SIR parameters fitting with Maximum likelihood estimation

Dear community,

I am trying to fitting data in SIR model, I want to get a parameter (beta,gamma) based on MLE, but a have the following error:

Supplied objective function must return a scalar value.

The SIR model>

I am using differte functions like:

(System ODE)

function dudt=sirpar(~,u,theta)
%Asignar compartimenteos 
S=u(1);
I=u(2);
R=u(3); 
%Escritura del sistema de ecuaciones diferenciales 
%ds/dt%
dudt(1,1)=-theta(1)*S*I;
dudt(2,1)=theta(1)*S*I-theta(2)*I;
dudt(3,1)=theta(2)*I;

end

Resolution

function [y] = odefit(t_obs,theta)
S0=1; I0=1/5072000; R0=0;
    [t,y] = ode45(@(t,u)sirpar(t,u,theta),t_obs,[S0 I0 R0]);
end 

Log_like function

function val=log_lik(theta,datos_obs)
t_obs=1:length(datos_obs(:,2));
[y] = odefit(t_obs,theta);
I_mod=y(:,2);
val=-sum(log(normpdf(datos_obs,mean(I_mod),1)));

and the principal script

function [y,val,theta]= SIR_MLE_2
tol=1e-10 

while  nomr((I_obs-I_mod).^2)<tol 
% Datos experimentales
datos_obs=xlsread('datos_sir');
S_obs=datos_obs(:,1);
I_obs=datos_obs(:,2);
R_obs=datos_obs(:,3);
% Sisitema de ecuaciones SIR

%Condiciones iniciales
theta=[0.341 0.141];
%Tiempo  
t_obs=1:length(I_obs);
%Parámetros
%Resolución del sistema en función del parámetro
[y] = odefit(t_obs,theta);
%Control Gráfico
%plot(t_obs,y)
%plot(t_obs,err)
%Datos del modelo
S_mod=y(:,1);
I_mod=y(:,2);
R_mod=y(:,3);
%plot(t_obs,S_mod,'b')
hold on 
plot(t_obs,S_obs,'r')
%MLE SUm
val=log_lik(theta,datos_obs)
%Optimizacion 
options=optimset('Display','off','MaxIter',10000,'TolX',10^-30,'TolFun',10^-30);
theta0=[0.341 0.141];
[theta,fval,exitflag,output,grad,hessian]=fminunc('log_lik',theta0,options,datos_obs);
theta0=theta; 

end


end

The data attachmet here!

I hope anyone can help me!

Thanks :)