<p>我认为需要<a href="http://pandas.pydata.org/pandas-docs/stable/generated/pandas.Series.str.split.html" rel="nofollow noreferrer">^{<cd1>}</a>按<code>str[0]</code>选择第一列<code>list</code>或按<code>[0]</code>选择第一列:</p>
<pre><code>df['new'] = df['location.display_name'].str.split(',').str[0]
#alternative
#df['new'] = df['location.display_name'].str.split(',', expand=True)[0]
print (df)
location.display_name new
0 Kelso, Scottish Borders Kelso
1 Manchester, Greater Manchester Manchester
2 Northampton, Northamptonshire Northampton
3 Reading, Berkshire Reading
4 Leicester, Leicestershire Leicester
5 Newport, Wales Newport
6 Swindon, Wiltshire Swindon
7 Perth, Perth & Kinross Perth
8 Manchester, Greater Manchester Manchester
9 Perth, Perth & Kinross Perth
10 Cardiff Cardiff
11 Hull, East Riding Of Yorkshire Hull
12 Chester, Cheshire Chester
13 Southampton Southampton
14 Leamington Spa, Warwickshire Leamington Spa
15 Swindon, Wiltshire Swindon
16 Slough, Berkshire Slough
17 Portsmouth, Hampshire Portsmouth
</code></pre>
<p>如果数据中没有<code>NaN</code>s和<code>None</code>s,则可以使用<code>list comprehension</code>:</p>
<pre><code>df['new'] = [x.split(',')[0] for x in df['location.display_name']]
</code></pre>