Hi, I've been using Zorro and love its back test and live trading features.
Zorro currently integrates with R and Python, which is great when you could utilize existing libraries for more advanced numerical/statistical calculations. However, R and Python are slow, so including calls to them would slow down back test.
Julia is a relatively new high-level language that is as easy as Python and Matlab, but much faster, and can be as fast as C/C++ if some basic care is taken. For details see: Julialang.org and Julia Computing
I wonder if there is a possibility to integrate with Julia? Julia itself can easily call C libraries with almost no overhead. Would that help? Or is there some documentation on how R and/or Python is integrated, so we could borrow the approach for Julia as well?
I'm new to Zorro - it looks like a fabulous system.
I'm trying to put up an algo where I sell 0.05 delta puts on the SPY.
I have Workflow 8 as a base and tweaked it to the following. However, I want to exit my put options when the profit hits 50% of initial premium, rather than wait till expiration. How do I do it in the code below?
Thanks a lot.
// Workshop 8: Simple Option system ////////////////////////////////////////// #include <contract.c>
Hi, I agree with MatPed, I think shortening the optimization OOS period may likely lead to overfitting and undermining the basic idea of the edge. For some reason though, the Z strategies are re-parametrized periodically, in terms of months, so I am still probably missing some parametrization fundaments in Zorro :-(.
Just my approach is to stay with the simplest original result, as you posted on 29.9.2019. Especially when you say it is consistent over many assets. I trade it small with microlots and focus to adopt other strategies into the ensemble, to balance/compensate the flat periods and draw downs.
not sure if this is a coding question or a general question about money management and portfolio setup.
For the latter case, I have used the booklet "Embrace the Mayhem" from the Robotwealth team. The link is on the Zorro support page.
The abstract of their approach: - besides the profit, one should always consider the annualized volatility of the portfolio, and keep it well below 15%. - trade small, trade broad - details explained there.
There certainly is a possibility to code in Zorro and compare the performance reports. I leave the space here to more experienced members though.
You misunderstood my approach: I wouldn't cache the signals (which is of course impossible because of the virtually unlimited number of combinations), but the predictions. The predicted value only depends on time, because the signals themselves only depend on time, unless you try to optimize parameters on which they depend. Is that even possible? At least for me, this isn't desirable. If it is possible, you could implement a flag DEEPCACHE or something, to switch caching on.
So if I was to implement that cache, I would pass a timestamp - a bar number or wdate() - along with the signals into the R session. So one of the columns of my signals.csv would be that timestamp. I would put that aside before my neural net gets trained. When the training is finished, I would first add a new column with my predictions to the signals table, then add the timestamp column. Then I would export only the timestamp and the predictions column to a file. So I would get a timeseries with all the predictions which I would read out from zorro.
Maybe the term "cache" doesn't describe the approach. It's more some pre-calculated time series.