Pandas conditional creation of a new dataframe column

Question

This question is an extension of Pandas conditional creation of a series/dataframe column . If we had this dataframe:

    Col1       Col2
1    A          Z
2    B          Z           
3    B          X
4    C          Y
5    C          W

and we wanted to do the equivalent of:

if Col2 in ('Z','X') then Col3 = 'J' 
else if Col2 = 'Y' then Col3 = 'K'
else Col3 = {value of Col1}

How could I do that?

Answer 1

You can use loc with isin and last fillna :

df.loc[df.Col2.isin(['Z','X']), 'Col3'] = 'J'
df.loc[df.Col2 == 'Y', 'Col3'] = 'K'
df['Col3'] = df.Col3.fillna(df.Col1)
print (df)
  Col1 Col2 Col3
1    A    Z    J
2    B    Z    J
3    B    X    J
4    C    Y    K
5    C    W    C

Answer 2

Try this use np.where : outcome = np.where(condition, true, false)

  df["Col3"] = np.where(df['Col2'].isin(['Z','X']), "J", np.where(df['Col2'].isin(['Y']), 'K', df['Col1']))

  Col1 Col2 Col3
1    A    Z    J
2    B    Z    J
3    B    X    J
4    C    Y    K
5    C    W    C

Answer 3

A simple (but likely inefficient) way can be useful when you have multiple if condition. Like you are trying to put values into (say) four buckets based on quartiles.

df holds your data, col1 has the values, col2 should have the bucketized values (1,2,3,4) quart has the 25%, 50% and 75% bounds. try this

1. create a dummy list as dummy = []
2. iterate through the data frame with: for index, row in df.iterrows():
3. Set up the if conditions like: if row[col1] <= quart[0]:#25%
4. append proper value in dummy under the if: dummy.append(1)
5. the nested if-elif can take care of all the needed optional values which you append to dummy.
6. add dummy as a column: df[col2] = dummy

You can find the quartiles via A = df.describe() and then print(A[col1])