How to do a modification to the StdDev() function?

[YngvaiMalmsteve]

I'm currently making a modification to the StdDev() function, calling it StdDevPlus(). Right now, StdDev() can take one input data series and a period. I want it to take 2 input data series, with the 2nd series being a moving average of my choice which will be used in the calculation of the standard deviation.

The relevant section of code for StdDev() is here:

protected override void OnBarUpdate()
{
if (CurrentBar < 1)
{
Value.Set(0);
sumSeries.Set(Input[0]);
}
else
{
sumSeries.Set(Input[0] + sumSeries[1] - (CurrentBar >= Period ? Input[Period] : 0));
double avg = sumSeries[0] / Math.Min(CurrentBar + 1, Period);
double sum = 0;
for (int barsBack = Math.Min(CurrentBar, Period - 1); barsBack >= 0; barsBack--)
sum += (Input[barsBack] - avg) * (Input[barsBack] - avg);

Value.Set(Math.Sqrt(sum / Math.Min(CurrentBar + 1, Period)));
}
}

The bolded section is the section I want to change. Rather than using the calculated simple average, I want to pass the average from a series (like TEMA) and use that value.

Let's say the second input is called Input2. So the bolded section would now look like this:

sum += (Input[barsBack] - input2[barsBack]) * (Input[barsBack] - input2[barsBack]);

What I can't figure out is to how to have an iDataSeries object passed as input2 as an argument in the function and then use it in this manner.


Here's an example of MLQ4 code that does this and I'm trying to do the 2nd option:


//
//
//
// In the original article the second method is how deviation is
// calculated . It is a questionableble method , but if we apply
// standard calculation the bands tend to be "spiky".That is the
// reason to have the "UseClassicalDeviations" option
//
//
//

double iDeviationPlus(double& array[], double& ma[], int period, int i)
{
double sum = 0;
if (UseClassicalDeviations)
for(int k=0; k<period; k++) sum += (array[i+k]-ma[i]) *(array[i+k]-ma[i]);
else for( k=0; k<period; k++) sum += (array[i+k]-ma[i+k])*(array[i+k]-ma[i+k]);
return(MathSqrt(sum/period));
}

[YngvaiMalmsteve]

I've also tried just to get the standard deviation of the zero lag TEMA indicator (posted in a recent thread) using the StdDev function but I get an error message when I try to compile.

double stdDevValue = StdDev(ZeroLagTEMA, Period)[0];

...gives me two compiling error messages CS1502 and NT1503

Argument 1 cannot convert from 'method group' to NinjaTrader.Data.IDataSeries

[NinjaTrader_CodyB]

Hello,
Thank you for your post.
What you are wanting to do would be possible through using enumerations. We actually have a reference file on creating a moving average indicator that allows you to select different moving average. You would be able to add this into your StdDev indicator. Please see the reference at the following link,

http://ninjatrader.com/support/forum/showthread.php?t=3420

The compiler error that you are receiving is because you will need to define a period for the ZeroLagTEMA and enter in its overloads.

[Harry]

I've also tried just to get the standard deviation of the zero lag TEMA indicator (posted in a recent thread) using the StdDev function but I get an error message when I try to compile.

double stdDevValue = StdDev(ZeroLagTEMA, Period)[0];

...gives me two compiling error messages CS1502 and NT1503

Argument 1 cannot convert from 'method group' to NinjaTrader.Data.IDataSeries

You cannot call the ZerolagTEMA without parameters. The correct syntax would be one of

double stdDevValue = StdDev(ZeroLagTEMA (temaPeriod), stdDevPeriod)[0];
double stdDevValue = StdDev(ZeroLagTEMA (Input, temaPeriod), stdDevPeriod)[0];

where temaPeriod is the period selected for the ZeroLagTEMA, and stdDevPeriod is the period selected for the standard deviation.