metawards.mixers.merge_matrix_multi_population(network: Networks, nthreads: int, profiler: Profiler, **kwargs)

This merge_function merges the FOIs across all demographic sub-networks according to the interaction matrix stored in networks.demographics.interaction_matrix, using the number of people in each demographic separately, so as to better model demographics that have different contact parameters. In this case, the different demographics are assumed to be in different populations and so have different contact probabilities. The FOIs are merged together where they are divided by N_j (where N_j is the number of individuals in the jth demographic), e.g.

FOI_i = FOI_i / N_i + FOI_j / N_j + FOI_k / N_k …