I was wondering if anyone could help interpret two lines of ThinkScript code from the VSA indicator.

The first is:

rec avgVolume = CompoundValue(volumeEMALength, ExpAverage(volume, volumeEMALength), Double.NaN)

and the second:

rec sAvgVolume = CompoundValue(volumeEMALength, Average(volume, volumeEMALength), Double.NaN);


I initially thought that the first one was simply an EMA of the volume, but the presence of the CompoundValue(volumeEMALength, is throwing me off.

The second appeared to be just the simple average of the volume, but again, I am having a hard time understanding the beginning.



Could someone shed some light on what these two lines of code are actually measuring and the logic behind it.


I checked the TOS website and am having hard time making sense of the explantion in regard to compound value and this specific code.

Thank you in advance for the help

Hey,

Did you figure out what this compoundvalue function does and why its being used - whats any other similar function in other languages (Python,c#) if any...

BTR411,

from the code snippets it appears the CompoundValue should make sure that the averages have enough bars present before they are used.
If the bar number on the chart is greater than the value for volumeEMALength CompoundValue would return the value for ExpAverage(volume, volumeEMALength) otherwise it would return Double.NaN.

Additionally the link below might help clarifying things for you:
https://toslc.thinkorswim.com/center/reference/thinkScript/Functions/Others/CompoundValue.html

Regards,

ABCTG

ABCTG 

from the code snippets it appears the CompoundValue should make sure that the averages have enough bars present before they are used. If the bar number on the chart is greater than the value for volumeEMALength CompoundValue would return the value for ExpAverage(volume, volumeEMALength) otherwise it would return Double.NaN.

@ABCTG,

Solid explanation -- you nailed the core concept here. The "cold start" problem trips up a lot of folks writing custom ThinkScript indicators, and CompoundValue is exactly how you handle it cleanly.

For anyone following along: when you're calculating something like an EMA, you need X bars of history before the math even makes sense. Without CompoundValue, your indicator either throws errors or plots garbage values on those early bars. By returning Double.NaN when bar_number is less than your required length, you get clean charts instead of misleading data points.

The three-parameter structure is worth highlighting:

• First param: the lookback length needed
• Second param: what to return when you have enough data
• Third param: the fallback (usually Double.NaN)

This pattern becomes essential once you start building recursive calculations or any indicator that references its own previous values. The official thinkorswim documentation you linked covers the basics well.

Good looking out for BTR411 -- this kind of knowledge sharing is what makes the community work.

-- Fi
"The best code doesn't just calculate correctly -- it fails gracefully when conditions aren't right."