Clash Royale CLAN TAG#URR8PPP

pandas dataframe fast apply function on multiple columns

I have a dataframe df with multiple columns (not sure how many). One of the columns is called x. I have defined a function my_function that takes 2 columns as inputs and does something them and returns a new column.
I want to perform the my_function(df[x], df[y]) on all columns of the dataframe (where y is all columns one by one except for x) and return a new df with exact same column names but update values from my_function:

x

my_function

my_function(df[x], df[y])

y

x

my_function

new_df = my_function(df[x], df[y])


where y = all columns in df except for x. What is the fastest way to do this?

my_function

2 Answers
2

if your function performs row-wise preprocessing, you can use the method apply (see doc ) for each column:

columns = [item for item in df.columns.values if item!=x]
for c in coulmns:
df[c] = df.apply(lambda row: my_function(row[c], row[x]) ,axis=1)


If not, can you share "my_function"'s body ?

Bests,

You could try with something like this:

x = 'col_name'
[my_function(df[x], df[col]) for col in df if col!=x]


Example:

df = pd.DataFrame( 'a':[1,2],'b':[4,3],'c':[5,9], 'd':[2,9])
x = 'a'
df1 = pd.DataFrame()
def my_function(a,b, name):
c = a+b
df1[name] = c

[my_function(df[x], df[col], name) for name, col in enumerate(df) if col!=x]
print df1


Output:

1 2 3
0 5 6 3
1 5 11 11


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.