% given already binned data:
% a is the value, and b is the count
% create a logarithmically binned histogram using the optional argument
% logbase for the width of the bins

function [midpoints,histsum] =logbinfromlinbin(a,b,logbase)

if (nargin < 3)
    logbase = 2
end

if (logbase <=1 )
    disp('warning: logbase should be > 1. exiting...');
    return;
end

[a1,ia] = sort(a);
a2 = b(ia);


max2 = ceil(log(max(a1)+1)/log(logbase));
tmp = 0:max2;
x = (logbase).^(tmp);
midpoints = ceil((x(2:length(x))-x(1:(length(x)-1))-1)/2)+x(1:(length(x)-1));
disp(midpoints);
disp(x);
i = 1;
j = 1;
histsum = zeros(size(midpoints));

while (i <= length(a1)) 
   if (a1(i) < x(j+1)) 
      histsum(j) = histsum(j) + a2(i);
   else
      j = j + 1;
      histsum(j) = histsum(j) + a2(i);
   end
   i = i+1;
end
   
histsum = histsum./(x(2:length(x))-x(1:(length(x)-1)));