Using regex in second series.str.replace() argument

Question

I'm working with a DataFrame with a year column with the following format:

I want to replace strings like '2015-2016' with just '2015' using regex. I tried something like this:

df['year']=df['year'].str.replace('[0-9]{4}\-[0-9]{4}','[0-9]{4}')

But that doesn't work. I know I could do smething like:

df['year']=df['year'].str.replace('\-[0-9]{4}','')

But sometimes you need something more flexible. Is there any way to keep a portion of the match in the regex or is this one the standard approach?

Thanks in advance.

sacuL · Accepted Answer · 2018-09-11 17:49:29Z

2

If you just want to keep the first year, and all years have 4 digits, use:

df['year'] = df.year.str.extract('(\d{4})')
>>> df
   year
0  2015
1  2015
2  2016

If you want to keep the first year before any -, use:

df['year'] = df.year.str.split('-').str[0]

>>> df
   year
0  2015
1  2015
2  2016

answered Sep 11, 2018 at 17:49

sacuL

51.6k9 gold badges88 silver badges115 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Juan C Over a year ago

That exctarct function was something I should have already known, quite useful ! Thanks sacul

sacuL Over a year ago

Glad to help! Happy coding!

Steven Rumbalski · Accepted Answer · 2018-09-11 18:01:45Z

2

You can capture the good year in parenthesis and refer to it in your replacement with \1:

df['year'].str.replace(r'([0-9]{4})\-[0-9]{4}', r'\1')

Or you can make parenthesis around the good year into a non-capturing positive lookbehind assertion with ?<= and then the replacement string will be blank because only \-[0-9]{4} was matched (but only when preceded by [0-9]{4}).

df['year'].str.replace(r'(?<=[0-9]{4})\-[0-9]{4}', '')

answered Sep 11, 2018 at 18:01

Steven Rumbalski

45.8k10 gold badges96 silver badges125 bronze badges

2 Comments

Juan C Over a year ago

First time I hear about lookaround assertions. Do you know any site where there's a good and understandable explanation of how they work? The explanations I've found haven't been very helpful to really understand the difference between negative/positive and ahead/behind

Steven Rumbalski Over a year ago

@JuanC: Here's a short explanation and a longer one on lookaround assertions.

Collectives™ on Stack Overflow

Using regex in second series.str.replace() argument

2 Answers 2

2 Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related