Hello,
I'm using the PointEstimateMethod package and found out a strange behaviour with a particular choice of the distribution. In particular, when extracting 3 points from a truncated normal distribution $N(\mu,\sigma)$ in $[0,+\infty]$ with $\mu = 1.0$ and $\sigma = 0.2$, sometimes it returns the points but other times it returns an error. I have also created a file (attached) in order to reproduce the error and understand how many times it returns an error repeating the extraction 1000 times.
Thank you for your time.
minimal_error_pem.txt
