Incorrect logic in reproduction step of genetic algorihm example (knapsack problem)?

[fotto]

Grokking-Artificial-Intelligence-Algorithms/ch04-evolutionary_algorithms/knapsack_genetic_algorithm.py

Line 171 in 89002a4

children = one_point_crossover(chosen_selections[parent_index],

I know it's just an example but IMHO there are two problems with the actual implementation of the reproduction step:

the reproduce_children() function is called with 100 genes but

• half of the parents are not used at all:
  • it combines chosen_selections[0] with chosen_selections[1]
  • then chosen_selections[1] with chosen_selections[2]
  • and so on and until chosen_selections[48] is combined with chosen_selections[49]
  • chosen_selections[50:] is not used at all
• it forget the results of reproduction:
  • only the two children of last iteration are returned
  • all others are ignored