Extract numbers from row pandas-CodePudding

I got an CSV file that looks like this.

0	1
33 abcdefg abcdr 5 kjksk 6 jnjs 555	row
34 abcdefg abcdr 8 kjksk hf 8 jnsj 665	row
35 abcdefg abcdr 7 kjksk hffd 9 jsdi 667	row

I want to extract only the third number.

0	1
5	row
8	row
7	row

I use something like:

df[0].str.extract('(\d )')

But I need some help to only extract the third number.

Hope someone can help!

CodePudding user response：

Use str.findall:

df[0] = df[0].str.findall('\d ').str[1]
print(df)

# Output:
   0    1
0  5  row
1  8  row
2  7  row

CodePudding user response：

With your shown samples, please try following regex.

^(?:\S \s ){3}(\S )\s

Online demo for above regex

Explanation:

^(?:\S \s ){3} ##Matching 1 or non-spaces followed by 1 or more spaces in a non-capturing group 3 times.
(\S )          ##Creating capturing group to get all non-spaces values which are required by OP.
\s             ##Matching 1 or more spaces here.