setup-siblings produces an exaggerated number of siblings

[ambitious-octopus]

During standard setup the setup-siblings procedure generates an exaggerated number of siblings. The problem does not extend to connections that are created during the generate_households procedure. The number of siblings that are generated is proportional to the number of initial agents of the model.

You can try it yourself by entering this code at the end of setup-siblings:

  ask persons with [ any? out-offspring-link-neighbors ][
    show count sibling-link-neighbors]

With 2k initial agents this is the result:

Possible solution:
During the mesa translation I solved this problem by eliminating within the cycle the agents that have already been connected. Without eliminating these agents, it happens that the agents who have already been assigned siblings are fished out and added an additional group of siblings among the candidates, creating an exaggerated number of siblings.