Python Forum

Please consider the following Python code:

import pandas as pd


data = {
  "l1": [50, 100, 150, 200, 250, -350, 450],
  "l2": [100, 120, 150, 190, 240, 200, 500],
}

#load data into a DataFrame object:
df = pd.DataFrame(data)

I want to write code to add a new column to the data frame df where the new values are the larger of the values in l1 and l2. I can do this using a loop with the following code:

maxValue = []
limit = len ( df )
for i in range( limit ):
    if df.l1[i] > df.l2[i]:
        maxValue.append( df.l1[i] )
    else:
         maxValue.append( df.l2[i] )
df['l3'] = maxValue

However, I believe there should be a better way. I tired this:

df['l3'] = df['l1'] if df['l1'] > df['l2'] else df['l2']

This produced the following error message:

Error:
  File "C:\ProgramData\Anaconda3\lib\site-packages\pandas\core\generic.py", line 1478, in __nonzero__
    raise ValueError(

ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

What is the right way to add the new column l3 to df?

with numpy, like ifesle statment in R...

df['l3'] = np.where(df['l1'] > df['l2'], df['l1'],  df['l2'])

Alternatively

df['l3'] = df.apply(lambda row: max(row.l1,row.l2), axis=1)

Output:
	l1	l2	l3
0	50	100	100
1	100	120	120
2	150	150	150
3	200	190	200
4	250	240	250
5	-350	200	200
6	450	500	500