Playback Question

[mikepugliano]

Gentlemen,
I have a question. If I download a tic file and play it back and compare the close price of every minute on a 60min candle to the close of the same candle as it was forming live, I get different results. The closing prices have some varying differences. In other words the playback prices should be the same as the live chart; I would believe that should be the case anyway. If you can help me I would appreciate it. Thank You, Brad

[eSignal_AveryH]

Brad,

We're not clear on how you're preforming the test. Can you please provide the steps taken?

Thanks in Advance,

[mikepugliano]

Hi and thanks for getting back to me, I used the following program, first running it on a live GBPJPY chart, and then I downloaded tics for the GBPJPY for one day and ran the program on it. I then brought the results of the two text files created into Excel and compared them. That’s when I saw the differences. Should the prices not be the same for the same period in time?



function preMain() {

setPriceStudy(true);

setStudyTitle("Two");

setShowCursorLabel(false);

}

var mm = null;

var ss = null;

var oldMin = null;

var txtFlags = Text.ONTOP | Text.RELATIVETOLEFT | Text.RELATIVETOBOTTOM;



function main() {

var gDate = getDay();

var hh = hour();

var mm = minute(inv(1));

var ss = second(inv("1s"));



if(getMonth(0) == 4 && gDate == 20 && hh == 13){

if(oldMin != mm){

if(getSymbol() == "$PLAYBACK") Rep = "PlayBack";

if(getSymbol() != "$PLAYBACK") Rep = "RealTime";

timeNow = hh + ":" + mm + ":" + ss;

price = close();

Report(timeNow,price);

debugPrintln("At " + hh + ":" + mm + ":" + ss + ", close() is " + close() + "\n");

oldMin = mm;

}

}

}



function Report(timeNow,price){

fpp = new File(Rep+".report");

if(!fpp.exists()){

fpp.open("at+")

fpp.writeln("Time" + "\tPrice");

fpp.close();

}

fpp.open("at+")

fpp.writeln(timeNow + "\t"+price);

fpp.close();

return;

}

[eSignal_AveryH]

Before we get involved with EFS we first have to determine if this is a data issue. What I would suggest is removing the EFS and run the chart for the day. At the end of the day export the data from the chart using the Tools-->Data Export.

Then run the Tick Downloader and run the file. Once completed run the Data Export again and compare the two files. I will do the same and if we can see differences in the two files I'll report it to our Data QA to further investigate.

If we do not see any differences we'll then consider the next step of EFS.

[mikepugliano]

Avery, I did as you suggested and there was no difference in the files. I even modified my playback script to report high(-1), low(-1), close(-1) and open(-1) of each previous candle on a NEWBAR for the day and they also matched. So it seems there might be a problem with how my program is reporting close() when running playback versus live every minute on a sixty minute candle and trying to get close() every minute where it matches playback. Would you have any ideas on this ? Thank you, Brad

[eSignal_AveryH]

Brad,

Thank for the information, at this point for better support on EFS I've moved the post to the EFS Studies Forum where someone there should be able to assist you further.