# credit to the Stack Overflow user in source link # vector is the label of a column containing fixed - size numpy (np) array df.groupby('your_label')['vector'].apply(np.mean)