Rename Multiple pandas Dataframe Column Names using function

Question

Trying to rename pandas dataframe column using regex, I know how to do it rename the list as per below but could not get success results with df.rename.

Input:

df.columns.values = ['Time', '101 <RoomTemperature> (C)', '102 <ChemberTemperature> (C)', '103 <U1100> (C)', '103 <U1200 (C)', '103 U1500> (C)']

Trials of the Renaming dataframe column as per below code using regex but it does not work. I could not think of how to put multiple instruction together in df.rename method.

df.rename(columns={c: c.strip() for c in df.columns.values.tolist()
                                if "<" and ">" in c: 
                  re.search(r"(?<=<).*(?=>)",c).group(0)}, inplace=True)

I want it to follow regex and rename it to as per below:

df.columns.values = ["Time", "RoomTemperature", "ChemberTemperature", "U1100", "103 <U1200 (C)", "103 U1500> (C)"]

Dani Mesejo · Accepted Answer · 2020-03-29 11:57:59Z

3

You could extract the functionality into a function and do the following:

import re
import pandas as pd

# sample data 
df = pd.DataFrame(
    columns=['Time', '101 <RoomTemperature> (C)', '102 <ChemberTemperature> (C)', '103 <U1100> (C)', '103 <U1200 (C)',
             '103 U1500> (C)'])


# replacement function 
def repl(name):
    match = re.search(r"<(.*?)>", name)
    return match.group(1) if match else name


df.rename(columns={c: repl(c.strip()) for c in df.columns}, inplace=True)

print(df.columns)

Output

Index(['Time', 'RoomTemperature', 'ChemberTemperature', 'U1100',
       '103 <U1200 (C)', '103 U1500> (C)'],
      dtype='object')

That being said, you also need to fix your regular expression.

answered Mar 29, 2020 at 11:57

Dani Mesejo

62.2k6 gold badges56 silver badges86 bronze badges

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

Comments

Shubham Sharma · Accepted Answer · 2020-03-29 12:18:09Z

2

You can use regular expressions to extract the matching group as per your requirements and then you can use DataFrame.rename to alter the column labels.

Try this:

import re

col_dict = {}
for col in df.columns:
    mobj = re.search(r"\<(.*?)\>", col)
    if mobj:
        col_dict[col] = mobj.group(1)

df.rename(columns=col_dict, inplace=True)

After renaming df.columns will be:

['Time', 'RoomTemperature', 'ChemberTemperature', 'U1100', '103 <U1200 (C)', '103 U1500> (C)']

edited Mar 29, 2020 at 12:18

answered Mar 29, 2020 at 11:56

Shubham Sharma

71.8k6 gold badges26 silver badges58 bronze badges

4 Comments

Dani Mesejo Over a year ago

You are applying rename for each column, this could be applied just once

Shubham Sharma Over a year ago

@DaniMesejo I don't think thats going to make much difference.

ansev Over a year ago

you are using the rename method n times instead of just 1, obviously there is a performance issue here

OO7 Over a year ago

Thanks, Shubham and Dani

sammywemmy · Accepted Answer · 2020-03-29 12:12:31Z

1

Another solution, regex can have some unfriendly look about it, despite its power:

columns = ['Time', '101 <RoomTemperature> (C)', '102 <ChemberTemperature> (C)', '103 <U1100> (C)', '103 <U1200 (C)', '103 U1500> (C)']
df = pd.DataFrame([[1,2,3,4,5,6]],columns=columns)

   p = re.compile(r'((?<=<).*?(?=>))')

  #create a dict for the replacement

  replace_dict = {w:p.search(w).group() for w in df.columns if p.search(w)}

 #pass dictionary into rename method

 df.rename(columns=replace_dict)

answered Mar 29, 2020 at 12:12

sammywemmy

28.9k4 gold badges21 silver badges35 bronze badges

Collectives™ on Stack Overflow

Rename Multiple pandas Dataframe Column Names using function

3 Answers 3

Comments

4 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

4 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related