將datetime列添加到數字數據幀會強制將值設為數字，因此NaN

Question

我有一個帶有數字條目的數據框，我希望向其中添加一些數字代碼作為字符串，這些數字代碼也被當作數字，盡管我嘗試指定dtype ='str'，並且由於它們看起來不像數字，所以它們變成了NaN：

dates = pd.date_range('20130101', periods=6)
df = pd.DataFrame(np.random.randn(6,4), index=dates, columns=list('ABCD'))
df['codes'] = pd.Series(['3/27', '3/22', '3/23', '3/27', '3/58', '3/29'], dtype='str')
df

產生：

Out[1]: 
                   A         B         C         D codes
2013-01-01 -1.071662  0.322842 -1.364833  1.046144   NaN
2013-01-02 -1.779425  1.403387 -1.603079  2.117234   NaN
2013-01-03 -0.759267 -0.305942  1.310631  0.606185   NaN
2013-01-04  1.610275 -0.681264  0.800195  0.775496   NaN
2013-01-05  1.145720  0.252765 -1.512279 -0.222186   NaN
2013-01-06  1.267579 -1.412583 -0.270927  0.584454   NaN

我該怎么做才能將這些代碼輸入為字符串？

Answer 1

添加index=df.index以使新Series與DataFrame對齊：

df['codes'] = pd.Series(['3/27', '3/22', '3/23', '3/27', '3/58', '3/29'], index=df.index)

或分配列表：

df['codes'] = ['3/27', '3/22', '3/23', '3/27', '3/58', '3/29']