12-13-2009 02:54 AM
Hello,
Im looking for a way to implement 'Moving Standard Deviation' into the FPGA module.
Are there any examples out there?
I'm using cRIO-9012 chassis.
Thanks,
Yuval
12-13-2009 07:14 PM
12-14-2009 10:05 AM
I don't know much about the algorithm, but from what I've seen it seems like just a few simple calculations. It doesn't seem like it would be too tough to implement using the numeric palette calculations.
http://www.transitionstrading.com/mstd.htm
12-14-2009 11:11 AM
stu@viewpointusa.com wrote:
how many data points are you considering? since the algorithm requires a pass through all the data for each new data point, it will be cycle consuming. how fast does it need to run?
Sorry, the algorithm does NOT require a pass through the system. is the same calculation as computing the AC RMS value of X which can be computed point by point on a fairly arbitrary set of data. You do have to keep the set of data around though. There is a built in calculation of standard deviation pt by pt that will show you an implementation but it is not coded for FPGA.
12-14-2009 11:47 AM
I've calculated variance and standard deviation in FPGA using FXP math toolkit. Attached are the VIs.
12-14-2009 02:09 PM
Thanks very much to all of you.
I will look at it today and come back with the results.
Yuval
12-14-2009 08:02 PM
dwisti wrote:I've calculated variance and standard deviation in FPGA using FXP math toolkit. Attached are the VIs.
Whenever i code an FPGA algorithm, i create a test case with known inputs. I am not sure if you did for this
code or not but i put together a test case that does not produce expected results. I am not sure if i implemented the
your code correctly but you can check it out.
12-15-2009 12:42 AM
Many thanks Stu. I'll be back with results.
Yuval
12-15-2009 07:39 AM
12-15-2009 07:54 AM