How to test string contains elements in list and assign the target element to another column via Pandas

问题

I have a one column list presenting some company names. Some of those names contain the country names (e.g., "China" in "China A1", 'Finland' in "C1 in Finland"). I want to extract their belonging countries based on the company name and a pre-defined list consisted of country names.

The original dataframe df shows like this

    Company name    Country
0   China A1        
1   Australia-A2        
2   Belgium_C1      
3   C1  in  Finland
4   D1  of  Greece
5   E2  for Pakistan

For now, I can only come up with an inefficient method. Here is my code:

country_list = ['China','America','Greece','Pakistan','Finland','Belgium','Japan','British','Australia']
for t in country_list:
    df.loc[df['company name'].contains(t),'country']=t

The result shows like

    Company name    Country
0   China A1        China
1   Australia-A2    Australia   
2   Belgium_C1      Belgium
3   C1  in  Finland Finland
4   D1  of  Greece  Greece
5   E2  for Pakistan Pakistan

I thought that when the country_list contains large amount of elements, i,e, countries, it would be time-consuming via loop method. Is there any simpler method to tackle with my problem?

回答1:

Here's one way using str.extract:

df['Country'] = df['Company name'].str.extract('('+'|'.join(country_list)+')')

       Company name    Country
0          China A1      China
1      Australia-A2  Australia
2        Belgium_C1    Belgium
3   C1  in  Finland    Finland
4    D1  of  Greece     Greece
5  E2  for Pakistan   Pakistan

回答2:

You need series.str.extract() here:

pat = r'({})'.format('|'.join(country_list))
# pat-->'(China|America|Greece|Pakistan|Finland|Belgium|Japan|British|Australia)'
df['Country']=df['Company name'].str.extract(pat, expand=False)

回答3:

Maybe using findall in case you have more than one country name in one cell

df["Company name"].str.findall('|'.join(country_list)).str[0]
Out[758]: 
0        China
1    Australia
2      Belgium
3      Finland
4       Greece
5     Pakistan
Name: Company name, dtype: object

回答4:

Using str.extract with Regex

Ex:

import pandas as pd
country_list = ['China','America','Greece','Pakistan','Finland','Belgium','Japan','British','Australia']

df = pd.read_csv(filename)
df["Country"] = df["Company_name"].str.extract("("+"|".join(country_list)+ ")")
print(df)

Output:

           Company_name    Country
0      China A1              China
1  Australia-A2          Australia
2      Belgium_C1          Belgium
3       C1  in  Finland    Finland
4        D1  of  Greece     Greece
5      E2  for Pakistan   Pakistan

来源：https://stackoverflow.com/questions/56278553/how-to-test-string-contains-elements-in-list-and-assign-the-target-element-to-an