Reading Pandas DataFrame

tralaritralara

Greetings,

I am just starting with Backtrader, I think it is a really complete tool, but I am finding some difficulties to get started with it. I am trying to follow the Quickstart Guide in documentation, but I am not even able to load price history. This is my code:

from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import datetime  # For datetime objects
import os.path  # To manage paths
import sys  # To find out the script name (in argv[0])

# Import the backtrader platform
import backtrader as bt

class dataFeed(bt.feeds.PandasData):
    params = (
        ('datetime', None),
        ('open', None),
        ('high', None),
        ('low', None),
        ('close', 1),
        ('volume', None),
        ('openinterest', None)
    )

def data_mining():
    start_date = pd.to_datetime('1970-01-01')
    end_date = pd.to_datetime('2020-12-31')
    auth_tok = '##########'
    data = quandl.get('LBMA/GOLD', trim_start=start_date, trim_end=end_date, authtoken=auth_tok)
    close = data['USD (AM)']
    all_weekdays = pd.date_range(start=start_date, end=end_date, freq='B')  ## B = Business days (L-V)
    close = close.reindex(all_weekdays)
    close = close.fillna(method='ffill')
    close = close.to_frame().rename(columns={'index': 'Date', 'USD (AM)': 'Close'})
    return close  ## With this I return a DataFrame of 1 column, I only want close prices

if __name__ == '__main__':
    # Create a cerebro entity
    cerebro = bt.Cerebro()

    close = data_mining() ## My data in Pandas DataFrame format
    # Create a Data Feed
    data = dataFeed(dataname=close)  ### Class defined above

    # Add the Data Feed to Cerebro
    cerebro.adddata(data)

    # Set our desired cash start
    cerebro.broker.setcash(100000.0)

    # Print out the starting conditions
    print('Starting Portfolio Value: %.2f' % cerebro.broker.getvalue())

    # Run over everything
    cerebro.run()

    # Print out the final result
    print('Final Portfolio Value: %.2f' % cerebro.broker.getvalue())

Whatever I do (I tried many more similar ideas), I get this error:

Traceback (most recent call last):
  File "<input>", line 13, in <module>
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\cerebro.py", line 1127, in run
    runstrat = self.runstrategies(iterstrat)
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\cerebro.py", line 1212, in runstrategies
    data.preload()
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\feed.py", line 688, in preload
    while self.load():
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\feed.py", line 479, in load
    _loadret = self._load()
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\feed.py", line 710, in _load
    return self._loadline(linetokens)
  File "C:\Proyectos\cockroach_portfolio\venv\lib\site-packages\backtrader\feeds\csvgeneric.py", line 148, in _loadline
    csvfield = linetokens[csvidx]
IndexError: list index out of range

I would like to know how to import Pandas data, I am mainly interested only in close prices and large history backtestings.

Thanks in advance

dasch

@tralaritralara try thhis:

class dataFeed(bt.feeds.PandasData):
    params = (
        ('datetime', 0),
        ('open', None),
        ('high', None),
        ('low', None),
        ('close', 1),
        ('volume', None),
        ('openinterest', None)
    )

you may need to define the dtparam for your data. but you need to provide some date time information to backtrader.

tralaritralara

@dasch My datetime is given in index column (according to documentation, that should work), it has %Y-%m-%d format, I will update my OP with a sample

tralaritralara

It seems I cannot edit my posts, so I will post it here. This is a sample of my data, historical prices for Gold obtained from Quandl. As I said before, I am only interested in close prices, so there is no OHLC structure.

              Close
1970-01-01      NaN
1970-01-02    35.12
1970-01-05    35.10
1970-01-06    35.12
1970-01-07    35.00
             ...
2020-12-25  1872.55
2020-12-28  1872.55
2020-12-29  1873.90
2020-12-30  1877.55
2020-12-31  1891.10
[13306 rows x 1 columns]

dasch

@tralaritralara you can reset the index by:

df.reset_index(inplace=True)

before creating the pandas feed, set datetime to 0.

class dataFeed(bt.feeds.PandasData):
    params = (
        ('datetime', 0),
        ('open', None),
        ('high', None),
        ('low', None),
        ('close', 1),
        ('volume', None),
        ('openinterest', None)
    )

run-out

@dasch I haven't tried this, but I think from previous reading you have to set the open/high/low to the close value not None

tralaritralara

@dasch @run-out That made the fix. run-out is right, although dasch's proposal also works out, it will later create other problems as soon as cerebro tools get into action. So it is better to leave all the columns with close prices.

Thanks a lot to both of you :)