IF statement for dataframes in Python

Question

I have a dataset that looks like this.

Conatct 1	Contact 2	Contact 3
NA	XYZ	STU
NA	NA	LMN
ABC	PQR	NA

I'm trying to write a new column which will consolidate the contact information from the three different columns into one consolidated column such that if Contact 1 is NA then contact information from Contact 2 will be picked up and if Contact 2 is also empty then Contact 3 will get picked up. Below is the code I have so far.

def rm_contact(row):
   if row["LOB Contact #1"] != "NA":
     return row["Contact 1"]      
   else:
     return row["Contact 2"] 

df2 = df2.assign(rm=df2.apply(rm_contact, axis=1))

The else part of the code doesn't seem to work. So even though Contact 1 is NA the details from Contact 2 or Contact 3 don't get filled in properly. Any suggestions?

jezrael · Accepted Answer · 2021-02-15 07:01:21Z

1

Use if NA is missing value - back filling missing values per rows and then select first column:

cols = ["Contact 1","Contact 2","Contact 3"]
df2['rm'] = df2[cols].bfill(axis=1).iloc[:, 0]

If NA is string first replace values to missing NaNs:

df2['rm'] = df2[cols].replace('NA',np.nan).bfill(axis=1).iloc[:, 0]

answered Feb 15, 2021 at 7:01

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Rob Raymond · Accepted Answer · 2021-02-15 07:04:36Z

1

Simple case of chain together fillna() in order you want values to be picked

df = pd.read_csv(io.StringIO("""Contact 1   Contact 2   Contact 3
NA  XYZ STU
NA  NA  LMN
ABC PQR NA
"""), sep="\t")

df.assign(rm_contact=df["Contact 1"].fillna(df["Contact 2"]).fillna(df["Contact 3"]))

	Contact 1	Contact 2	Contact 3	rm_contact
0	nan	XYZ	STU	XYZ
1	nan	nan	LMN	LMN
2	ABC	PQR	nan	ABC

answered Feb 15, 2021 at 7:04

Rob Raymond

31.5k3 gold badges19 silver badges34 bronze badges

Collectives™ on Stack Overflow

IF statement for dataframes in Python

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related