Strategy Robustness Tester

Viewing 15 posts - 91 through 105 (of 162 total)
  • Author
    Posts
  • #114446 quote
    Paul
    Participant
    Master

    Looks like tony has some nice advanced ideas but I will make it work for myself and post the excel sheet.

    #114447 quote
    Tony87
    Participant
    Junior

    Thanks Paul, for me it’s on the back burner at the moment, but I do plan to work on it some more in the future, I think the main advantage is, to easily establish a robustness baseline that can be compared to when adapting a strategy, I will also update if I make progress on it

    #114460 quote
    Paul
    Participant
    Master

    Looking forward to your improvements, it sounds interesting!

    They way I compare often, on a mac, is taking screenshots, which I can open at the same time to quickly compare. But that’s basic!

     

    Here’s what I got so far.

    I had an easy macro included, but it’s not allowed to be uploaded so it’s removed.

    Don’t take the results for granted. It needs checking.

    #114470 quote
    Vonasi
    Moderator
    Master

    Perhaps we should also be calculating the average win rate from all the tests and the average gain per trade from all the tests as this gives us a truer idea of the performance of a strategy based on all the tests. Perhaps giving some weighting to tests that have more trades in and less weighting to those with only a few trades in would be a good idea when calculating these averages.

    #114528 quote
    Paul
    Participant
    Master

    a bit more to play with, but looking at these figures it’s incredibly boring!

    Didn’t do direct weighting to it but it could be done on this basis.

    Vonasi thanked this post
    #114531 quote
    Paul
    Participant
    Master

    tried another system with less trades.

    splitting it on average trades in 3 ways has it flaws because of div/0 error. Anyway.

    #114587 quote
    Paul
    Participant
    Master

    fixed the div/0.

    even when not all sheets are filled, all is calculated correctly

    report sheet is locked to prevent accidental changes, password is in cell a1

    GraHal thanked this post
    #114611 quote
    Paul
    Participant
    Master

    final one p6

    if tested with  2-10 1-5, the 45 test results can also be copied. The graphs aren’t dynamic and are stuck with 100 tests so lines go half way.

    2 reports, added one extra, with one split less. Probably better for strategies with fewer trades.

    Bard and Dow Jones thanked this post
    #117165 quote
    nonetheless
    Participant
    Master

    Hello all, apologies for being dense — I have the test results but how do I get the optimize report into Excel?

    Yes, I did read the idiot’s guide on page 1 but, drag and drop what exactly? Nothing moves and cant select for copy/paste. I don’t geddit.

    #117169 quote
    Vonasi
    Moderator
    Master

    Have the optimised results window open on the same screen as Excel. Click anywhere on the optimised results window and hold the mouse button and drag it into the Excel data page and then release the mouse button. The optimised results window contents will appear in the Excel page.

    nonetheless thanked this post
    #117172 quote
    nonetheless
    Participant
    Master

    Yeah, tried that about a hundred times … nada. Nothing moves. Nothing copies over to Excel.

    #117175 quote
    Vonasi
    Moderator
    Master

    Sometimes I find that Excel disappears when I click and hold on the results window but if I drag down to the Excel item on the task bar at the bottom of the screen (Windows 10) then Excel suddenly appears again.

    Perhaps you should try shutting down your platform and starting it again as switch it off and on again often fixes things!

    #117185 quote
    Paul
    Participant
    Master

    ran a test on the dow with 1 hour timeframe and spread 2.4 200k bars

    GraHal thanked this post
    #117188 quote
    Paul
    Participant
    Master

    another one, dow with 5 min timeframe and spread 2.4 200k bars

    GraHal thanked this post
    #117207 quote
    nonetheless
    Participant
    Master

    as switch it off and on again often fixes things

    Primary tech advice – see image 1! And yes, seems to have done the trick, thanks.

    So … what counts as a good result here? and which do you consider as the key metric – Average Gain, % Win or Average of the two?

    Do you regard this as superior to WF analysis, or to be used in conjunction with?

    @Paul, nice work on the VRT template, dead easy to import data! 👍

    Paul thanked this post
Viewing 15 posts - 91 through 105 (of 162 total)
  • You must be logged in to reply to this topic.

Strategy Robustness Tester


ProOrder support

New Reply
Author
author-avatar
Vonasi @vonasi Moderator
Summary

This topic contains 161 replies,
has 20 voices, and was last updated by Roger
5 years ago.

Topic Details
Forum: ProOrder support
Language: English
Started: 09/01/2019
Status: Active
Attachments: 94 files
Logo Logo
Loading...