3. Importing data from pandas_datareader
For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/

Importing data from pandas_datareader

  • C

    Hi all, I am trying to test importing data from pandas_datareader (a spin out of pandas.io) and it throws an error. Anyway I could import those macro economic data into backtrader data stream?

    Here's the code where it throws error:

    import pandas_datareader.data as web

def runstrat():
    cerebro = bt.Cerebro()
    
    start = datetime.datetime(1949,12,31)
    end = datetime.datetime(2017,1,3)
    spx = web.DataReader("^GSPC", 'yahoo', start, end)

    And here's the error output:

    TypeError                                 Traceback (most recent call last)
<ipython-input-55-af84f08c4efa> in <module>()
     58 
     59 if __name__ == '__main__':
---> 60     runstrat()

<ipython-input-55-af84f08c4efa> in runstrat()
     49 
     50 
---> 51     data = bt.feeds.PandasData(spx)
     52     cerebro.adddata(data)
     53     cerebro.broker.setcash(10000.0)

C:\Anaconda3\envs\backtrader\lib\site-packages\backtrader\metabase.py in __call__(cls, *args, **kwargs)
     85         _obj, args, kwargs = cls.donew(*args, **kwargs)
     86         _obj, args, kwargs = cls.dopreinit(_obj, *args, **kwargs)
---> 87         _obj, args, kwargs = cls.doinit(_obj, *args, **kwargs)
     88         _obj, args, kwargs = cls.dopostinit(_obj, *args, **kwargs)
     89         return _obj

C:\Anaconda3\envs\backtrader\lib\site-packages\backtrader\metabase.py in doinit(cls, _obj, *args, **kwargs)
     75 
     76     def doinit(cls, _obj, *args, **kwargs):
---> 77         _obj.__init__(*args, **kwargs)
     78         return _obj, args, kwargs
     79

    pandas-datareader parses the data like this (I've downsample to monthly):

    Open	High	Low	Close	Volume	Adj Close	
Date									
1950-01-31	17.049999	17.049999	17.049999	17.049999	
1950-02-28	17.219999	17.219999	17.219999	17.219999	
1950-03-31	17.290001	17.290001	17.290001	17.290001	
1950-04-30	17.959999	17.959999	17.959999	17.959999	
1950-05-31	18.780001	18.780001	18.780001	18.780001
    0
  • C

    Nevermind, fixed the error by adding 'dataname=' in front of the name of the pandas df.
    data = bt.feeds.PandasData(dataname=spx)

    0