SAR Formula

[nhoak]

Does anyone know how eSignal determines the number of prior candles to use in the SAR calculation?

From what I'm seeing, it looks very random as to how many past candles it calculates the SAR for.

I pulled up a 60 minute chart and added the custom Parabolic SAR study. In this study I put a print statement to print a count so that I knew how many candles it was calculating the SAR for.

For the symbol APPL, I'm getting a count of 145. For the symbol IBM, I'm getting a count of 344. For the symbol GOOG, I'm getting a count of 326. For the symbol CSC, I'm getting a count of 513. As you can see the counts are different. Does anyone know why?

[Dave180]

I assume this is the number of candles in the chart? If so the timetemplate will determine the initial count, but if you leave the chart runnng, or have a dynamic template and scroll, then when you reload the symbol or open another chart with the same symbol you will (unless they've changed it) get whatever the number of candles in the cache is, not the template.

[nhoak]

Thanks Dave180!

That explains the differences for the candle counts.