merge_ordered

# good for time-series data or handling missing data
# standard
pd.merge_ordered(df1, df2, on='id', suffixes=('_str1', '_str2'))

# forward-fill
# missing values are filled with information from previous row
fill_method='ffill'