Matlab turns scalar parameter into array when numerically integrating

When I numerically integrate a function, for example:

function [ y ] = funci(x)
z = [1 ; 2 ; x];
A = diag(z);
...
end

by writing

integral(@(x)funci(x),0,2)

then the integration will turn x into an array with values from 0 to 2. This doesn't matter for simple integrations but in this case, it causes an error in the second line of the function (the one starting with "z="). Any solutions? I have been stuck on this for the last hours :/ cheers

Answer 1

[Youssef Khmou]

hi Fabian

I think the error( besied i do not own the "integral" function) is that your function returns a matrix , but integral takes an input as one dimensional function then if your function returns 2D data, you need to use other function such as double integral or dblquad,..or you have to map your function from 2D to 1D .

Take this as proposition , not definite answer .

cordially

[Fabian]

Hi, thanks for your answer

The function actually returns a scalar. For example the determinant of A. Hence, this should not be the problem. Moreover, the error occurs in the second line already because Matlab can't fit the array (x) into the vertical array A, even though x should be a scalar.

Cheers