spline for increasing-decreasing data

[looser]

 Hi every body.

I am obtaining data from two multimeter contiously , lets call them and x and y. And i plot them in a xy graph, which   forms a circle every 200 data points. Nest 200 x,y data forms another circle and i want to average it withe the previous one. In anothe words, i have two functions F1(x1i)=y1, and F2(x2i)=y2,   and reprsents 2 circles to be averaged. The problem is that but the X values are not the same so i can not sum both y values  for the corresponding x value , and divide them by 2. I thought to use interpolate each circle and the average them. but then spline function does work only for increasing x values, in the case of a circle x values increases and decreases around 0..

anyone has an idea what could be the result

[tinnitus]

hello,

If 200 is a real periodic cycle, why don't you average (x0 & x200) (y0 & y200) (x1&x201) (y1&y201) and so on ...

Best regards

Tinnitus

[Darin.K]

In this example I bin by angle and average.  I would have preferred to experiment with an In Place Structure but I still haven't upgraded LV on this computer.

[looser]

Actually, i named them as function just for simplicity. Indeed, they are hysteresis loops, shown below. In the picture, there are just 2 loops ( red and black) and as you see that x data points are not matching excatly- though they are close to each other-.. i need a very precise averaging of 100 hysteresis loops like in the picture...Thats why i thought that interpolation each loop and average the interpolated loop would be the beast solution. but it s kinda hard to do.

any better idea?

[tinnitus]

 !!!

a little bit lost

what do you really want to do, in a frst step you could also reduce noise filtering signals independently at fmax for exemple...

could you be more precise on what information you want to extract from this measures 

best regard

Tinnitus

[looser]

Well, the aim to enhance the signal to noise ratio by averaging individual loops, every 800 xy data taken from 2 multimeters forms a loop. Due to the thermal drifts and other external effects each loop does not  match exactly the other one. meaning  x1 could be different than x801, and y1 is different than y801. you think is this the best way for averaging? 

Wouldnt be better to intorpolate 2 loops in the same limits, and averaging the interpolated loops?