A confused question about Pandas DataFeed

life

Hi, maybe it is very basic, but I am glad that some one cloud give some tips about it:

when I load the orcl-1995-2014.txt from github datas with command:

class PandasData(feed.DataBase):
    params = (
        ('datetime', None), # here I use "None" to  indicate the  index of dataframe  is datetime
        ('open', 'Open'),
        ('high', 'High'),
        ('low', 'Low'),
        ('close', 'Close'),
        ('volume', 'Volume'),
        ('openinterest', None),
    )
df = pd.read_csv('orcl-1995-2014.txt',
                                skiprows=0,
                                header=0,
                                parse_dates=True,
                                index_col='Date')
data = bt.feeds.PandasData(dataname=df)
# ... and go on ....
cerebro.run( )

in this case above, everything runs well, but if I make a little different from the example code, as below:

class PandasData(feed.DataBase):
    params = (
        ('datetime', "Date"), # here I use "Date" to  indicate the "Date" column  is datetime
        ('open', 'Open'),
        ('high', 'High'),
        ('low', 'Low'),
        ('close', 'Close'),
        ('volume', 'Volume'),
        ('openinterest', None),
    )
df = pd.read_csv('orcl-1995-2014.txt',
                                skiprows=0,
                                header=0,
                                parse_dates=True) # remove index_col='Date'
data = bt.feeds.PandasData(dataname=df)
# ... and go on ....
cerebro.run( )

I make those little change follow the docs, but then, I got this error:

    266 
    267         # convert to float via datetime and store it
--> 268         dt = tstamp.to_pydatetime()
    269         dtnum = date2num(dt)
    270         self.lines.datetime[0] = dtnum

AttributeError: 'int' object has no attribute 'to_pydatetime'

It is very confused for me, It seems the docs maybe out of date or misunderstood?

run-out

@life said in A confused question about Pandas DataFeed:

df = pd.read_csv('orcl-1995-2014.txt',
skiprows=0,
header=0,
parse_dates=True) # remove index_col='Date'

I the second instance tell pandas where to parse dates:

dataframe = pd.read_csv(datapath, header=0, parse_dates=["Date"], index_col=None)

data = bt.feeds.PandasData(
    dataname=dataframe,
    datetime="Date",
    open='Open',
    high='High',
    low='Low',
    close='Close',
    volume='Volume',
    openinterest=None

)