Pivot and convert rows to columns in pandas groupby

# pivot args: 1st -> index, 2nd -> columns, 3rd -> values
 
df.pivot(["a", "b"], "c", "d").rename_axis(columns=None).reset_index()