[Math] PolynomialCurveFitter

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

[Math] PolynomialCurveFitter

Bruce Johnson

Issue MATH-1009 was closed as resolved in version 3.3.

The issue raised was about the fact that the PolynomialFitter (now PolynomialCurveFitter) does the curve fit with non-linear regression (and requires initial guess and number of iterations).

I can’t think of any reason why a polynomial curve fit (with a least squares target) should be done this way instead of doing it with a linear method (like QR or SVD) and I don’t know of any other math package that does polynomial fits this way.  

I’m not a numerical methods expert, so maybe I’m wrong, but it seems that this really should be resolved before the release of Math 4.0 (or even 3.6).

Should MATH-1009 be reopened, or should I create a new issue (which will just be a copy of 1009 changed to refer to PolynomialCurveFitter).

Bruce




---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [Math] PolynomialCurveFitter

Luc Maisonobe-2
Le 05/11/2015 19:00, Bruce Johnson a écrit :

>
> Issue MATH-1009 was closed as resolved in version 3.3.
>
> The issue raised was about the fact that the PolynomialFitter (now
> PolynomialCurveFitter) does the curve fit with non-linear regression
> (and requires initial guess and number of iterations).
>
> I can’t think of any reason why a polynomial curve fit (with a least
> squares target) should be done this way instead of doing it with a
> linear method (like QR or SVD) and I don’t know of any other math
> package that does polynomial fits this way.
>
> I’m not a numerical methods expert, so maybe I’m wrong, but it seems
> that this really should be resolved before the release of Math 4.0
> (or even 3.6).
>
> Should MATH-1009 be reopened, or should I create a new issue (which
> will just be a copy of 1009 changed to refer to
> PolynomialCurveFitter).

Yes, a new issue would be better.

best regards,
Luc

>
> Bruce
>
>
>
>
> ---------------------------------------------------------------------
>
>
To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>
>


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]