Cancelling pending orders

Forums ProRealTime English forum ProOrder support Cancelling pending orders

Viewing 4 posts - 16 through 19 (of 19 total)
  • #148736

    Replace line 4 with

    Once C1=1

    After line 264, add line

    C1=0

    What this will do is that for the first candle, C1 being true will place a long order. However, for second candle, C1 being false will not place order but first order should be cancelled automatically. This will explain how the candle loops work.

    You are facing this confusion because you have hardcoded the condition to be ALWAYS TRUE regardless of what happens. Your strategy need to update your conditions to be “true” or “false” based on some criteria (whatever criteria you chose) and it will work like a breeze.

     

    Rgds, kaq

    2 users thanked author for this post.
    #148738

    Thank you very much Kaq. Will try…

    #148803

    @kaq52 @juanj @Nicolas @martentornquist @robertogozzi @Bard

    Dear All, thank you very much for your support. The idea suggested by kaq52 is very much what I was looking for. For what it’s worth, I’m sharing here my code and the result of the backtest. Please bear with me if you find erros or room for improvement.

    I’m not proficient in PRT, but I tired to assemble a few pieces here and there found on this forum. I used ML based on code published by Bard here https://www.prorealcode.com/reply/128486/

    Variables have obviously been over-optimized for DAX 15 minutes 1€ with IG CFDs. WF provides a relatively high result… Not sure how this will work in live trading.

    I’d apprecite if some of you can test it (live if possible) and share their feedback. I’ve doubts about:

    1/ the price distance between certain orders calculated by the system vs what IG accepts.

    2/ the interaction between Stop Loss calculated by this system, Spread and actual execution price… I’ve seen that for orders with nearly breakeven results in backtest I ended losing 1 or 2 euros in actual trading.

    3/ overnight holding cost.

    4 /not sure how these impressive Backtest results will translate in reality over a few weeks

    4/ Last but not least… I tried ML with one and with two variables on a very simple SMA cross Algo and I’ve observed on different TF that ML with only one variable provides better results than ML with two variables… I’m very interested in your thoughts on this.

    Next steps: add conditions/optimize entry points to reduce false signals, try wider SL (?),

    Again, thank you and look forward to hear your feedback.

     

     

    #210671

    I’d love to try it, but where is KST1 Indicator?

    KST1 is not the standard Indicator as Line 14 (copied below) needs 2 values and the standard KST returns only 1 value.

    indicator7, indicator6 = CALL KST1 //Standard KST Indicator

Viewing 4 posts - 16 through 19 (of 19 total)

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