Pathfinder swing TS

Forums ProRealTime English forum ProOrder support Pathfinder swing TS

Viewing 15 posts - 1,756 through 1,770 (of 2,005 total)
  • #53021

    Looks great O-jay8. Thank you very much.

    #53022

    Do we call this v3.1?

    1 user thanked author for this post.
    #53026

    Careful Dajvop, the 2015 trade is not coming from the sell condition.
    This issue still remains unfortunately. I didnt find out yet what is causing the additional trade in V2.

    But at least all the other trades are identical between V3 and V2 with the sell condition.

    #53028

    If we want to work with that onwards, we could name it V3.1.

    For me the questions remains whether we even should use V3, if you say the results of the periods are affecting each other.
    I mean the only reason for V3 was to create a full year algo with changing parameter or am I wrong?
    But if that is not working correctly why are we using it then? I do still use V2 to be honest.

    If I find time I would take a look again in the code but it is in my case more try and error.
    Otherwise it seems we have to hope Reiner comes around with some of his great ideas again.

    Best regards

    #53029

    @O-jay8

    If you change line 73 in your v2 to OpenDay (as in v3) then you get the same result as v3 with the sell command.

    1 user thanked author for this post.
    #53030

    And if you change that line to “Date – ((CurrentYear * 10000) + CurrentMonth * 100)”  in the original v3 (without the sell command) you get even better results.

    #53031

    Thanks Dajvop,
    great so we solved the discrepancies then.

    Did you use the same parameters?
    If so, That would be an interesting exercise.

    #53032

    Yes, I used DAX nov1 that you uploaded.

    Though now the question is what is “correct” to use. OpenDay or Date – ((CurrentYear * 10000) + CurrentMonth * 100)?

    I tried the change in another system and got worse results. So I think it depends on the optimisation. Some systems may benefit from the change and others not.

    #53040

    That is the question right question. Which one do we use.
    I am off for almost 2 weeks so I cant check it for a while unfortunately.

    Hear from you guys next time 🙂

    #53111

    So we’ll use the code from this one and call it v 3.1? Or did I miss something?

    1. PF-DAX-Nov-V3-RE-Mod-V2.itf
    #53125

    @ozz87 it turns out that it is quite not that simple. The main differences between v2 and v3 is how CurrentDayOfTheMonth is calculated and the additional sell command in v2 that is missing in v3.

    I try each combination for the systems I’m optimising now. So far it is better with CurrentDayOfTheMonth = OpenDay (as in v3) and it rarely matters if the sell command is present or not. Maybe it depends on how you optimise the systems..? There is not a clear answer here it seems.

    #53126

    @dajvop: Hmm ok. Because I got some time to optimize for the next period this weekend, but unclear which one to use then. I changed some things in my latest V3 FTSE and got completely different results if I changed the previous periods numbers, so will not use that one.

    #53136

    Yes, if you change the ThirdMA for the previous period you might change the results of the period you are optimising. Also the maxcandles affect the next period. In the original v3 that is.

    Either use my modified v3 or go back to v2. I am trying to do several periods in a row so I’m sticking with v3 and just accept the side effects for now.

     

    1 user thanked author for this post.
    #53137

    I’ll use V2 for now. Sometimes I don’t load all history, concentrating on the latest years instead (like with CC below).

    Here is Soy and CC.

    3 users thanked author for this post.
    #53142

    AU short. Long on roadmap but didn’t get any good results at all.

    AG

    3 users thanked author for this post.
Viewing 15 posts - 1,756 through 1,770 (of 2,005 total)

Create your free account now and post your request to benefit from the help of the community
Register or Login