convertUnweightedPajekToAdjMatrix.m

% (c) 2012-2013 M Schaub -- michael.schaub09@imperial.ac.uk
function A= convertUnweightedPajekToAdjMatrix(filename)
% Convert undirected, unweigthed pajek graph into sparse adjacency matrix. 
% Assumes node numbering in pajek file starts with 0 or 1, 
% and is contiguous.
%
% Inputs:
%
%           filename:   string with filename of input file,
%                       e.g. 'pajekgraph.net'. Links in the Pajek file should have no weight, i.e. should be in a format
%                       [from] [to]
%                       
% Outputs:
%           A:          Adjacency matrix of undirected graph in sparse data
%                       format

% open file
fid = fopen(filename ,'r','n','ascii');

% skip "intro" part in pajek file
line=fgets(fid);
while(~feof(fid) && ~strcmp(line(1:6),'*Edges') && ~strcmp(line(1:6),'*edges') && ~strcmp(line(1:5),'*Arcs') && ~strcmp(line(1:5),'*arcs'))
    line=fgets(fid);
end

% scan adjacency list into G
G = fscanf(fid,'%e %e', [2 inf]);
fclose(fid);

% check if node numbering starts with 0 and in case make it start from 1
if min([G(1,:) G(2,:)]==0)
    G(1,:) = G(1,:)+1;
    G(2,:) = G(2,:)+1;
end

% find number of Nodes
N= max([G(1,:) G(2,:)]);

% allocate and fill in adjacency matrix
A= sparse(G(1,:),G(2,:),1,N,N);

% undirected pajek graphs are "one way"/nonsymmetric, hence symmetrize and remove double counting of self loops
A = A+A'-diag(diag(A));

end