Comparing columns in two different dataframes for matching values Pandas

Question

Here is what my two dataframes look like:

DF1

NAME   EMAIL                ID
Mark   [email protected]      8974
Sam    [email protected]       9823
June   [email protected]      0972
David  [email protected]     2143

DF2

ID     ROLE-ID
2143      22
0972      34
8974      98
9823      54

What I need to help doing:

I need to COMPARE the ID column for both dataframes and if the ID from DF1 MATCHES with the ID of DF2, I need to replace the ID column in DF1 with the respective ROLE-ID from DF2.

The output would look like this:

Updated DF1

NAME   EMAIL               ROLE-ID
Mark   [email protected]        98
Sam    [email protected]         54
June   [email protected]        34
David  [email protected]       22

I am using the Pandas library and tried the merge function with conditions but it didnt work

    print(pd.merge(df1, df2, on=(df1['Id'] == df2[])))

The goal isnt simply just to merge the two dataframes. The "ID" for both DFs has to match before it is replaced — abudda, Aug 28 '19 at 23:12
That is the whole purpose of joining. It's simply `df1.merge(df2, on='ID')` — Erfan, Aug 28 '19 at 23:33

score 0 · Answer 1 · answered Aug 28 '19 at 23:22

0

Try:

df = df1.merge(df2, on='ID', how='left')
df[['NAME', 'EMAIL', 'ROLE-ID']]

It gives the following:

Screenshot

answered Aug 28 '19 at 23:22

kelmouloudi

13
4

score 0 · Answer 2 · answered Aug 29 '19 at 00:43

You did not exactly state what should happen if id is not found or is avail multiple times this may not be 100% what you want. It will leave the id untouched then.B ut guess otherwise its what you wanted.

import pandas as pd
import numpy as np

df1 = pd.DataFrame([[1,'a'],
                    [7,'b'],
                    [3,'e'],
                    [2,'c']], columns=['id', 'name'])

df2 = pd.DataFrame([[1,2],
                    [3,8],
                    [2,10]], columns=['id', 'role'])

# collect roles
roles = []
for id in df1.loc[:, 'id']:
    indices = df2.loc[:,'id'] == id
    if np.sum(indices) == 1:
        roles.append(df2.loc[indices, 'role'].iloc[0])
    else:
        # take id if role id is not given
        roles.append(id)  # could also be None if not wanted

# add role id col
df1.loc[:,'role-id'] = roles

# delete old id
del df1['id']

print(df1)

DF1:

DF2:

   id  role
0   1     2
1   3     8
2   2    10

Output

  name  role-id
0    a        2
1    b        7
2    e        8
3    c       10

score -1 · Answer 3 · answered Aug 28 '19 at 22:55

-1

Seems like a merge problem

pd.merge(df2, df1, how='inner')

answered Aug 28 '19 at 22:55

stahamtan

848
6
10

1

you can also include `on` variable `pd.merge(df2, df1, how='inner', on='ID')` – stahamtan Aug 28 '19 at 22:57
Did you these answers? Because they are both wrong – Erfan Aug 28 '19 at 23:07
I might have misunderstood the problem. Could you help me understand why I was wrong? – stahamtan Aug 28 '19 at 23:51

Comparing columns in two different dataframes for matching values Pandas

3 Answers3