MaplePrimes - answers and comments on Question, A more accurate line equation

A better fit

Markiyan Hirnyk — Fri, 08 Jul 2011 22:50:35 Z

Increasing degree, we obtain a better fit:

Exponential fit

Robert Israel — Sat, 09 Jul 2011 02:13:44 Z

As Axel said, polynomials are probably the wrong class of functions. I'll try constant + linear + exponential.

> fit2:= Fit(a + b*x + c*exp(-d*x),X,Y,x,parameterranges=[d=0..0.1]);

Thanks!

press9761 — Sat, 09 Jul 2011 21:23:40 Z

I have a lot of information to work with now.

And I now see what to do.

Thank you all so much!!

hm ...

Axel Vogt — Sat, 09 Jul 2011 00:46:37 Z

looking at it it seems data become linear in the large and thus doubt an polynomial would satisfy

without knowing more, that probably is not a way for a 'solution'

Residualleastsquares

Markiyan Hirnyk — Sat, 09 Jul 2011 07:51:54 Z

> fit3 := Fit(a+b*ADC+c*exp(-d*ADC), X, Y, ADC, parameterranges = [d = 0 .. .1], output = [residualsumofsquares, leastsquaresfunction]);

[1164.31223443563, 628.533114813602764 + 0.113596427586220516 ADC

- 655.479174035531400 exp(-0.00245278681774733576 ADC)]

PS. Take look at

Download fit4.mw