The generalized inverse Gaussian distribution with parameters
\(\theta\), \(\psi\), and \(\chi\)
has density proportional to
$$
f(x) = x^{\theta-1} \exp\left( -\frac{1}{2} \left(\psi x + \frac{\chi}{x}\right)\right)
$$
where \(\psi>0\) and \(\chi>0\).
An alternative parametrization used parameters
\(\theta\), \(\omega\), and \(\eta\)
and has density proportional to
$$
f(x) = x^{\theta-1} \exp\left( -\frac{\omega}{2} \left(\frac{x}{\eta}+\frac{\eta}{x}\right)\right)
$$
The domain of the distribution can be truncated to the
interval (lb,ub).