Merge branch 'stochastic_genetic' of gitlab.ai.vub.ac.be:sdeproost/bnaic2024 into stochastic_genetic

Merge branch 'stochastic_genetic' of gitlab.ai.vub.ac.be:sdeproost/bnaic2024 into stochastic_genetic

ID can appear in many places in the genome, and makes "gaps" that can be
filled by other operators when mutating. This allows programs of 20
genes to produce short code, with many identities, and learn stably on
SimpleGoal

"rank" simply does not work nor learns anything...

Randomize programs in a fixed way: values close to x.5 get cast to int(x), values further away may slip to x-1 or x+1

Constant propagation does not resolve things like (x[0] != x[0]) as
False, which then prevents the evaluation from detecting that the
program is not actually looking at x. By removing any operator that may
map a function of x to a constant, we make sure that a program that has
"x" in its expression actually has an x-dependent output.

NOTE: This commit does not address a corner case, of exp(large-number)
being a constant (exp(10)). Maybe some programs may learn to use exp and
then a division by a constant to fabricate a constant even though "x"
appears in their string representation

Add negative versions of functions that have a positive image, so, on
average on random DNAs, the output of a program is not biased towards
positive numbers.