function y=linfit(k)
%This function takes in the values of ka and kb
%from the non-linear optimization routine and
%computes the best fit values for the linear
%fitting parameters A and B.  The function returns
%the sum of the square of the deviation which
%we are trying to minimize.  We make use of the
%global command between this function and the
%main script so that we can pass the best fit
%values of A and B back and forth, and also pass
%in the data values.  The latter is done to minimize
%the data I/O.
%
%Thus:
global tpass cpass capass cbpass
clear a
t=tpass;
c=cpass;
a=[exp(k(1)*t),exp(k(2)*t)];
x=a\c;
capass=x(1);
cbpass=x(2);
y=norm(a*x-c)^2;