Denis Boyer. Departamento de Sistemas Complejos. Instituto de Física. UNAM

FURTHER ANALYSIS OF THE .txt FILES:
Two other separate Fortran programs can be compiled to perform further analysis of the existing .txt files. A)Preferential associations: The ASCII program preferassoc.rnd.time.F computes, for each independent run, the relative affinity defined in the manuscript, averaged over all foragers. At the same time, the program computes the value of the relative affinity that would be given if each forager met the same individuals and had the same number of total encounters, but distributed with equiprobability among these individuals. These two values are finally averaged over the independent runs. The values are written as ascii in the output file affinity.time.rnd, each line corresponding to a different run. The last line corresponds to the average values. Note: In order to run preferassoc.rnd.time.F you need the file namerun.f in the same directory. That file contains the names of the .txt results files to be open. Editing parameters: Directly in preferassoc.rnd.time.F: ntime =number of iterations in a run; nmonos=number of foragers; nbrun=number of independent runs to analyze; t0=iteration where the analysis begins (1, typically); outputdat=name of the file where the results are written. B) Network properties: The program network.perco.F does various tasks (all the final quantities correspond to averages over the independent runs): a)It determines the strong links according to the criteria described in the manuscript (strong bonds) and analyses the resulting network: it computes its clustering coefficient, the degree distribution and mean degree. b)It also compute the mean degree of the network of all bonds (weak + strong), i.e. the average number of distinct individuals met by a forager. c)It computes the number of nodes (individuals) that belongs to the largest cluster of the network. It also compute the total number of (disconnected) clusters in the network. All these results are stored in the output affinity.network.perco. Note: In order to run network.perco.F you need to have the file namerun.f in the same directory, that contains the names of the .txt results files to be open. You also need to have the file subrout.clust.net.f, that contains the subroutines that do the cluster analysis. Editing parameters: Directly in network.perco.F. The parameters are the same as in A) above.

FURTHER ANALYSIS OF THE .txt FILES:

Two other separate Fortran programs can be compiled to perform further analysis of the existing .txt files.

A)Preferential associations:
The ASCII program preferassoc.rnd.time.F computes, for each independent run, the relative affinity defined in the manuscript, averaged over all foragers. At the same time, the program computes the value of the relative affinity that would be given if each forager met the same individuals and had the same number of total encounters, but distributed with equiprobability among these individuals. These two values are finally averaged over the independent runs.
The values are written as ascii in the output file affinity.time.rnd, each line corresponding to a different run. The last line corresponds to the average values.

Note: In order to run preferassoc.rnd.time.F you need the file namerun.f in the same directory. That file contains the names of the .txt results files to be open.
Editing parameters: Directly in preferassoc.rnd.time.F: ntime =number of iterations in a run; nmonos=number of foragers; nbrun=number of independent runs to analyze; t0=iteration where the analysis begins (1, typically); outputdat=name of the file where the results are written.

B) Network properties:
The program network.perco.F does various tasks (all the final quantities correspond to averages over the independent runs):
a)It determines the strong links according to the criteria described in the manuscript (strong bonds) and analyses the resulting network: it computes its clustering coefficient, the degree distribution and mean degree.
b)It also compute the mean degree of the network of all bonds (weak + strong), i.e. the average number of distinct individuals met by a forager.
c)It computes the number of nodes (individuals) that belongs to the largest cluster of the network. It also compute the total number of (disconnected) clusters in the network.

All these results are stored in the output affinity.network.perco. Note: In order to run network.perco.F you need to have the file namerun.f in the same directory, that contains the names of the .txt results files to be open. You also need to have the file subrout.clust.net.f, that contains the subroutines that do the cluster analysis. Editing parameters: Directly in network.perco.F. The parameters are the same as in A) above.