> I have to admit that I modified JBookTrader in such a way that it now > accepts also OHLC data from JSystemTrader and other sources. I have > done this because I liked the many small improvements of JBookTrader > over JSystemTrader but still wanted to have OHLC data. Basically I > have extended the 'MarketSnapshot.java' object by open, high and low > fields. The rest were program changes in 'BacktestFileReader.java', > and consequently small changes in 'Bar.java' and > 'PerformanceChartData.java'. I can't imagine that this has such an > impact but of course it is possible. JBookTrader is tolerant to > missing data (gaps of minutes or even hours) and therefore accepts > 5min data, for instance. That means it accepts JSystemTrader generated > 5min data and performs the optimization and backtests flawlessly with > my modifications, except for the memory problem. >
I'd mention several things: 1. In my opinion, 8 parameters is WAY too many. With such a large parameter space, chances are that the optimizer will find a "profitable" strategy by chance, simply by permutation through the many millions of combinations, and whatever it find would not be very meaningful. 2. Since you've modified JBT, there is a possibility that you've introduced a memory leak of some sort, and that's why you are running out of memory. To verify, try optimizing your strategy using the same data, but with the original, unaltered JBT source code. 3. Another test to run is to run a brute force optimization with your altered code and the same data set. If you set the steps so that the total number of strategies to optimize is still 4800, does it run to completion? -- You received this message because you are subscribed to the Google Groups "JBookTrader" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en.
