循環遍歷列中的多個值

Question

有人可以告訴我如何循環數據框列中的多個值。

例：

col1    col2
High street qwe.723,qwe.2,qwe.17,qwe.1000,qwe.23
Must street qwe.34,qwe.17,qwe.1000,qwe.23

我想要以下輸出：

High street
qwe.723
High street
qwe.2
High street
qwe.17
High street
qwe.1000
High street
qwe.23

Must street
qwe.34
Must street
qwe.17
Must street
qwe.1000
Must street
qwe.23

我的嘗試：

lines = open('file.txt','r')
for line in lines:
    line=line.strip().split('\t')
    vals=line[1].split(',')
    for val in vals:
        print(line[0],'\n',val)

Answer 1

嘗試這個：

In [136]: df
Out[136]:
          col1                                  col2
0  High street  qwe.723,qwe.2,qwe.17,qwe.1000,qwe.23
1  Must street         qwe.34,qwe.17,qwe.1000,qwe.23

In [137]: df.set_index('col1').col2.str.split(',', expand=True).stack().reset_index(level=1, drop=1).to_frame('col2').reset_index().stack()
     ...:
Out[137]:
0  col1    High street
   col2        qwe.723
1  col1    High street
   col2          qwe.2
2  col1    High street
   col2         qwe.17
3  col1    High street
   col2       qwe.1000
4  col1    High street
   col2         qwe.23
5  col1    Must street
   col2         qwe.34
6  col1    Must street
   col2         qwe.17
7  col1    Must street
   col2       qwe.1000
8  col1    Must street
   col2         qwe.23
dtype: object

我敢肯定必須有更好的方法來做到這一點......

Answer 2

另一個：

(df.set_index('col1')
   .col2.str.split(',', expand=True)
   .stack()
   .reset_index(level=-1, drop=True)
   .to_csv('output.txt',sep='\n')

Answer 3

因為我正在玩cytoolz和numpy
超級快！

import cytoolz

c2 = np.core.defchararray.split(df.col2.values.astype('str'), ',')
col1 = df.col1.values.repeat([len(c) for c in c2.tolist()])
col2 = list(cytoolz.concat(c2))
np.stack([col1, col2]).ravel('F')

array(['High street', 'qwe.723', 'High street', 'qwe.2', 'High street',
       'qwe.17', 'High street', 'qwe.1000', 'High street', 'qwe.23',
       'Must street', 'qwe.34', 'Must street', 'qwe.17', 'Must street',
       'qwe.1000', 'Must street', 'qwe.23'], dtype=object)

---

時間測試