As discussed here #71 (comment), the CSV import of the locations dataset turns the CA climate zones to floats because the empty cells in the column are read as numpy NaNs. Revise to avoid this type conversion (maybe set the blanks to a negative integer like -999?)