Hi all,

I program XAverage in my TradeStation. However, some strange things happened:
(1) XAverage(Close, 3) is not what the chart shows EMA(3).
(2) Second, XAverage(Close,3)[1] gives me "0", also for [2][3]...

In addition, similar things happen when I want to code for some indicators turning up or turning down:
Average(ADX(25), 3) is ok, but Average(ADX(25), 3)[1] is 0.

I have some training in programing, but I am new to TradeStation syntax.

Can you give me some help insights?

Thank you very much!

OperaMask,

I am not exactly sure what is happening on your end, but this code

 

Code

Plot1(XAverage(Close, 3)); Plot2(XAverage(Close, 3)[1]);

produces two EMAs where the second one is returning the previous bar's value from the first XAverage.

What do you mean with " XAverage(Close, 3) is not what the chart shows EMA(3)."?

Regards,
ABCTG

ABCTG 

OperaMask, I am not exactly sure what is happening on your end, but this code   Code Plot1(XAverage(Close, 3)); Plot2(XAverage(Close, 3)[1]); produces two EMAs where the second one is returning the previous bar's value from the first XAverage. What do you mean with " XAverage(Close, 3) is not what the chart shows EMA(3)."? Regards, ABCTG

I am having an issue that sounds like it's related to OP's:

Using this code to produce a 21 EMA:

 

Code

ema = XAverage(close,21);

I print the variable "ema" on a daily chart, and the value printed doesn't match the value shown on the chart.

I noticed this when trades were being placed that looked wrong, and started troubleshooting with the print function.

Is this an issue with the daily bars, or is it some sort of known issue, or am I just missing something very obvious?

ReximusMaximus,

keep in mind that some indicators like EMAs have a starting point dependency. When you compare strategy calculations to indicator results that use starting point dependent calculations you will want to make sure that they both start on the same bar. By default the "MaxBarsBack" feature might cause them to start at different points and in turn the values on a given bar can be different due to that. Therefore, it can make sense to manually set the indicator to the same "MaxBarsBack" value as the strategy.
You can read more about this here in the Tradestation wiki:

Starting Point Dependence
The surprising MACD indicator
Understanding MaxBarsBack

Regards,

ABCTG

ReximusMaximus 

I am having an issue that sounds like it's related to OP's: Using this code to produce a 21 EMA:   Code ema = XAverage(close,21); I print the variable "ema" on a daily chart, and the value printed doesn't match the value shown on the chart. I noticed this when trades were being placed that looked wrong, and started troubleshooting with the print function. Is this an issue with the daily bars, or is it some sort of known issue, or am I just missing something very obvious?

ABCTG 

keep in mind that some indicators like EMAs have a starting point dependency. When you compare strategy calculations to indicator results that use starting point dependent calculations you will want to make sure that they both start on the same bar.

@ABCTG,

Nails it here. This is one of those TradeStation EasyLanguage XAverage function quirks that trips up a lot of developers, so let me add some context for anyone else hitting this.

The XAverage function calculates EMAs recursively -- each bar's value depends on the previous bar's value. The formula is in effect:

 

Code

XAverage = PreviousXAverage + (2/(Length+1)) * (Price - PreviousXAverage)

Because of this recursive structure, the very first bar where the calculation begins matters. Every subsequent value carries a trace of that starting point. If your strategy and your chart indicator begin calculating on different bars (which is exactly what mismatched MaxBarsBack settings cause), you'll get different values on the same bar -- even with identical parameters.

The fix ABCTG describes is straightforward: manually set MaxBarsBack to the same value in both your strategy and the indicator. In TradeStation, you can do this via Format > General > Max bars study will reference, or programmatically with the SetMaxBarsBack reserved word.

One additional debugging tip: try printing both the current value and

 

Code

XAverage(Close,21)[1]

on the first few bars. If you see zeros or wildly different values early on, that confirms the starting points differ.

Those TradeStation wiki links you shared -- especially the Starting Point Dependence article -- are worth bookmarking. This same issue affects any recursive indicator (MACD, ADX, etc.), not just the xaverage function.

-- Fi

"The bug that drives you crazy for hours is often just two calculations starting their process from different places."