Conversion to double from sym is not possible.

Hii,

I have calculated something with matrix and want to plot it after wards. But i got this message from matlab "Conversion to double from sym is not possible".

here is my code. There are a lot of variables, but the point is the plot. Hope that someone can help me

thanx

syms k q a b 

beta=(1/2)*(1+((i*k)/q));
betah=(1/2)*(1+(q/(i*k)));
betahcon=(1/2)*(1-(q/(i*k)));
betacon=(1/2)*(1-((i*k)/q));

Sm=[betah betahcon; betahcon betah];
sp=[beta betacon; betacon beta];
thb=[exp(q*b) 0; 0 exp(-q*b)];
tha=[exp(-q*a) 0; 0 exp(q*a)];
thab=thb*tha;
ta=[exp(i*k*a) 0; 0 exp(-i*k*a)];
tb=[exp(-i*k*b) 0; 0 exp(i*k*b)];

Vh=Sm*thab*sp;
M=tb*Vh*ta;
M1=1./M(2,2);

AA=10^-10;
eV=1.602*10^-19;
a=10*AA;
b=500*AA;
V=1*eV;
m=0.067*9.109*10^-31;
hbar=(6.626*10^-34)/2*pi;
E=linspace(0.001*V,4*V,1000);
q=((sqrt((V-E).*2.*m))./hbar);
k=((sqrt(2.*m.*E))./hbar);

abs(M1).^2 ;
plot(E,M2)

[Smail]

M2=abs(M1).^2

[Smail]

I have also tried this

M2= double(abs(M1).^2) and

M2= vpa(abs(M1).^2)

but it doesnt work

Answer 1

[Walter Roberson]

Remove the assignments

a=10*AA;
b=500*AA;
q=((sqrt((V-E).*2.*m))./hbar);
k=((sqrt(2.*m.*E))./hbar);

Then after assigning to E, use

M1M2 = double( subs({M1, M2}, {a,b,q,k}, {10*AA, 500*AA, ((sqrt((V-E).*2.*m))./hbar), ((sqrt(2.*m.*E))./hbar)}) );

M1n2 = abs(M1M2{1}).^2;

M2n = M1M2{2};
plot(E, M2n);

[Smail]

this does not work.. Can you see what the problem is?

[Walter Roberson]

Sorry "does not work" is not specific enough for me to know what you are seeing.

[Smail]

Ohh, my bad

i get this message from matlab

??? Error using ==> mupadmex

Error in MuPAD command: not a square matrix

[(Dom::Matrix(Dom::ExpressionField()))::_power]

Error in ==> sym.subs>mupadsubs at 152

G = mupadmex('symobj::fullsubs',F.s,X2,Y2);

Error in ==> sym.subs at 127

G = mupadsubs(F,X,Y);

Error in ==> subs at 64

r_unique_name = subs(sym(f_unique_name),varargin{:});

Error in ==> plotenbarrier at 35

M1M2 = double( subs({M1, M2}, {a,b,q,k}, {10*AA, 500*AA,

((sqrt((V-E).*2.*m))./hbar), ((sqrt(2.*m.*E))./hbar)}) );

Answer 2

[Sean de Wolski]

You probably need to subs-titute a value in for your symbolic variable.

doc subs

[Smail]

okay, how do i do that? Can you please show me