy versions (see NEP 50 ).
numpy.uint8(-1)has therefore been replaced withnumpy.iinfo(numpy.uint8).maxinget_default_fill_value, and further related modifications were made. The constraint onnetcdf4was there to constrain the numpy version to <2.0.0 and is therefore no longer required and has been removed.