How to Create new dataframe columns from existing column
new_df = df.filter(like='n_') \
.replace(0., np.inf) \
.apply(lambda x: sorted(x), axis=1, result_type='expand') \
.replace(np.inf, 0.0)
new_df.columns = ['new_1', 'new_2', 'new_3']
out = pd.concat([df, new_df], axis=1)