# Pandas group by a column looking at the count unique/count distinct values of another column df.groupby('param')['group'].nunique()