I am dealing with a large set of data with a lot of different label names in it. Is there a way to only demonstrate the top few high percentages of attributes to the pie cha