根据具有条件的其他列设置列的值

Question

I encountered a problem and it can probably be done by iterating over the rows of this DataFrame but there might be a more elegant solution.

I try to create the column 'desired' as a string if the value of col1 is higher than 20. I tried np.where without success.

DataFrame

Who can help me? Thanks!

Answer 1

This should work:

df['desired'] = ''
df.loc[df['col1'] > 20, 'desired'] = 'col1 is ' + df['col1'].astype(str)

Example

import pandas as pd

df = pd.DataFrame({'col1': [25, 10, 15, 21]})

df['desired'] = ''
df.loc[df['col1'] > 20, 'desired'] = 'col1 is ' + df['col1'].astype(str)

#    col1     desired
# 0    25  col1 is 25
# 1    10            
# 2    15            
# 3    21  col1 is 21

The problem with this

The power of pandas is in holding structured data. As soon as you combine strings with numeric data, you lose that structure. Manipulating strings is tedious, for example you can't add 1 to the "desired" column.

A better idea

It is better to use a Boolean column to signify a desired condition. For example:

df['desired'] = df['col1'] > 20

This will give a Boolean [True or False] series depending on the condition specified.

Answer 2

Use numpy.where for new column by condition:

df['desired'] = np.where(df['col1'] > 20, 'col1 is ' + df['col1'].astype(str), '')