% Ella Bingham, May 2005 & Feb 2006.  
%
% Discard the disconnected rows of the similarity matrix of data rows. Also
% discard the corresponding rows in siteinfo.
%
% Inputs:
% data - a data matrix of sites=rows, taxa=columns.
% siteinfo - a structure containing auxiliary information about the sites.
% similarity_measure - how the similarity matrix is formed. 'dot' for plain
% dot product similarity, 'wdot' for weighted dot product similarity (takes
% into account the total number of attribute appearances in the whole data).
%
% Outputs: 
% outputdata, outputsiteinfo - same as the input variables, except for the
% disconnected rows.
%
function [outputdata,outputsiteinfo] = findconnecteddata(data,siteinfo,similarity_measure);

[spcoeff,simm,lapm] = laplacian(data,similarity_measure); 
[n,n] = size(simm);
m = ceil(log2(n));

T = simm;
for i=1:m
  T = T^2;
  T = (T>0); 
end;

connectedsites = find(sum(T)>=mean(sum(T)));
outputdata = data(connectedsites,:);

% Update the contents of siteinfo correspondingly
outputsiteinfo = siteinfo;
names = fieldnames(siteinfo);
if ~isempty(names)
   for field=1:max(size(names))
      fieldcontents = getfield(siteinfo,char(names(field)));
      outputsiteinfo=setfield(outputsiteinfo,char(names(field)),fieldcontents(connectedsites));
   end
end