See the corresponding manuscript (Proc. R. Soc. B) here.

 

The Fortran code is greedy.weight.f

N point-like targets are distributed randomly in a square domain of length 1. Both x(i) and y(i), the coordinates of target number i (i=1,…,N), are random numbers uniformly and independently chosen in the interval (0,1). rkiss() is the random number generator.

The fruit content of target number i is a positive integer, stored as eta(i), distributed according to an inverse power law p(k)=C k^–λ (k=1,2…,kmax). C is the normalization constant such that Σ_k=1...kmax p(k) = 1. The exponent λ is β in the manuscript. The maximum value kmax is a cut-off that depends on N and λ: kmax is chosen such that there is at most one site, among the N ones, with size larger or equal to kmax. kmax therefore satisfies the equation: N Σ_k=kmax...∞ p(k) = 1.

 

A forager starts from a site i₀ located near the center of the square domain. The values of the ratio d(i₀,j)/eta(j) are calculated for all j different from i₀, where d(i₀,i) is the Euclidian distance between sites i₀ and j, and eta(j) fruit content of target j.The site with lowest ratio is i_m. The forager’s new position is set to x(i_m), y(i_m). The target i₀ is removed from the box (x(i₀)=y(i₀)=100, eta(i₀)=1) and i_m becomes the new i₀. The process is carried out recursively. (The program has not been optimized.)

The result file is <namerun>.<run number>.pos: for instance, the file <namerun>.024.pos contains the results of the 24^th independent run, if various runs are performed with the same parameters. Each result file records 3 quantities per line: x(i₀), y(i₀), eta(i₀). The lines correspond to the sequence of visited targets.

 

Parameters of the program:

 

ntrap :total number of targets

tmax : number of visited targets by a forager during the run (‹‹ ntrap).

namerun : first letters of the name of the result files. It is a character chain of 13 letters, see line: character*13 namerun. The length can be changed accordingly.

lambda : parameter β in the manuscript.

iruni : usually set to 1.

irunf : number of independent runs (different environments and starting positions).