current community

your communities

more stack exchange communities

Stack Exchange
tour help
careers 2.0
Take the 2-minute tour ×
Stack Overflow is a question and answer site for professional and enthusiast programmers. It's 100% free, no registration required.
up vote 1 down vote favorite

I'm new to pandas and trying to figure out how to convert multiple columns which are formatted as strings to float64's. Currently I'm doing the below, but it seems like apply() or applymap() should be able to accomplish this task even more efficiently...unfortunately I'm a bit too much of a rookie to figure out how. Currently the values are percentages formatted as strings like '15.5%'

for column in ['field1', 'field2', 'field3']:
    data[column] = data[column].str.rstrip('%').astype('float64') / 100
share|improve this question
add comment

2 Answers

up vote 7 down vote accepted

Starting in 0.11.1 (coming out this week), replace has a new option to replace with a regex, so this becomes possible

In [14]: df = DataFrame('10.0%',index=range(100),columns=range(10))

In [15]: df.replace('%','',regex=True).astype('float')/100
Out[15]: 
<class 'pandas.core.frame.DataFrame'>
Int64Index: 100 entries, 0 to 99
Data columns (total 10 columns):
0    100  non-null values
1    100  non-null values
2    100  non-null values
3    100  non-null values
4    100  non-null values
5    100  non-null values
6    100  non-null values
7    100  non-null values
8    100  non-null values
9    100  non-null values
dtypes: float64(10)

And a bit faster

In [16]: %timeit df.replace('%','',regex=True).astype('float')/100
1000 loops, best of 3: 1.16 ms per loop

 In [18]: %timeit df.applymap(lambda x: float(x[:-1]))/100
1000 loops, best of 3: 1.67 ms per loop
share|improve this answer
    
Can you please tell me how can I implement this for specific columns? df['Column1'].replace('%','',regex=True).astype('float')/100 didn't work. –  erantdo Feb 10 at 9:40
add comment
up vote 1 down vote 
df.applymap(lambda x:float(x.rstrip('%'))/100)
share|improve this answer
1  
slightly more efficient to use df1.applymap(lambda x: float(x[:-1]))/100... –  Andy Hayden May 20 '13 at 10:24
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.