Now a set of useful mutations are implemented, and balanced so that the number of nodes or edges doesn’t explode.
The mutations I’ve implemented (node are genes are nodes):
add random gene delete node delete group of nodes (range of indexes) split edge create new node in place of an edge (insertNode) flip edge duplicate node duplicated group of nodes (range of indexes) change node index (regrouping/separating functional groups) change group of nodes index (transposable elements) create random edge delete random existing edge scale existing edge weight negate weight redirect existing edge to random node scale parameter (k1, b, k2) negate bias
This is the next installment from the Gene Regulatory Network saga.
previously: lemmy.sdf.org/post/1967056######