Indicators and Other-Timeframe Data

[BlueHorseshoe 10]

Hello all,

 

Any suggestions as to how I can correctly calculate indicator values using Close of Data2, where Data2 is of a different timeframe to Data1? I've even tried introducing the Data2 reference into the EL Function, but to no avail . . .

 

Many thanks for any help you're able to give,

 

Bluehorseshoe

[Tams 20]

Hello all,

 

Any suggestions as to how I can correctly calculate indicator values using Close of Data2, where Data2 is of a different timeframe to Data1? I've even tried introducing the Data2 reference into the EL Function, but to no avail . . .

 

Many thanks for any help you're able to give,

 

Bluehorseshoe

 

what have you tried?

how does it not work?

[onesmith 10]

Data1 must be smallest timeframe.

Data2 must be equal or greater than data1 timeframe.

 

After the indicator is inserted in a chart .. Format indicator, General tab, Base study on data2

but you might have to reconfigure the scaling tab depending on how you want to display the result.

 

if you're specifing the data2 hooks within the code then leave the base_indicator_on setting at it's default data1.

 

// both examples produce identical results

value1=stochastic(H,L,C,14,3,3,1,oFastK,oFastD,oSlowK,oSlowD) data2; // data 2 outside parens at function level

value1=stochastic(H data2,L data2,C data2,14,3,3,1,oFastK,oFastD,oSlowK,oSlowD); // data2 inside parens at HLC

 

Series functions sometimes have complications that simple functions don't. I remember being told it was necessary to explicitly declare data2 vars but I don't recall how that's done and never understood why it was necessary because I never had a problem skipping that step.

 

BTW, like Tams says, the more info you provide the more likely you will get the right answer the first time you ask.

[BlueHorseshoe 10]

It's taken me a while to get around to this, but here is some more information about this 'problem'.

 

Data2 Indicators.bmp

 

The screenshot above shows two charts.

 

The one on the right has Data1=60min. The indicator plotted beneath it is a 2-period RSI.

 

The chart on the left has Data1=15min, Data2=60min. The indicator plotted beneath it is a 2-period RSI, with the input set to 'Close of Data2'.

 

The two indicators are clearly displaying differently, and the one on the left hand chart is clearly wrong.

 

So what I am hoping for is a suggestion as to how this (or any other) indicator could be correctly displayed using the close of Data2. Hopefully now I've been a bit more specific you'll be able to help me more easily.

 

Many thanks,

 

BlueHorseshoe

[onesmith 10]

Your picture did not post but from your description I sense the cause of the problem.

 

Format indicator,

 

Input tab, Price should be close (not close data2).

General tab, Base study on, .. Select the 2nd choice, ie 2 (not 1).

[BlueHorseshoe 10]

Your picture did not post but from your description I sense the cause of the problem.

 

Format indicator,

 

Input tab, Price should be close (not close data2).

General tab, Base study on, .. Select the 2nd choice, ie 2 (not 1).

 

Hi Onesmith,

 

I'll give this a go, thanks. Not sure why the image isn't appearing as I can see it within the thread without any problems.

 

How would I overcome this same problem in terms of strategy code and backtesting?

 

Cheers,

 

BlueHorseshoe

[onesmith 10]

Before coding a strategy .. code it as indicator to verify everything is calibrated correctly. Leave the indicator properties set to the default (base study on data1). The references to data2 are within the code.

 

Uncommenting plot1 demonstrates data1 rsi doesn't interfer with data2 rsi.

Barstatus(2)=2 isn't necessary within the context of your question but it might be helpful.

 

var: d1rsi(0), d2rsi(0), bs2.d2rsi(0);

d1rsi=rsi(c,2);
d2rsi=rsi(c data2,2);

// plot1(d1rsi,"d1rsi");
plot2(d2rsi,"d2rsi");

if barstatus(2)=2 then begin
   bs2.d2rsi=rsi(c data2,2);
   plot3(bs2.d2rsi,"bs2.d2rsi");
end;

[BlueHorseshoe 10]

Before coding a strategy .. code it as indicator to verify everything is calibrated correctly. Leave the indicator properties set to the default (base study on data1). The references to data2 are within the code.

 

Uncommenting plot1 demonstrates data1 rsi doesn't interfer with data2 rsi.

Barstatus(2)=2 isn't necessary within the context of your question but it might be helpful.

 

 

Hi Onesmith,

 

That seems to have done the trick - thanks again!

 

BlueHorseshoe