Getting the max of three calculations in pandas

Question

I tried looking for an answer but couldn't find it. If someone has a link, all the better. My question goes as follows.

I have a sheet I am reading in pandas which has numerous columns with values in it.
There are three calculations I need to run which use bits from different columns at a time.
I need to return the max of those calculations in one bit of code, which will then be added to a new column.

I'm having problems with number 2.

Here is what my code would look like.

df = read_csv('file.csv')

df['Get New'] = maximum(df[Long] - df[Short], df[Long] + df[Short], df[Long] / df[Short])

df.to_csv('newFile.csv', index=False)

I know maximum doesn't work in this scenario but I can't seem to find what would. Any help is appreciated. Thanks!

EDIT: Here is the solve.

df['Get New'] = np.maximum(df['Long'] - df['Short'],  df['Long'] + df['Short'])
df['Get New'] = np.maximum(df['Get New'], df['Long'] / df['Short'])

jezrael · Accepted Answer · 2021-04-09T06:34:26.577

3

Use np.maximum with reduce:

import numpy as np


df = pd.DataFrame({
         'Long':[7,8,9,4,2,3],
         'Short':[1,3,5,7,1,7],
})

df['Get New'] = np.maximum.reduce([df['Long'] - df['Short'], 
                                   df['Long'] + df['Short'], 
                                   df['Long'] / df['Short']])
print (df)
   Long  Short  Get New
0     7      1      8.0
1     8      3     11.0
2     9      5     14.0
3     4      7     11.0
4     2      1      3.0
5     3      7     10.0

Alternative is use np.maximum for pairs only:

df['Get New'] = np.maximum(df['Long'] - df['Short'],  df['Long'] + df['Short'])
df['Get New'] = np.maximum(df['Get New'], df['Long'] / df['Short'])

edited Apr 09 '21 at 06:34

answered Apr 06 '21 at 06:55

jezrael

822,522
95
1,334
1,252

I've adjusted to use np.maximum, but I get this error "RecursionError: maximum recursion depth exceeded". I'm also using np.absolute in each of these calculations so I'm not sure if that's causing additional problems. – Ian Lang Apr 07 '21 at 21:09
@IanLang - Are columns `Lang` and `Short` numeric? What return `print (df.dtypes)` ? – jezrael Apr 08 '21 at 04:59
They come back as float64 – Ian Lang Apr 08 '21 at 05:12
@IanLang - You are right, answer was edited. – jezrael Apr 08 '21 at 05:18
It's getting closer it seems, but I'm getting this error. "TypeError: 'Series' object cannot be interpreted as an integer". Is there another thing to use besides reduce? – Ian Lang Apr 09 '21 at 06:31
@IanLang - What is your pandas version? – jezrael Apr 09 '21 at 06:32
@IanLang - Or use alternative, added to answer. – jezrael Apr 09 '21 at 06:34
1

That works! Thanks! I've edited the original post above to take that into account. – Ian Lang Apr 10 '21 at 02:34

Getting the max of three calculations in pandas

1 Answers1