I'm attempting to code up a Signal Noise Ratio as seen in Trend Following with Managed Futures by Greyserman and Kaminski

The formula is essentially:

Abs((Price(t) - Price(n)) / Sum of Abs(Price(t-k) - Price (t-k-1)

An example if N = 100 I'm looking at the $ price change from today vs 100 days ago divided by the sum of all the absolute price changes for the same period.

I made an indicator with a parameter of lookbackdays for N then have two indicators

How did you go with this? I'm getting a divide by zero error when calculating the indicator (I assume due to an instance of zero price movement over the lookback period for an instrument).

I coded in :

IF absoluteValue(instrument.close-instrument.close[atrLookBackNoise]) =0 then
noisePriceMovement = -1
else
noisePriceMovement = absoluteValue(instrument.close-instrument.close[atrLookBackNoise])

noisePriceMovement obviously being the denominator of the indicator expression.

It contains the Perry Kaufman's KAMA Indicator. At the heart of his indicator is a noise calculation he uses to make his KAMA indicator adjust the period length of his average calculation. His indicator seems to be the same as what is posted in this thread.

Blox has a lot of comments. If you find something causes a problem look in the TBB Help file to see if any of the names have changed. If that doesn't work, post your issue and I'll take a look.