I'm trying to show that:
$$ 1 + \frac{xy}{e^{\frac{1}{2}(x^2+y^2-2)}} \ge 0, \forall~x,y \in \mathbb{R^2} $$
But I end up with the following inequality:
$$ \ln{xy} - \frac{1}{2}(x^2+y^2-2) \ge \ln{(-1)} $$
Which looks like nonsense to me since $xy$ could easily be negative (say $x=-2, y=+2 \rightarrow xy = -4$).
So I'm guessing I'm approaching this in a wrong way. I've also plotted the function and it's indeed $\ge 0, \forall~x,y \in \mathbb{R^2}$ converging to 1 when $x,y \rightarrow \pm \infty$.
Any tips or sources on how to approach the problem (analytically) would be appreciated. Also, why is it that I end up with such a nonsense expression following the initial approach?
Edit:
- I missed a $\frac{1}{2}$ in the denominator, just added it.
- $\ln{(-1)}$ is not defined in the real numbers, that's what I mean by "nonsense".
- $\ln{(xy)}$ causes the same issue with $xy<0$.