I'm considering the following Robin problem. $$ \begin{cases}\Delta u-Au=-\lambda_{1}u & \text { on } \Omega \\ \frac{\partial u}{\partial n}+Bu=0 & \text { on } \partial \Omega.\end{cases} $$
Here $A,B$ are constans,$\lambda_{1}$ is the first eigenvalue.$\Omega$ is a domain with smooth boundary.
I wonder why $u$ is strictly positive on $\Omega$.I know we only need to show it doesn't change sign.
In fact I know $u$ doesn't change sign when it's the first eigenfunction of laplacian.But I don't know how to prove the same conclusion in this kind of operator and boundary condition.
Thanks in advance.