Python - 从包含文本的 pandas 系列中提取数字

Question

这个问题有一个引用( here )。我对 Python 很陌生，因此陷入了一些琐碎的问题!我有一个数据系列如下

         Text
0        some texts...qualifications: BE year of passing 2012
1        MCOM from XYZ University in 2007. In 2009 he obtained his MBA 
2        Academics: University / Board: XYZ University   year of passing:2014

目标是提取上述年份(仅前几个)，即2012,2007,2014。现在我的做法如下:

corpus = pd.Series('the above series')
corpus = corpus.str.replace(r'^[A-Za-z0-9]+')
corpus = corpus.str.lower()
if corpus.str.contains('qualifications').any():
    corpus.str.extract('.*qualifications.*?(\d{4})', expand = False)
if corpus.str.contains('university').any():
    corpus.str.extract('.*university. *?(d\{4})', expand=False)
if corpus.str.contains('academics').any():
    corpus.str.extract('.*academics. *?(d\{4})',expand=False)

上述方法是创建一个空白系列。请帮助我解决这个问题。

Answer 1

我认为你可以将该表达式简化为:

代码:

corpus = corpus.str.lower().str.extract(
    '(university|academics|qualifications).*?(\d{4})', expand=False)

测试代码:

corpus = pd.Series("""
    some texts...qualifications: BE year of passing 2012
    MCOM from XYZ University in 2007. In 2009 he obtained his MBA 
    Academics: University / Board: XYZ University   year of passing:2014
    """.split('\n')[1:-1], name='Text')

corpus = corpus.str.lower().str.extract(
    '(university|academics|qualifications).*?(\d{4})', expand=False)

print(corpus)

结果:

                0     1
0  qualifications  2012
1      university  2007
2       academics  2014