useing regular expression for column filter

Question

I have a pandas DataFrame that contains a year and a title. Some of rows.name are Batman, Catman etc. I am trying to get all rows of the format .*man, using the following expression:

t[t.title & re.search(r'.*man', t.title)]

This Fails. Is there an idiomatic way to accomplish this?

score 3 · Answer 1 · answered Jun 18 '16 at 16:27

3

t[t.title & (t.title[-3:] == 'man')]

answered Jun 18 '16 at 16:27

Julien Goupy

165
6

great; this will work ; but I may want put a complex regular expression and I want work around match object is there an operator like ~ or re_match ? – forvaidya Jun 18 '16 at 16:28
Use a function : `def select_(t): return t.title[-3:] == 'man'` then t[t.title & select_(t)]. When you will need a more complex selection, just change select_ – Julien Goupy Jun 18 '16 at 16:34
means select_t() : must return True or False ? correct . – forvaidya Jun 18 '16 at 16:45
I do not know how your DF lib (panda i assume) is working. I just replace my solution in your code. if a bool is expected, yes. – Julien Goupy Jun 18 '16 at 16:48

shivsn · Answer 2 · 2016-06-18T17:20:22.170

2

use str.contains:
df[df['title'].str.contains('.man')].

edited Jun 18 '16 at 17:20

answered Jun 18 '16 at 16:57

shivsn

7,680
1
26
33

This will not work as 'mangggggggg' will be selected thus not wanted. – Julien Goupy Jun 18 '16 at 17:11

score 0 · Answer 3 · edited May 23 '17 at 10:29

0

following solution worked

df.ix[[x for x in df.index if re.search(r'.*e.*', x)]]

Inspired from || by Pandas - filter and regex search the index of DataFrame

edited May 23 '17 at 10:29

Community

1
1

answered Jun 19 '16 at 14:30

forvaidya

3,041
3
26
33

useing regular expression for column filter

3 Answers3