fitting double exponential curve fitting to data - Success!

[American Citizen <website.reader3@gmail.com>]

Hello all:

I am glad to report that using a technique originally developed by J 
Jacquelin, documented at 
https://math.stackexchange.com/questions/1428566/fit-sum-of-exponentials/3808325#3808325, 
I was able to avoid computing the gradient and Hessian matrices and find 
the answer directly to fitting data to the equation

y = a + b*exp(p*x) + c*exp(q*x)

I attached a png file showing how well this works.

Now I can efficiently estimate the regulator for some elliptic curves 
who have a conductor so huge, it's not possible to directly compute the 
L-series.

If you want, I can send you the GP-Pari code, just ask.

Randall

Attachment: testing.png
Description: PNG image