Analysis: arithmetic multiplication

[DavidLeoni]

Currently in analysis page there are some integer multiplications by constant two: I considered them as $O(1)$ as small ints in CPUs would actually require just a bit shift, but after some research I fear in Python they are still really slow $O(k^2)$ with $k$ bits

Alas with multiplication there are to many levels of abstraction: theoretical (for single core), parallel implementation in actual CPU for small ints, implementation in Python... To avoid confusion I think it would be better to just get rid of multiplications and substitute with sums (when it makes sense).