Nice work Bruce. I was trying to make up my own algebraic equivalent in lieu of a reference source, or sources.
Did you get 'your' version of StDev from a reference of is it your own work? --- In [email protected], "bruce1r" <bru...@...> wrote: > > Herman, Mike, all - just checking in and think I can offer a quick > suggestion and some code that I have handy. > > 1. StDev() is a function that is not implemented as a variable period > function, but can be with a slight caveat. Because this uses an > alternative, but algebraically equivalent formulation, that employs a > "data^2" term, it loses precision at about the 5th significant digit. > That is close enough for most work, though. So, anyway, this one liner > function will give a fast, variable period StDev ( IOW, vper can be a > varying array) - > > function StDevVarPer( data, Vper ) > { > return sqrt( MA( data ^ 2, vper ) - MA( data, vper ) ^ 2 ); > } > > 2. This is fast even with large numbers of 1 minute bars, but it is > usually best to control QuickAFL with a statement like the following as > the LAST STATEMENT in the program to yield maximum performance - > > SetBarsRequired( 500, 0 ); > > Remember that with QuickAFL in an indicator, the first execution after > an edit or insert passes the entire array to set the bar requirement, > then QuickAFL is in force on a subset. > > -- BruceR >
