Constrained optimization of cubic polynomial

[Sammo]

Hi all,

I'm new to Python (and to coding in general). If anyone can help out with the problem below or point me in the right direction I'd really appreciate it.

I have a noisy dataset through which I'd like to fit the best possible cubic polynomial using the least squares method. numpy.polyfit is almost what I want, but for physical reasons there are some additional constraints on the curve:

1. It should pass through the origin (0, 0) and

2. All of the coefficients should be non-negative.

Basically I want to fit a curve of the form

y = ax**3 + bx**2 + cx where a, b and c are all >= 0

Is there a way of using numpy.polyfit with these additional constraints, or is there another method entirely that's accessible to a relative beginner?

Thanks very much for any help or advice!

[dwblas]

I would also try the numpy list http://www.scipy.org/Mailing_Lists This is beyond me but perhaps someone else will have some insights.

[Sammo]

Thanks dwblas, I'll try the numpy list too.