[CppAD] LuSolve

[Brad Bell]

The infinities in the original matrix that you were factoring were the 
source of the problem (see Matrix Values below). The corresponding error 
message has been changed to
    LuFactor can't determine an element with maximum absolute value.
    Perhaps original matrix contains not a number or infinity.
    Perhaps your specialization of AbsGeq is not correct.

Source code changes:
    https://projects.coin-or.org/CppAD/changeset/1132
    https://projects.coin-or.org/CppAD/changeset/1133


Matrix values:
{
    1.0547634880448762,
    0.71587033175396086,
    0.54731272363254424,
    .
    .
    .
    ,
    2.0377661555509993e+245,
    4.8284883691331531e+244,
    1.0908259726547921e+245,
    inf,
    inf,
    inf,
    inf,
    inf,
     inf,
     inf,
    inf
}




Mitch Watrous wrote:
> Hi.
>
> When the factors created by LuSolve contain NaN's, it should return the 
> signdet equal to 0.
>
> But in fact it sometimes generates the following assert:
>
> 		CPPAD_ASSERT_KNOWN(
> 			(imax < n) & (jmax < n) ,
> 			"AbsGeq must return true when second argument is zero"
>
>
> Thanks,
> Mitch
> _______________________________________________
> CppAD mailing list
> CppAD at list.coin-or.org
> http://list.coin-or.org/mailman/listinfo/cppad
>
>
>