I don't know how accurate this is, but this ended up working:

import backtrader as bt
import pandas_datareader as web
...
feed = web.DataReader("AAPL", "google", datetime(2016, 1, 1), datetime(2017, 1, 1))
data = bt.feeds.PandasData(dataname=feed, name="AAPL")
cerebro.adddata(data)
cerebro.run()
cerebro.plot(style='candlestick', barup='green', bardown='red')

The graph looks as expected. Not sure if the data is exactly correct though.

I don't know how accurate this is, but this ended up working:

import backtrader as bt
import pandas_datareader as web
...
feed = web.DataReader("AAPL", "google", datetime(2016, 1, 1), datetime(2017, 1, 1))
data = bt.feeds.PandasData(dataname=feed, name="AAPL")
cerebro.adddata(data)
cerebro.run()
cerebro.plot(style='candlestick', barup='green', bardown='red')

The graph looks as expected. Not sure if the data is exactly correct though.

Oh wow, thanks for the feedback. That post certainly has a lot of information. I'm a bit confused though, getting this data is an absolute critical part to be able to perform meaningful backtesting. How is everyone getting their data? Surely everyone isn't making one-off modifications, are they? I have to assume I'm being dumb here and there is a common source/service/method that everyone uses to get reliable, meaningful data.

Hi everyone,

I'm doing the most basic examples from the README page and simple tutorials online, and I'm plotting my graph the following way:

cerebro.plot(style='candlestick')

or

cerebro.plot(style='candlestick', barup='green', bardown='red')

However, my graph looks almost entirely red when plotted:

Verbatim copy/paste examples are showing the above type of output, but blog posts show red/green gray/green mixtures (as it should). Any thoughts on what is going wrong?

I appreciate the help, thanks!