fsolve : tolerance ignored
Reported by Alain Lamy
BUG DESCRIPTION:
----------------
The "tolerance" argument does not seem to be taken into account by fsolve.
ERROR LOG:
----------
none
HOW TO REPRODUCE THE BUG:
-------------------------
function [y] = f(x)
y = x * x * x - 8;
endfunction
x0 = 1;
[x1, y1] = fsolve(x0, f); // default tolerance value
[x2, y2] = fsolve(x0, f, 0.1); // explicit tolerance value
disp(x2-x1, y2-y1);
---------------------------------------
Results with Scilab 6.0.2
--> disp(x2-x1, y2-y1);
0.
0.
=> same results (not OK)
---------------------------------------
Results with Scilab 5.5.2
-->disp(x2-x1, y2-y1);
- 0.0224257
- 0.0018706
=> results are not the same (OK)