> Is there a particular reason that 0**0 is undefined in GMPL? Of course it > is a reasonable choice to have it like that, but it seems that an as > reasonable choice is to let it evaluate to 1.
> (Maybe this has been up for discussion before, but I just came across it > as a problem now...) 0^0 = 1 is just a technical convention like sqrt(-2) = 0, which may be convenient in some cases. However, in the strong mathematical sense due to discontinuity this convention is incorrect, because it may lead to wrong conclusions; for example, from the identity x^(m-n) = (x^m)/(x^n) it would follow that 0^0 = 0^(m-m) = 0^m/0^m = 0/0 = 1. _______________________________________________ Help-glpk mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-glpk
