Nan returned when multiplying large zero vector by its transpose. This is my
Reported by Kenn Sebesta
Originally assigned to Allan CORNET
-- Bug description --
Nan returned when multiplying large zero vector by its transpose.
This is my output on my Windows 2000 machine running 5.0.2. I was alerted to this problem during a training session, and saw it recreated on at least two other machines, both running 5.0.1.
Repeatedly running the same command yields different locations for the Nan.
On this machine, Scilab 4.2gtk does not have this problem. I did not check 4.2gtk on the other machines to see if the problem is machine specific or not. However, it seems not to be.
This is a sample output:
=============================================================
-->A=zeros(1000,1);
-->A*A'
ans =
column 1 to 12
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. Nan 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. Nan 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. Nan 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
=============================================================
-- Scilab error message --
-- How to reproduce the bug --
A=zeros(1000,1);
A*A';