Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.


ariel
Posts:
4
Registered:
12/8/12


Re: fitting ( cftool )
Posted:
Dec 10, 2012 7:43 AM


"Steven_Lord" <slord@mathworks.com> wrote in message <ka3mcf$cs1$1@newscl01ah.mathworks.com>... > > > "ariel " <arn335@msn.com> wrote in message > news:k9veu1$7s$1@newscl01ah.mathworks.com... > > im trying to do a fitting between a vector and a function . > > my vector is a intensity summed by 20 degree angle (160180) as a function > > of the beam radius, as the following pics describe : this is the picture : > > http://postimage.org/image/baff0lf63/ > > the imagesc pic after cutting by angle plus intensity graph by radius ( > > the angle is summed ) : http://postimage.org/image/s9l2d8j5b/ > > > > my try to to do fitting with cftool : > > http://postimage.org/image/6so1ugaeh/ > > > > the function i try to do fitting with : sin(b*sin(atan(x./0.05)).^2).^2+c > > which i took from theory . > > as you can see the fitting is total joke, and ideas ? > > 1) My guess is that the Curve Fitting window told you that there was some > difficulty with the fit. If so, what did it say? > > 2) How certain are you of your function? From the image of your data it > looks more like a sinusoidal times a quadratic or something similar. Taking > sine of sine of arctangent seems odd to me. >
1) didnt get any error message .
2) 99% sure, in the books the forumla appears without atanx but with X but the X in the books equalts to atan(x'/0.05) when x' is the x' in my data (came from simple geometry)
any ideas?



