Who will use matlab to solve SBM model

Please refer to the following procedure:

% curvefitLogistic.m

% in practical application, only when the obtained biological data is proved to have the law of logistic curve can the fitting result be

% is meaningful, and the following set of data has the law of logistic curve!

n =[3929 5308 7240 9638 12866 17069 23 192 3 1443 38558 50 156 62948 ...

75995 9 1972 1057 1 1 122775 13 1669 150697 179323 203 185 226500];

t = 1790: 10: 1980；

% draw the original data map

plot(t，N，t，N，' r* ')

continue

Construction of% logic function

func = @(beta，x) beta( 1)。 /( 1+(beta( 1)/N( 1)- 1)* exp(-beta(2)*(x- 1790))；

% Key: Selection of optimization initial value: beta0( 1) is the maximum value of the function, and the smaller beta0( 1) is, the better, so we can directly take 0 here.

% because of gauss-Newton algorithm and gauss-

% Newton's improved algorithm Marquardt method, these two algorithms only select a set of initial values close to the actual value.

When the value of% is k, a, b, the convergence of iterative process is good, that is, the discovery value of k, a, b can minimize the value of r!

beta 0 =[max(N)0]；

% lsqcurvefit call

[beta，resnorm，residual，exitflag，output]= lsqcurvefit(func，beta0，t，N)；