symbolic variable

'G' is a function of symbolic variable of 'b' the solve for b gives d=solve(G) a symbolic n*1 matrix but I'm not able to find the min or max even >(greater than sign)of this matrix. Matlab shows this following error Undefined function or method 'gt' for input arguments of type 'sym'.

Answer 1

[Walter Roberson]

re=solve(GG); will return symbolic numbers. You need to apply double() to the symbolic numbers to convert them to floating point numbers.

Answer 2

[Mike]

Since you haven't given explicit code one, can only speculate on the contents of your matrix d. However, here is an explicit example that I believe illustrates the issue. Say you have

>> syms a b c x;
>> results=solve('a*x^2 + b*x + c')

This gives

results =

 -(b + (b^2 - 4*a*c)^(1/2))/(2*a)
 -(b - (b^2 - 4*a*c)^(1/2))/(2*a)

Lets try to find the max of that matrix.

>> max(results)
??? Undefined function or method 'max' for input arguments of type 'sym'.

If you think about it, this should not surprise you since we do not know the values of the symbolic variables a,b and c and the results of max will depend on these values. For example

   a=1;b=1;c=1
   >> y=subs(results)
 y =
  -0.5000 - 0.8660i
  -0.5000 + 0.8660i
>> max(y)

 ans =
  -0.5000 + 0.8660i

So for a=1;b=1;c=1, the second element of results is the maximum. However, for a=-1;b=1;c=1, the first element of results is the maximum:

>> a=-1;b=1;c=1;
>> y=subs(results)

y =
  1.6180
 -0.6180

>> max(y)

ans =
  1.6180

Hope this helps, Mike

[Hassan]

Dear Mike

x1=sym('x1','real');

x2=sym('x2','real');

x3=sym('x3','real');

z1=sym('z1','real');

z2=sym('z2','real');

z3=sym('z3','real');

b=sym('b','real');

alpha=zeros(1,3);

m=[95,50,4000];

sig=[10,2.5,1000];

r=[-(1/3)^.5,(1/3)^.5,-(1/3)^.5];

digits(4)

x=[x1,x2,x3];

z=[z1,z2,z3];

for i=1:3;

x(1,i)=sig(1,i).*z(1,i)+m(1,i);

end

g=1.25*x(1,1).*x(1,2)-x(1,3)./2;

G=vpa(expand(subs(g)));

grad1=diff(G,'z1');

grad2=diff(G,'z2');

grad3=diff(G,'z3');

bb=0;

for i=1:10;

z1=alpha(1,1).*bb;

z2=alpha(1,2).*bb;

z3=alpha(1,3).*bb;

grad=[subs(grad1),subs(grad2),subs(grad3)];

k=(grad*grad')^.5;

for j=1:3;

alpha(1,j)=grad(1,j)/k;

end

z1=-alpha(1,1).*b;

z2=-alpha(1,2).*b;

z3=-alpha(1,3).*b;

GG=subs(G);

re=solve(GG);

a=max(size(b));

for k=1:a;

if re(k,1)>re(1,1);

y=re(1,1);

else

y=re(k,1);

end

end

end

this is my m.file