Not logged in
Viewing 9 posts - 1 through 9 (of 9 total)

Forums>StrategyQuant>Application Support>Error with testing on tick data Integral/TrueFX/Pepperstone

  • #113085 |
    Customer
    7 Posts
    Hello,
     
    I have a problem with build/retest strategy on tick data Integral/TrueFX/Pepperstone (downloaded from Pepperstone which redirects to TrueFX website).
     
    When I downloaded individual months for 1 pair, my next step was merging that with using cmd into one .csv file. Firstly I merged years and than years together to final .csv file for specific pair. When I control it with software for such a big files, there was everything ok (about split individual csv).
     
    I also successfully imported this data into SQ. But when I want to retest or start build on this data, I had a problem. In SQ log has been (mainly) this:
     

    21:59 SEVERE EXCEPTION testing strategy Error 1: Times between base and main TF don’t match!! STACK TRACE: java.lang.Exception: Error 1: Times between base and main TF don’t match!!

     

     

     
    I am not sure, what I should do with that. I am using DukasCopy data and that´s fine. Are you using this data and If yes, how did you split it and imported into SQ? Maybe I did some mistake here.
     
    Thank you for your help.
    #128278
    Customer
    7 Posts

    Nobody? I still can not to resolve it. For example, I do not understand, why problems shows when I want to test/build and not in importing.

    #128281
    Mark Fric
    Administrator
    1182 Posts

    it looks like some bug in SQ caused by the way you connected the data.

     

    I’d be glad if you could send me the data for testing (or post a link to download them here), then I can find out what is wrong.

    Mark
    StrategyQuant architect

    #128296
    Customer
    7 Posts

    Here are direct links for storage on TrueFX website, where you have to registrate for download.

     

    https://pepperstone.com/mt4-forex-trading/mt4-tick-chart-history-data.php

     

    I would be glad for you help, I spent a lot of time working on it..

    #128300
    Mark Fric
    Administrator
    1182 Posts

    the problem is most likely caused by the way you connected the data files, could you send or give me link to your data file that causes these problems after import?

     

    Or maybe try to delete the data in SQ and reimport it again, it could possibly solve the problem with testing.

    Mark
    StrategyQuant architect

    #128313
    Customer
    7 Posts

    I have successfully merge 2013 and 2014, these data works correctly in SQ, but 2009,10 and 11 doesn’t work. I wrote to “Patrick” and he confirm my problem. It can not be my mistake while importing data to SQ. I am not sure, what is wrong there. I am mainly interested, what could be fault. You can download data without registration on Pepperstone website (link above). Data for 2013 and 14 are relatively ok for me, but there can be some bug in SQ data manager, so, it is good to know about it.

    #128382
    Customer
    4 Posts

    Timestamp is confused

    #128897
    Mark Fric
    Administrator
    1182 Posts

    stepni – if you have problems with 2009 data specifically, could you send me your file?

     

    I can download it from their website and merge it, but I might merge it in a different way than you. If possible can you use some sharing site to send me your data file?

    Mark
    StrategyQuant architect

    #129381
    Customer
    432 Posts

    To reproduce a problem there is no way around than to send Mark your EXACT data file that YOU are using and which gives you this problem.

     

    Also be aware that the data from the TrueFX website is full of holes and errors. I´ve played with it in the past via Birts Tick Data Suite and it´s just not good data. The timestamps are wrong sometimes and there are holes of weeks in the data as well, especially in the years < 2011. In my opinion their data is completely useless and it´s clear it would generate such errors in SQ as timestamps are sometimes completely wrong in their data, simple as that.

     

    You are better off using Dukascopy tick data which is way more quality than that TrueFX crap.

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic.