Simple Hull MA screener won't work

Forums ProRealTime English forum ProScreener support Simple Hull MA screener won't work

Viewing 10 posts - 1 through 10 (of 10 total)
  • #100889

    Hi there, I’ve had issues with screeners and using the ‘Average’ command before (apparently ProScreener doesn’t like the second argument in the parenthesis so much…), but thought they would have been fixed by now. Can anyone spot the problem with this (doesn’t return any results on any list of instruments)?

    It seems ProScreener will only return results up to the value of 6 as the MA type…

    Even when I manually calculate the Hull MA I don’t get any results, even though translating to an indicator to obtain e.g. a histogram when the condition is true works just fine:

    Doesn’t fill me with confidence in ProRealTime as a platform if simple things like this can’t be programmed. I know v11 is coming out soon with a designated HullAverage instruction, but still…

    Thanks in advance for your help.

     

    #100923

    I get ProOrder / Auto-Systems working okay for MA[20,7] and MA[20,8] but it seems MA type 7 and 8 are not supposed to work in PRT v 10.3.

    Maybe it is just as a Screener that MA Type 7 and 8 don’t work in V 10.3

    MA type 7 and 8 do work in PRT v 11 … so I am told, as I don’t have v 11.

    #100927

    Install yjis indicator on your chart, to find out if there have been signals before:

     

    #100930

    Yes I did that thanks, Roberto, as alluded to in my post (“even though translating to an indicator to obtain e.g. a histogram when the condition is true works just fine”). I can confirm it works fine as an indicator, yet ProScreener doesn’t pick it up…

    The screener is designed for cryptos so am not so interested in backtesting on ProOrder, GraHal, but thanks for your observation.

    #100931

    With HULL periods = 100 it works.

    I also added line 19 and modfied line 20 to make a signal valid for longer than just one bar:

     

    #100934

    Thanks for that, Roberto. A 100-period lookback is not ideal – it should work for all (positive/non-zero) values.

    I have tried the Summation command before to detect signals from previous bars, but for the code I posted above that did not work either.

    #100945

    A 100-period lookback is not ideal

    What myself and Roberto were trying to provide is information to show if Hull MA works in a ProScreener as your original post suspected that Hull MA didnt work in ProScreener.

    Roberto’s code / results proves that HullMA does work in ProScreener.

    #100946

    Yes, but only with certain lookback periods (beyond what is a high initial value). I think most people would regard that as sub-optimal.

    #100948

    I think most people would regard that as sub-optimal.

    Maybe and we take / agree your point,  but Roberto code shows that ProScreener can read the Hull MA type 8 … which your original post read / doubted that HullMA could be used at all at any lookback period / in any way at all in ProScreener?.

    I guess now we are onto the next phase of coding … to get a Screener that gives you want you want (now we know / have proven that Hull MA works in ProScreener).

    #100949

    Thank you, GraHal.

Viewing 10 posts - 1 through 10 (of 10 total)

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