[Octave-bug-tracker] [bug #55134] fplot failing on non-vectorized function

[Rik]

Update of bug #55134 (project octave):

                  Status:                    None => Need Info              

    _______________________________________________________

Follow-up Comment #1:

I'm very surprised that Matlab handles this, but I suppose they must have
written a lot of input validation.

The trouble is not that your original function is not-vectorized, but that the
subfunctions aren't.

The line


fplot(@findX1, [0 12]);


tries to plot the function findX1


function [x1] = findX1(K)
    % FINDX1 Finds the first element of the vector x for a given K
    % Inputs
    %   K - a constant
    % Outputs
    %   x1 - the first element of x
    x = findXall(K);
    x1 = x(1);

end


, but this function depends on findXall


function [x] = findXall(K)
    % FINDXALL Finds the vector x for a given K
    % Inputs
    %   K - a constant
    % Outputs
    %   x - a vector
    A = [5-K -2 4;
         -5 4+K 3;
         5 2 1+K];

    b = [10; 4; -1];

    x = A\b;

end


You can see that the construction of the matrix A is going to fail if K is not
a scalar constant or a row vector.  Can you modify testfplot.m so that you
take the transpose of K in the construction of A and then re-run it under
Matlab?

Here is the new function I would like tested.


function [x] = findXall(K)
    % FINDXALL Finds the vector x for a given K
    % Inputs
    %   K - a constant
    % Outputs
    %   x - a vector
    A = [5-K' -2 4;
         -5 4+K' 3;
         5 2 1+K'];

    b = [10; 4; -1];

    x = A\b;

end


I made that change and have uploaded the file as new_tst_fplot.m.  With this
change, Octave runs fine.  I'm attaching the plot which you can also verify
looks more or less like the one from Matlab.



(file #45555, file #45556)
    _______________________________________________________

Additional Item Attachment:

File name: fplot.png                      Size:17 KB
File name: new_tst_fplot.m                Size:0 KB


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?55134>

_______________________________________________
  Message sent via Savannah
  https://savannah.gnu.org/