Supplement 2. MATLAB function files to perform MCMC analysis and the analysis of covariance.

File List Road_map_to_programs.pdf (MD5: 6cabea12cc599775aadafa59fa0c87d2) readdata.m (MD5: 3bdfbc7154cb9aac0aa62d280824a911) nantest.m (MD5: 1053c9d8b0bb9605d7e5dd848d7c8555) matrixmcmcgr.m (MD5: 2ef6a1d52b626d669abcad0630dc2fdb) polyover.m (MD5: f00007a67b901771ff1ec5853a73861b) conpoly.m (MD5: 634b9340a0f7870e69b5f0a65967ae09) overlap.m (MD5: bc4038bbd510126a527ccfd4843087a3) permissible.m (MD5: 33ee427066e5ae51c3975d24a04362f6) actual_count.m (MD5: 7061c1d5e10b7c540c4c0fb0bc839a8a) actual_count_numshar.m (MD5: e9a93094a093c39e9a23bca65518fed0) pvals.m (MD5: 0dc6f6c67cbdfc69328ffb93f2ccfa97) pvals_numshar.m (MD5: 2aa826aa93072c4171ed06015331af01) polyover2.m (MD5: 406c9c9bb45c08d7b1f51fda35fe0a0f) fixpartial.m (MD5: a5ac36ba552b19151576c2b022b10d9b) simulated_count.m (MD5: f804cd8ab2638d8669de5a0f14006bd9) simulated_count_numshar.m (MD5: b131ba01d7ef613644e6082d144ad57d) ancova_prog.m (MD5: d3b1b7232d2f02a7bada58ade3d78911) Description PDF file Road_map_to_programs.pdf provides an overview of how the programs work together. MATLAB script file readdata.m reads in the data on species incidences, island group membership, and island coordinates. MATLAB script file nantest.m is called by readdata.m to convert not-a-number and uncertain status entries in the species occurrence matrix into absences. MATLAB script file matrixmcmcgr.m is the main program function for the MCMC analysis. MATLAB script file polyover.m uses the observed data to calculate overlap between each pair, and also creates index variables to code each pair as congeneric, within-guild, sharing island groups or not. MATLAB script file conpoly.m is called by polyover.m repeatedly to generate the coordinates on the convex hull for each species. MATLAB script file overlap.m uses the convexhulls to calculate the overlaps between pairs. MATLAB script file permissible.m creates a binary permissibility matrix. MATLAB script file actual_count.m generates the actual number of true complete checkerboards. MATLAB script file actual_count_numshar.m generates the actual number of true partial checkerbaords. MATLAB script file pvals.m generates the expected values, sd’s, and p values associated with tests based on defining a true complete checkerboard as observed overlap > expected overlap. MATLAB script file pvals_numshar.m generates the expected values, sd’s, and p values associated with tests based on defining a true partial checkerboard as observed overlap > expected overlap. MATLAB script file polyover2.m calculates the pair-wise overlap for each pair in a simulated matrix. MATLAB script file fixpartial.m determines which pairs are true partial checkerboards (TPC) at a variety of α levels. MATLAB script file simulated_count.m is used to determine in each simulated matrix which pairs are true complete checkerboards. MATLAB script file simulated_count_numshar.m is used to determine in each simulated matrix which pairs are true partial checkerboards. MATLAB script file ancova_prog.m performs the analysis of covariance using the observed and expected values generated by the simulation.