Matlab newbie

  1. I have homework due and no help from the TA's due to a language problem.

    How do you multiply a scalar with a matrix? It seems so simple, but I keep getting errors, such as "matrix must be square".

    .m funtion code:

    function X=parabola(a,x,b)
    X=(x*a+b)^2;

    Command window code :

    EDU>> x=0:10:100;
    EDU>> a=2;
    EDU>> b=4;
    EDU>> f=parabola(x,a,b)
    ??? Error using ==> mpower
    Matrix must be square.

    Error in ==> parabola at 2
    X=(x*a+b)^2;
    EDU>>

    so now we try

    EDU>> f=parabola(x',a,b)
    ??? Error using ==> mpower
    Matrix must be square.

    Error in ==> parabola at 2
    X=(x*a+b)^2;
    EDU>>

    I know it can't be that hard, but the help section does not address this explicitly, and I have not been able to use logic to find out where I am going wrong. By the way, this is Matlab 7.0, Student version.

    Thank You
    Bill
     
  2. jcsd
  3. I used many days ago.......

    I can say... in matlab everything 1 dimensional array is considered as vector and two are matrix....here may be you've to use a dot (.) before or after *
     
  4. enigma

    enigma 1,815
    Staff Emeritus
    Science Advisor
    Gold Member

    The problem isn't with the scalar. You can add and multiply scalars to vectors with no problems.

    x is a 1 by 11 vector.

    You are trying to multiply a 1 by 11 vector by a 1 by 11 vector, which isn't possible.

    What you're trying to do is have each component of the vector squared.

    Anytime you need to do this, whether you're using *, /, or ^, you need to prefix the command with a '.'

    You want your code to look like this:
    Code (Text):

    function X=parabola(a,x,b)
    X=(x*a+b).^2;
     
    The period means do a point by point operation.
     
    Last edited: Oct 16, 2004
  5. Enigma,

    Thank you. That is exactly what I was trying to do. I even went so far as to transpose the matrix, but that did not work. Using the '.' behind the parentheses instead of inside the parentheses is what did the trick.

    Now another question. How do you code this problem:

    code in the .m.file, by line item

    1) function [opt] = goldmax(xlow,xhigh,maxit,es,fx)
    .
    .
    9) f1=@(fx)xl;
    10) f2=@(fx)xu;
    11) if (f1>=f2)
    .
    .
    18) if (k<=maxit)

    and the code in the command window is:

    EDU>> t=goldmax(5,7,50,1.e-4,'parabola')
    ??? Function 'ge' is not defined for values of class 'function_handle'.

    Error in ==> ge at 18
    [varargout{1:nargout}] = builtin('ge', varargin{:});

    Error in ==> goldmax at 11
    if (f1>=f2)


    'fx' is a handle for a function, 'xl' and 'xu' are the inputs to the function, but 'f1' and 'f2' are variables, and 'k' is my counter, which equals 1 at this point. I think I am following the syntax in the Matlab help guide, but I must be missing something.

    Thanks again

    Bill

    :smile:
     
  6. enigma

    enigma 1,815
    Staff Emeritus
    Science Advisor
    Gold Member

    I've never used the '@' symbol in any of my code. The only time I can see it being useful is if you want to have a function call another function which can be changed using the input arguments. Is that what you're using it for here?

    The only time I've ever used handles is with graphics, file handling, and GUIs. All of them use handles as returns from functions, so I have never needed to code the function handles explicitly.

    Lines 9 & 10,

    Code (Text):

    f1=@(fx)xl;
    f2=@(fx)xu;
     
    don't look like typical Matlab syntax to me.

    What exactly do you want this code to accomplish?
     
    Last edited: Oct 20, 2004
  7. Simply multiply the matrix with a dot...

    ie

    >> A = [1 2 3;4 5 6; 7 8 9]

    A =

    1 2 3
    4 5 6
    7 8 9

    then multiply A by 3

    >> A.*3

    ans =

    3 6 9
    12 15 18
    21 24 27

    see the dot? hope that helped
     
  8. enigma

    enigma 1,815
    Staff Emeritus
    Science Advisor
    Gold Member

    You don't need the dot for scalar-vector multiplication.

    You only need it if you want to multiply the ith component of a vector with the ith component of another vector, because the '*' operator is defined as a vector multiplication.
     
Know someone interested in this topic? Share this thead via email, Google+, Twitter, or Facebook

Have something to add?