[Thread Prev][Thread Next][Index]

Re: [ferret_users] Running Variance



Demean the data, square all values, do @sbx:N of the squares.

If you want to demean within the running intervals instead of over the whole time series, then subtract @sbx (of the original unsquared data) for each interval first.

BK

On Aug 4, 2014, at 13:36 , terra brasilis wrote:

> Dear expert users :-)
> 
> I would like to calculate a "running VARIANCE" (like a running mean, but instead of mean, VARIANCE).  In other words how to I get a transform for @running_var that would the the same thing as @sbx?
> thank you!
> Terra




[Thread Prev][Thread Next][Index]
Contact Us
Dept of Commerce / NOAA / OAR / PMEL / Ferret

Privacy Policy | Disclaimer | Accessibility Statement