如何通过比较两个数据帧的唯一ID来创建新列？

Question

Hi, I have two dataframes which have OrderID and stage numbers. I want to create a new column for August Dataframe which returns values in July's stage number if the orderID matches. If not, please return "N/A".

How should I use lambda and apply functions to create this column? (don't use join...)

Any clues and suggestions will be appreciated! Thanks!

Answer 1

You can use pd.Series.map with a series. Note if you have NaN values your series will be forced to float as NaN is a float value. This is unavoidable without adding inefficiencies.

aug = pd.DataFrame({'ID': [111, 222, 333, 444, 555], 'Prior': np.nan})
jul = pd.DataFrame({'ID': [222, 333, 444, 555, 666, 777], 'Stage': [1, 2, 3, 4, 5, 6]})

aug['Prior'] = aug['ID'].map(jul.set_index('ID')['Stage'])

print(aug)

    ID  Prior
0  111    NaN
1  222    1.0
2  333    2.0
3  444    3.0
4  555    4.0

A more long-winded solution is possible via pd.Series.update and aligning indices:

aug.set_index('ID', inplace=True)
aug['Prior'].update(jul.set_index('ID')['Stage'])
aug = aug.reset_index()

Answer 2

Although I hate posting this as an answer but if you are still interested in using lambda and apply you can you as below:

df=pd.DataFrame({'Order_id_July':[222,333,444,555,666,777],'stage':[1,2,3,4,5,6]})
df2=pd.DataFrame({'Order_id_August':[111,222,333,444,555]})

Mapper function(similar to lookup)

def myfunc(row):
    if set([row[0]]).intersection(set(df.Order_id_July)):       
        return int(df[df.Order_id_July==row[0]]['stage'])
    return np.nan

df2['prior_stage']=df2.apply(lambda x:myfunc(x),axis=1)

Output:

Order_id_August prior_stage
    111         NaN
    222         1.0
    333         2.0
    444         3.0
    555         4.0

If later you change your mind and want to explore better ways to achieving this task try below code

df2.merge(df,left_on='Order_id_August',right_on='Order_id_July',how='left').drop('Order_id_July',axis=1)

Order_id_August prior_stage
        111         NaN
        222         1.0
        333         2.0
        444         3.0
        555         4.0