Fractal Software > Programming

 Pertubation Theory Deltas Precision

(1/2) > >>

superheal:
I am kinda late to the party, but I decided to add perturbation theory and SA in Fractal Zoomer, following this excellent code base (https://github.com/ShiromMakkad/MandelbrotPerturbation) as an example.

I managed to implement it Perturbation theory and SA for Mandelbrot, and perturbation theory for Cubic Mandelbrot, Burning ship and Cubic burning ship.

I calculate the reference point from the center coordinates in high precision (using Apfloat in java) and then I store the Reference Xn on an array of double precision complex numbers.

In order to calculate the rest of the pixels, I find the Delta0 by by doing this : Delta0 = PixelComplexVal - ReferenceComplexVal, and then I create a Complex number of double precision from Delta0. PixelComplexVal  and ReferenceComplexVal are complex numbers with high precision.

This let me reach the limit of doubles as the Delta0 can reach +-e-300ish.

My question is what can I do to overcome this limitation.
In SA I had to implement FloatExp and ComplexFloatExp. Do I have to do something like that for the rest of the calculation?

Thank you very much for any input!

Linkback: https://fractalforums.org/index.php?topic=4385.0

claude:
https://fractalforums.org/programming/11/fast-extended-range-types/4224/msg28710#msg28710

superheal:
This is just a different way to do it right? Can it be done with FloatExp as well?

claude:
Yes but floatexp for everything is much slower.

superheal:
Is there any sample code available, which describes the comment that you posted?

Navigation

[0] Message Index

[#] Next page

Go to full version
Powered by SMFPacks Rates (Facepunch) Mod
Powered by SMFPacks SEO Pro Mod | Sitemap
Powered by Advanced Topic Prefix Pro
Powered by SMFPacks Advanced Attachments Uploader Mod