The flags to set from the re library (e.g. re.IGNORECASE). Multiple flags can be set by combining them with the bitwise | (e.g. re.IGNORECASE | re.MULTILINE).

Return Value

A multi-index DataFrame.

Examples

Basic usage

Consider the following DataFrame:


        
        
            
                
                
                    import pandas as pd
df = pd.DataFrame({'A':['k23','45k','67k89']}, index=['a','b','c'])
df
                
            
               A
a  k23
b  45k
c  67k89

To get extract substrings that match a given regex:


        
        
            
                
                
                    df['A'].str.extractall('(\d+)')   # returns a multi-index DataFrame
                
            
                      0
match  
a      0  23
b      0  45
c      0  67
       1  89

Here, the input string is a regex, and \d+ indicates a number, while () indicates the portion we want to extract.

Since the resulting DataFrame is a multi-index, we can obtain the matches for specific indexes like so:


        
        
            
                
                
                    df_result = df['A'].str.extractall('(\d+)')
df_result.loc['c']
                
            
                    0
match  
0       67
1       89

Multiple capturing groups

Consider the following DataFrame:


        
        
            
                
                
                    import pandas as pd
df = pd.DataFrame({'A':['k23','45y','67k89']}, index=['a','b','c'])
df
                
            
               A
a  k23
b  45y
c  67k89

We can capture multiple groups using multiple brackets:


        
        
            
                
                
                    df['A'].str.extractall('(\d+)([ky])')
                
            
                        0   1
    match    
b     0     45  y
c     0     67  k