With regard to information gain $($IG$)$ in decision trees, which of the following is true?

Question $3$ options:

$1)$

IG can be larger than $1\mathrm{.}0.$

$2)$

IG equal to $0\mathrm{.}0$ means the creating node is redundant.

$3)$

A negative IG is not possible.

$4)$

The small IG$,$ the better in classification.