Catalogue files used to generate the effective models described in arXiv:1710.01742 [cond-mat.dis-nn]. Files are structured as catalogue_prototype_matrix_type.dat where *prototype = the simulation where the catalogue was computed from. In this case it is "SiS_GG" to indicate it is Si:S with geometry relaxation effect included. *matrix = either Hamiltonian (ham) or overlap (ovlp). *type = it can be 0 (matrix elements pertaining to the impurity sites), 1 (silicon sites neighbouring a single impurity), 2-5 (silicon sites neighbouring a pair of impurities which are 1st to 4th nearest neighbours). For the silicon-only background, the filenames are catalogue_ham_-1.dat catalogue_ovlp_-1.dat All files consist of lines describing the matrix elements between sites i and j. These lines contain, in order, *an integer indicating the sublattice of site i *an integer indicating the sublattice of site j *the three real coordinates of the displacement vector from site i to j *the three real coordinates of the displacement vector from site i to the closest impurity *the three real coordinates of the displacement vector from site i to the second closest impurity *81 real numbers of the 9x9 matrix block, printed by rows.