Adding weight to data points

Question

LS 2011-7-21

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/12158-adding-weight-to-data-points

Hi all,

I'm working on basic model fitting in Matlab and now I'm trying to figure out how to add weight to certain data points and take away weight from others. I have the fitting down so that two of the parameters (r and K) are estimated successfully but the fit isn't great. I tried something (in the third cell of the first script copied below) that I had seen in previous code but it didn't work (no errors but estimated parameters that were way off). Any help would be much appreciated! Thank you!

%%Data import
load 'cultures1e3.csv' %load data for culture
global dataT dataN1
dataT = cultures1e3(:,1); %x = time (days)
dataN1 = cultures1e3(:,2); %y = cell count (cells/mL)
%%Giving data weight - what I've tried so far but it doesn't work
%points to take away weight
dataN1([14 16]) = 0;
%keep weight
dataN1([1 2 3 4 5 6 7 8 9 10 11 12 13 15 17 18]) = 1;
%%Curve fitting 
% Initial estimates for r and K ( p0 = [r;K])
p0 = [0.1; 3.5e6] ;
% Estimate parameters
tic 
pFit = lsqcurvefit(@logistic, p0, dataT, dataN1) ;
toc

~~~~logistic script (function used in previous script)

function funcN = logistic(p0, T_range)
  %logistic growth equation for data fitting
    %variables
    funcN_init = 1e3;
    %options
    options = [];
    % Solver
    %to estimate r
    [T funcN] = ode45(@odefun, T_range, funcN_init, options, p0(1), p0(2));
    function f_prime = odefun(T, funcN, r, K)
    f_prime = r * funcN * (1 - funcN / K);
    end
    end