python pandas date read_table

Question

我有以下输入文件:

2012,10,3,AAPL,BUY,200
2012,12,5,AAPL,SELL,200

我如何在以下列中将其读入pandas数据帧:

index: default int range # 0
column1: datetime(2012,10,3,16) # 2012-10-03 16:00
column2: string # AAPL
column3: string # BUY
column4: integer # 200

例:

0 2012-10-03 16:00 AAPL BUY  200
1 2012-12-05 16:00 AAPL SELL 200

尝试过(熊猫0.7):

In[2]: pandas.io.parsers.read_csv("input.csv", parse_dates=[[0,1,2]], header=None)
Out[2]: 
    X.1  X.2  X.3   X.4   X.5  X.6
0  2012   10    3  AAPL   BUY  200
1  2012   12    5  AAPL  SELL  200

Answer 1

尝试使用read_csv()函数.确保您的csv包含标头或传递header=None以进行正确的解析.parse_dates=[[0,1,2]]将促进所需的数据时间解析.

In [4]: pandas.io.parsers.read_csv("input.csv", parse_dates=[[0,1,2]], header=None)
Out[4]: 
              X0_X1_X2    X3    X4   X5
0  2012-10-03 00:00:00  AAPL   BUY  200
1  2012-12-05 00:00:00  AAPL  SELL  200