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

Alpha Vantage DataFeed

  • M

    Hello,

    I'm pretty new to Python coding and Backtrader.
    However I have done some backtests using the Alpaca API and it worked without to much problems.

    I would to use the Alpha-Vantage API because they provide not only stock data, but also FOREX and Crypto.

    First I tried the Code I found here:
    https://backtest-rookies.com/2019/01/04/backtrader-alpha-vantage-data-direct-ingest/

    from alpha_vantage.timeseries import TimeSeries
import pandas as pd
import numpy as np
import backtrader as bt
from datetime import datetime

# IMPORTANT!
# ----------
# Register for an API at:
# https://www.alphavantage.co/support/#api-key
# Then insert it here.
Apikey = 'XXXXX'


def alpha_vantage_eod(symbol_list, compact=False, debug=False, *args, **kwargs):
    '''
    Helper function to download Alpha Vantage Data.

    This will return a nested list with each entry containing:
        [0] pandas dataframe
        [1] the name of the feed.
    '''
    data_list = list()

    size = 'compact' if compact else 'full'

    for symbol in symbol_list:

        if debug:
            print('Downloading: {}, Size: {}'.format(symbol, size))

        # Submit our API and create a session
        alpha_ts = TimeSeries(key=Apikey, output_format='pandas')

        data, meta_data = alpha_ts.get_daily(symbol=symbol, outputsize=size)

        #Convert the index to datetime.
        data.index = pd.to_datetime(data.index)
        data.columns = ['Open', 'High', 'Low', 'Close','Volume']

        if debug:
            print(data)

        data_list.append((data, symbol))

    return data_list

class TestStrategy(bt.Strategy):

    def __init__(self):

        pass

    def next(self):

        for i, d in enumerate(self.datas):

            bar = len(d)
            dt = d.datetime.datetime()
            dn = d._name
            o = d.open[0]
            h = d.high[0]
            l = d.low[0]
            c = d.close[0]
            v = d.volume[0]


            print('{} Bar: {} | {} | O: {} H: {} L: {} C: {} V:{}'.format(dt, bar,dn,o,h,l,c,v))


# Create an instance of cerebro
cerebro = bt.Cerebro()

# Add our strategy
cerebro.addstrategy(TestStrategy)

# Download our data from Alpha Vantage.
symbol_list = ['LGEN.L','LLOY.L']
data_list = alpha_vantage_eod(
                symbol_list,
                compact=False,
                debug=False)

for i in range(len(data_list)):

    data = bt.feeds.PandasData(
                dataname=data_list[i][0], # This is the Pandas DataFrame
                name=data_list[i][1], # This is the symbol
                timeframe=bt.TimeFrame.Days,
                compression=1,
                fromdate=datetime(2018,1,1),
                todate=datetime(2019,1,1)
                )

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

print('Starting to run')
# Run the strategy
cerebro.run()

    Unfortunately it didn't work. I wasn't able to display any data with it.
    And yes, I used my API - key (which is working, i tested it in another code)

    I thought that the problem came from the Pandas dataframe from Alpha-Vantage (maybe not supported anymore)

    I tried different things and I done quiet a lot of research in the internet and I can't find any suitable solution.

    I also tried to request JSON datas and convert them to pandas or CSV, unfortunately this was also not working.

    But maybe someone can help with the code above.

    Thanks a lot in adavance
    marketwizard

    0
  • A

    I've developed alphavantage data feed to bt some time ago. Not sure if it works now, you can try -
    https://github.com/ab-trader/backtrader_addons

    0
  • M

    Hello ab_trader,

    thanks for the quick reply !
    did you have a sample code for using this datafeed ?

    0
    A
     1 Reply
  • R

    https://backtest-rookies.com/2019/01/04/backtrader-alpha-vantage-data-direct-ingest/

    this could help..

    0
  • M

    @rajanprabu I already test this code and its not working for me, that's why I opened this topic

    0
  • R

    I'm sorry for the oversight.

    0
  • A

    @marketwizard there is tests folder in the repo, which has examples.

    1
Log in to reply
 
});