CODEWITHSUNDEEP
pythonpandas
Lex Borodai
Lex Borodai

Reputation: 25

How to use OR with .filter method from Pandas?

I have a PandaSeries with indicies that represent particular group and a value assigned to it

data = pd.Series(data=[12875.30, 12561.31, 12654.38, 13091.60, 13062.22, 13345.32], index=['A1','A2','B1','B2','C1','C2'])
data
A1    12875.30
A2    12561.31
B1    12654.38
B2    13091.60
C1    13062.22
C2    13345.32

What I am trying to achieve is to save the data for the B1, B2, C1 and C2 indicies and get rid of everything else using .filter method. The code below works, but only in case if I pass 1 attribute per time:

new_data = data.filter(like='B', axis=0) 
new_data = data.filter(like='C', axis=0)

The question is how to combine those two lines into one using or | operator? I've tried many options, for example:

new_data = data.filter(like=('B' | 'C'), axis=0) 
new_data = data.filter((like='B') | (like='C'), axis=0) 
new_data = (data.filter(like='B', axis=0)) | (data.filter(like='C', axis=0))
new_data = ((data.filter(like='B', axis=0)) | (data.filter(like='C', axis=0)))

but everything I come up with results an error. Would appreciate any help. Thank you!

Upvotes: 1

Views: 58

Answers (1)

Chris
Chris

Reputation: 16147

You could try using a regular expression:

data.filter(regex='B|C', axis=0)

Output

B1    12654.38
B2    13091.60
C1    13062.22
C2    13345.32

Upvotes: 1

Related Questions