"Subscript indices must either be real positive integers or logicals" error from regula falsi function

Hello All,

I am need in need of some help with creating this regula falsi function. Below is the function I have created along with the definitions for the variables in the function, however when trying to run in the command prompt, I get the error "Subscript indices must either be real positive integers or logicals" and points to the line with f(b) = fun(b). I don't really understand what this error means and what exactly its looking at.

FYI, I am new to matlab, so any help would be much appreciated.

Thank you.

fun = @(in)(in^.2 - 5); %defined in the command window

function [ x ] = falsi(fun,a,b,epsilon) %functions I created

disp('   I          a          b         x');            
disp('=======    =======    =======    ======');
k=0;

f(a)=fun(a);
f(b)=fun(b);

%t=linspace(b,a,100);
%f = fun(t);
%plot(t,f);
%hold on 
%set(fig,'color','white')
%grid on 

while f(x) > epsilon
    k=k+1;
    x = (a*f(b) - b*f(a))/(f(b)-f(a));
    f(x)=fun(x);
    fprintf('%5i,%11.4f;%11.4f,%11.4f\n',k,a,b,x);
    if f(a)*f(x)< 0
        a = x;
        f(a) = f(x);        
    elseif f(a)*f(x)> 0
        b = x;
        f(b) = f(x);
    else 
        break
    end    
   end

[Adam]

How are you calling the function?

This error often implies that Matlab sees 'fun' as being a matrix to index into rather than a function. One of the joys of Matlab using the same syntax for function calls and for array indexing.

In this case it is probably 'b' that it has a problem with, at a guess, as an index into f.

[CAAJ]

I am calling the fun as:

fun = 0;
fun = @(in)(in^.2-5);

and as for the falsi function I am just calling it as:

falsi(fun,a,b,epsilon);

Answer 1

[Guillaume]

I don't understand what you're trying to do with the statements f(b) = fun(b) and similar. This assigns the result of fun at value b to the element at index b of matrix (or vector) f. Of course, since indices must be positive integers you get an error if b is not.

As said, I don't understand why you're trying to store these values at arbirtrary indices of a matrix. Two possible ways to fix your code:

• replace all instances of f(a) by f(1), f(b) by f(2) and f(x) by f(3). The values will be stored in a 3 element vector
• replace all instances of f(a) by fa, f(b) by fb and f(x) by fx. The values will be stored in individual variables