Re: data fitting/analysis - Spline?

On Sep 3, 11:52 pm, Uwe Schmitt <rocksportroc...@xxxxxxxxxxxxxx>
wrote:
On 3 Sep., 20:22, mariobard...@xxxxxxxxx wrote:



Hi.
I don't know if i'm writing in the right group but i wrote in the
matlab's one and i haven't received any answer so far.
So i hope here omebody can help me. This is my problem:

I'm analyzing a list of x-y values,http://www.geocities.com/mariopwr/data.png
(http://www.geocities.com/mariopwr/dataxy.zipforthe .mat file) ,and
i'd like to find out the best function that fits these points.
I tried using the matlab spline toolbox and I found a quite close (in
my opinion) fit using the least square approximation with order 14
(http://www.geocities.com/mariopwr/spline.png). Do you think it's a
good approximation or should i try a different method?

I'd like to find a way to fit these points and later on other similar
(with similar trend) data.

Anyway, after that i'd want to extract a few information from this
graph like amplitude and width of each wave. Any
suggestion?

Thank you for your hel

Mario

Can you say something about the origin of the data ? Which kind of
noise do you have ?
Do you have a modell for your data ?

You could use leave-one-out or crossvalidation for finding the best
parameters for fitting.

Greetings, Uwe

The values describe the movement (variation of y-coordinate) of the
centre of mass of an animal.
I don't have any model for my data. I just know that the ideal graph
should be periodic and every period describe a full stride. In a
period there are 4 peaks, when the (back or front) legs of the animal
overlap. In data.png you can see there are 8 peaks so that graph
describe almost 2 complete strides.
I was thinking about describing the data using a sum of sinusoidal
function and then compare different graphs paying attention to the
variation of the coefficients that describe the function. But I don't
know if it makes sense and if I get something interesting.

Cheers,
Mario.
.

Relevant Pages

  • Re: data fitting/analysis - Spline?
    ... graph like amplitude and width of each wave. ... should be periodic and every period describe a full stride. ... In data.png you can see there are 8 peaks so that graph ... So you could fit c*sinto the data or at least to ...
    (sci.math.num-analysis)
  • Re: VMR9 Stubbornly Insists on Format
    ... If the negotiation fails, the graph should fail. ... > real width or the stride (why isn't there a field in the format descriptor ... While the VMRs kindly attach the new media type ...
    (microsoft.public.win32.programmer.directx.video)
  • Re: Why is Microsoft Word Complaining About This Sentence?
    ... The variation of yy with xx is shown in Figure 7. ... Figure 7 displays a graph of yy against xx. ... The earth is kept in orbit by the gravitational attraction ... The gravitational attraction of the sun keeps the earth in ...
    (alt.usage.english)
  • Re: Slow Motion Cosmological Train Wreck
    ... You can look at one graph in one paper over and over and over again, ... the easiest peaks to observe and the first likely to be reported. ... definitive prediction of the dark matter mass spectrum is that it was ... worry about because nature will show that you are right. ...
    (sci.astro)
  • Re: Crazy, Murray is driving to cincinatti not catching a plane... id imagine its a 12 hour drive fr
    ... Erm....8 years of cooling is proof of 8 years of cooling...which if you scan ... When I look at that line, I see extreme warming, happening ... variation that does not come close to being as great as the variations ... What do you think that graph should look like? ...
    (rec.sport.tennis)