Hi,

This is extremely helpful. The issue I am having is that the pickle file saves only when my data classes are derived from bt.feeds.PandasData, but when the classes are derived from bt.feeds.PandasDirectData (to speed up first time loading) I get the following error. Any insights?

_pickle.PicklingError: Can't pickle <class 'pandas.core.frame.Pandas'>: attribute lookup Pandas on pandas.core.frame failed

Thanks,

AP

@rajanprabu Thannks for the comments! Would love to hear thoughts from other BT users.

@rajanprabu Thanks! I am assuming you are using the underlying and the related options as separate custom datafeeds, and using signals from the underlying to make trades on the options. The part where I am looking for some guidance -

1. If each option (say 5 ITM, 5 OTM for each expiry month) has to be treated as separate datafeeds for backtesting purpose, this could mean you are loading 120 data feeds per backtested year (10 options/expiry month * 12 months), or 600 for 5 years. That's a whole lot to load and process, unless there's a simpler way to feed the data. Not sure if chaining/rollovers would simplify things, but it could complicate the situations as there could be multiple options at same strike price for separate expiry months.
2. Can the volatility, greeks be fed as self-defined attributes within custom options datafeeds?
3. What will be the best way to tag the options relative to the underlying for selection purpose? In QuantConnect +1, 2, 3 etc let's you select option that are 1, 2, 3 strikes away from current price. Thinking if anything better can be set up here.

Again Thanks in advance!

AP

Hi,
I have been using Backtrader, mostly for Forex trading strategy backtesting, but want to use it for Options strategy backtesting (underlying stock, with various combinations of put/call options, selected based on the underlying stock). I tried QuantConnect but didn't like the fact that everything is remote/cloud-based and local setup was complicated. I know that Backtrader is not designed for this purpose but some of you have already been using it, so would really appreciate some pointers to start off with (things that worked, things to avoid) rather than me trying things from scratch.

Thanks in advance,
AP

@ab_trader
Thanks a lot!!

@ab_trader
Excellent idea! Any idea on how to extend the datafeed with the indicators, which are line objects. Basically, how to convert the line objects back to datafeed (csv or DataFrane), aligned by date?

Thanks!

Hi,

I am optimizing a strategy where the optimization is focused on identifying the correct exit strategy, and the entry is the same for every optimization run with the same set of indicators/signals. In other words, the difference in one run from the other is the exit strategy but the indicators/signals are unchanged and so are the entries. My questions - is there any way to run the indicators only once and reuse them for every optimization runs, rather than doing it separately for every run.

Thanks,

AP