OmerRosler/fractal_plotter

Introduce infinite zoom

Opened this issue · 0 comments

We want infinite zoom-in ability for all fractals.
The problem is that the current algorithm works with points i.e. doubles and when the zoom-in is large enough we run into the precision limit.

We could work around it if we introudce a "reference point", and instead of working with points, we work relative to this reference. Then we would mathematically know "the first k digits" agree, and only calculate with residues.

This can be done in two ways:

  1. Ad-hoc, where the reference is chosen explicitly in the algorithms.
  2. Use infinite-precision library, but that is problematic as the reference choice is part of the algorithm