CODEWITHSUNDEEP
pythonpandasdataframe
d84_n1nj4
d84_n1nj4

Reputation: 1852

Python -- add new column that takes other column values into account to create the values for the new column

So I want to create a new column in my dataframe, let's call it "outcome". I want each value in "outcome" to be determined based off values in two other columns. So let's say this is my dataframe:

raw_data = {
        'subject_id': ['1', '2', '3', '4', '5'],
        'first_name': ['Alex', 'Amy', 'Allen', 'Alice', 'Ayoung'], 
        'last_name': ['Anderson', 'Ackerman', 'Ali', 'Aoni', 'Atiches']}
df_a = pd.DataFrame(raw_data, columns = ['subject_id', 'first_name', 'last_name'])

This is what out dataframe looks like:

  subject_id first_name last_name
0          1       Alex  Anderson
1          2        Amy  Ackerman
2          3      Allen       Ali
3          4      Alice      Aoni
4          5     Ayoung   Atiches

So I want to create a new column that kind of takes in logic like this:

if(df_a[df_a['subject_id'] == 2] & df_a[df_a['first_name'] == 'Amy']):
  df_a['outcome'] = 2
elif(df_a[df_a['subject_id'] > 0] & df_a[df_a['first_name'] == 'Alice']):
  df_a['outcome'] = 1
else:
  df_a['outcome'] = 0

I cannot seem to get it to work properly though. I expect the new column to look like this:

  subject_id first_name last_name outcome
0          1       Alex  Anderson   0
1          2        Amy  Ackerman   2
2          3      Allen       Ali   0
3          4      Alice      Aoni   1
4          5     Ayoung   Atiches   0

Upvotes: 1

Views: 62

Answers (2)

Yale Newman
Yale Newman

Reputation: 1231

df_a.loc[(df_a['subject_id'] == 2) & (df_a['first_name'] == 'Amy'),'outcome']=2
df_a.loc[(df_a['subject_id'] > 0 ) & (df_a['first_name'] == 'Alice'),'outcome']=1
df_a['outcome'].fillna(0)

Not as good as the above, but looking at what you were trying to do, this will be relevant for you.

Upvotes: 0

jezrael
jezrael

Reputation: 862521

Use numpy.select or numpy.where:

#first convert `subject_id` to int
df_a['subject_id'] = df_a['subject_id'].astype(int)
m1 = (df_a['subject_id'] == 2) & (df_a['first_name'] == 'Amy')
m2 = (df_a['subject_id'] > 0) & (df_a['first_name'] == 'Alice')

df_a['outcome'] = np.select([m1, m2], [2,1], default=0)
print (df_a)
   subject_id first_name last_name  outcome
0           1       Alex  Anderson        0
1           2        Amy  Ackerman        2
2           3      Allen       Ali        0
3           4      Alice      Aoni        1
4           5     Ayoung   Atiches        0

Or:

df_a['outcome'] = np.where(m1, 2, np.where(m2, 1, 0))
print (df_a)
   subject_id first_name last_name  outcome
0           1       Alex  Anderson        0
1           2        Amy  Ackerman        2
2           3      Allen       Ali        0
3           4      Alice      Aoni        1
4           5     Ayoung   Atiches        0

Upvotes: 2

Related Questions