Prev: how to compare variable precision numbers
Next: setting color with data
From: Kevin Reilly on 1 Dec 2009 21:45
Hey all,

I am trying to fit parameters to a kinetic model from experimental data through least squares regression, basically just using fmincon, nothing fancy. I am, however, stuck with an issue; MatLab is telling me:

??? Error using ==> odearguments at 116
Solving RATE_FUNC requires an initial condition vector of length 8.

The annoying part is that I AM specifying and initial condition vector of length 8:

t_stamp = param.measured_data(:,1);
x0 = [0; 0; 0.0553; 0; 0; 0; 0; 0];

[t_traj,x_traj] = ode45(@rate_func, t_stamp, x0, [], k2, k3, k4, k5, k6, k7, param);

The variable t_stamp is just taking the time stamps of the measured data points so that the ODE solver returns the function value at those times. The function 'rate_func' contains seven chemical rate equations and eight differential equations; hence why I need an initial condition vector of length 8. The variables k2 to k7 are the parameters that I am varying.

I have tried even tried entering the code as:

[t_traj,x_traj] = ode45(@rate_func, t_stamp, [0; 0; 0.0553; 0; 0; 0; 0; 0], [], k2, k3, k4, k5, k6, k7, param);

but to no avail. I've also tried it a row vector; nothin'. If anyone is able to suggest where I might be going wrong or what may be causing this error, it would be greatly appreciated. I have spent hours trying to solve this. Please message me if it would help to see the complete code.

Thank you!
From: Eugene on 4 Dec 2009 03:25
"Kevin Reilly" <kevin.t.reilly(a)gmail.com> wrote in message <hf4kbt$46e$1(a)fred.mathworks.com>...
> Hey all,
>
> I am trying to fit parameters to a kinetic model from experimental data through least squares regression, basically just using fmincon, nothing fancy. I am, however, stuck with an issue; MatLab is telling me:
>
> ??? Error using ==> odearguments at 116
> Solving RATE_FUNC requires an initial condition vector of length 8.
>
> The annoying part is that I AM specifying and initial condition vector of length 8:
>
> t_stamp = param.measured_data(:,1);
> x0 = [0; 0; 0.0553; 0; 0; 0; 0; 0];
>
> [t_traj,x_traj] = ode45(@rate_func, t_stamp, x0, [], k2, k3, k4, k5, k6, k7, param);
>
> The variable t_stamp is just taking the time stamps of the measured data points so that the ODE solver returns the function value at those times. The function 'rate_func' contains seven chemical rate equations and eight differential equations; hence why I need an initial condition vector of length 8. The variables k2 to k7 are the parameters that I am varying.
>
> I have tried even tried entering the code as:
>
> [t_traj,x_traj] = ode45(@rate_func, t_stamp, [0; 0; 0.0553; 0; 0; 0; 0; 0], [], k2, k3, k4, k5, k6, k7, param);
>
> but to no avail. I've also tried it a row vector; nothin'. If anyone is able to suggest where I might be going wrong or what may be causing this error, it would be greatly appreciated. I have spent hours trying to solve this. Please message me if it would help to see the complete code.
>
> Thank you!

I don't have a solution, but I'm having the exact same problem
From: Eugene on 4 Dec 2009 03:38
Inside RATE_FUNC are you _returning_ a vector of length eight? I think you have to do that, if you're not.

"Kevin Reilly" <kevin.t.reilly(a)gmail.com> wrote in message <hf4kbt$46e$1(a)fred.mathworks.com>...
> Hey all,
>
> I am trying to fit parameters to a kinetic model from experimental data through least squares regression, basically just using fmincon, nothing fancy. I am, however, stuck with an issue; MatLab is telling me:
>
> ??? Error using ==> odearguments at 116
> Solving RATE_FUNC requires an initial condition vector of length 8.
>
> The annoying part is that I AM specifying and initial condition vector of length 8:
>
> t_stamp = param.measured_data(:,1);
> x0 = [0; 0; 0.0553; 0; 0; 0; 0; 0];
>
> [t_traj,x_traj] = ode45(@rate_func, t_stamp, x0, [], k2, k3, k4, k5, k6, k7, param);
>
> The variable t_stamp is just taking the time stamps of the measured data points so that the ODE solver returns the function value at those times. The function 'rate_func' contains seven chemical rate equations and eight differential equations; hence why I need an initial condition vector of length 8. The variables k2 to k7 are the parameters that I am varying.
>
> I have tried even tried entering the code as:
>
> [t_traj,x_traj] = ode45(@rate_func, t_stamp, [0; 0; 0.0553; 0; 0; 0; 0; 0], [], k2, k3, k4, k5, k6, k7, param);
>
> but to no avail. I've also tried it a row vector; nothin'. If anyone is able to suggest where I might be going wrong or what may be causing this error, it would be greatly appreciated. I have spent hours trying to solve this. Please message me if it would help to see the complete code.
>
> Thank you!
From: Kevin Reilly on 4 Dec 2009 03:47
I solved this issue. It turns out I had made a mistake in specifying the output of my function that was being called by the ODE solver. I was trying to pass back a vector with a single element rather than my vector of differential equations. So yes, you are right, I was not returning a vector of length 8.

Thanks!


"Eugene " <evulfsonREMOVETHIS(a)brandeis.edu> wrote in message <hfahpr$bav$1(a)fred.mathworks.com>...
> Inside RATE_FUNC are you _returning_ a vector of length eight? I think you have to do that, if you're not.
>
> "Kevin Reilly" <kevin.t.reilly(a)gmail.com> wrote in message <hf4kbt$46e$1(a)fred.mathworks.com>...
> > Hey all,
> >
> > I am trying to fit parameters to a kinetic model from experimental data through least squares regression, basically just using fmincon, nothing fancy. I am, however, stuck with an issue; MatLab is telling me:
> >
> > ??? Error using ==> odearguments at 116
> > Solving RATE_FUNC requires an initial condition vector of length 8.
> >
> > The annoying part is that I AM specifying and initial condition vector of length 8:
> >
> > t_stamp = param.measured_data(:,1);
> > x0 = [0; 0; 0.0553; 0; 0; 0; 0; 0];
> >
> > [t_traj,x_traj] = ode45(@rate_func, t_stamp, x0, [], k2, k3, k4, k5, k6, k7, param);
> >
> > The variable t_stamp is just taking the time stamps of the measured data points so that the ODE solver returns the function value at those times. The function 'rate_func' contains seven chemical rate equations and eight differential equations; hence why I need an initial condition vector of length 8. The variables k2 to k7 are the parameters that I am varying.
> >
> > I have tried even tried entering the code as:
> >
> > [t_traj,x_traj] = ode45(@rate_func, t_stamp, [0; 0; 0.0553; 0; 0; 0; 0; 0], [], k2, k3, k4, k5, k6, k7, param);
> >
> > but to no avail. I've also tried it a row vector; nothin'. If anyone is able to suggest where I might be going wrong or what may be causing this error, it would be greatly appreciated. I have spent hours trying to solve this. Please message me if it would help to see the complete code.
> >
> > Thank you!
 | 
Pages: 1
Prev: how to compare variable precision numbers
Next: setting color with data