Error using odearguments???

What does this error message mean?

 ??? Error using ==> odearguments at 120
  Inputs must be floats, namely single or double.

Error in ==> ode45 at 172
[neq, tspan, ntspan, next, t0, tfinal, tdir, y0, f0, odeArgs, odeFcn, ...

Error in ==> Testfunction at 32
[t,x] = ode45(@somefn, tspan, x0);

Thanks in advance for any help!

Answer 1

[Jan]

The message means, that the subfunction odearguments recieves non-float values, although it needs this type. The message is actually clear in this point.

Are tspan or x0 both floating point types? You can check this by:

class(tspan)
class(x0)

[Richard]

Thank you, @Jan! I have found that class(x0) is 'sym' by your suggested method.

But I don't know why it is a symbol, what I did was:

a0 = input('a0:');

b0 = input('b0:');

c0 = f(a0,b0) %where f is some function defined earlier in the form f = @(a,b) ...

x0=[a0, b0, c0]

[Jan]

What did you type in at the input prompts? Perhaps variables, which have been declared as SYMs before?

[Richard]

I have written up this question here:

http://www.mathworks.co.uk/matlabcentral/answers/33756-symbols-vs-floating-points

But no, the inputs are numbers and I found that if I deleted c0, MATLAB is quite happy.

I don't understand what's wrong with c0 though. When I run it without c0, I get, for example

{x0 =

10 10}

class(x0)=double

whereas with c0, I get

{x0=

[10, 10, 20]}

,say, and

class(x0)=sym

So Matlab does get numbers, I don't know why the square brackets though and why the class.

Answer 2

[shobhit kumar]

i have the same problem but when i check class of x0 and tspan it is double than why i got the same error