fzero raises error before setting exitflag

Question

Fitzgerald on 21 Jan 2013

0
Link

Direct link to this question

https://nl.mathworks.com/matlabcentral/answers/59385-fzero-raises-error-before-setting-exitflag

I have a function fh of which I would like to calculate the zero within a certain interval. If there is no zero in the interval, then I'd like to do something else.

The following code does not work:

fh = @(x) x + 1;
try
    [ds, ~, exitflag] = fzero(fh, [0 10]);
catch
end
switch exitflag
    case 1
        fprintf('found a solution\n')
    case -6
        fprintf('no solution in interval\n')
    otherwise
        error('something is wrong')
end

The problem is that fzero raises an error: "The function values at the interval endpoints must differ in sign," and doing so before setting the exitflag. I can catch the exception raised by fzero, but then there is no way to tell the difference between a true error or that there was just no zero in the specified interval.

How can I use exitflag, if the only possible value of exitflag is 1 (the case that no error is raised by fzero)?

Thanks!