Hello,
I have a simulator for trading and it's results are multiply parameters. I need a software to clear my eyes and give me an answer which combinations of parameters will be good in the future for trading.
I have (attached a sample) a csv file (could get to few millions lines) of buy/sell orders (column 3) each on a separate line (the program needs to match based on date+time which is the start of the position and which is the end of the position and to calculate the profit/lose), with price (column 5) and date (column 8) and time (column 9) and the parameters i used to generate them (column 1 with 8 parameters with "_" separator).
each parameter (permutation) combination's profit (or loss) should be aggregated per day and than use that pn/l to predict the next time frame best permutation to use. also need to take off commissions (which are not constant).
other parameters to use for the prediction will be:
1. profit/loss (as mentioned above) for the time frame
2. standard deviation of 1
3. draw down of 1
4. # of trade in the time frame
5. profit factor
6. parameters used in column 1
7. more ideas will be welcome (and considered before choosing a freelancer)
All the above csv file is both long and short trade. it needs to be separated (upon check box ) for long best combination prediction and for short trades best combination prediction.
Also (upon check box) a prediction should be sliced upon time brackets: 9:30-10:00, 10:00-10:30, 10:30-11:00, 11:00-11:30, 11:30-12:00, 12:00-12:30, 12:30-13:00, 13:00-13:30, 13:30-14:00, 14:00-14:30, 14:30-15:00, 15:00-15:30, 15:30-16:00 (and i need the ability to change this #s if trading another session). so each day/week/2 weeks/ any time frame I can have 13 combinations / parameters that will be changed along the day.
If both time brackets and long/short are checked i should get 26 combinations per time frame.
The time frame prediction can't be less than a day (i think it's should be even 1 week) and can range from 1 day to 3 months to predict the next day/week/2 weeks/etc.
The program need to accumulate the prediction results of the out of sample (predicted time frame) and to plot the accumulated pn/l for the all periods in order to test the prediction and now how it would have done on that time frame.
The program need to plot statistics sheet (common avg, min, max etc. ) for the combined out of sample periods
The program need to give me a single answer (prediction) which combination of parameters will be the best in the next time frame (and if needed for long/short and for 30 min. during the day)
Full documentation inside the code needed as well.
Can be VB/C++/delphi/c/.net/what ever as long as it a stand alone exe file