Help with Filtering an Array Using Standard Deviation

[Kevin_Price]

Glad you have a working solution.   FYI, "noise" is a term that usually implies a degree of randomness and unpredictability.  If you can solve your problem merely by keeping / rejecting alternating values, that's a very simple and regular pattern -- so "noise" is probably not the best way to describe it.

-Kevin P

[joshua.l.guthrie2.civ]

I know you say Standard Deviation.. but would middle third averaging work (trimmed mean)?  

[altenbach]

---

@joshua.l.guthrie2.civ wrote:

I know you say Standard Deviation.. but would middle third averaging work (trimmed mean)?  

---

Here we have two interlaced signal, one as shown and one all near zero. Both signal have equal weight. Any running average with an odd window will favor one or the other while with an even window things will be even more blurry.

If the undesired values are more random, we could start with a known good value and reject points that differ too much form the previous, but that will also make the time base irregular (nonlinear). In that case, we would typically replace the bad values based on interpolation from good neighbors, while keeping the number of points constant.