Re: [Wien] rotdef problem

2014-10-13 Thread Peter Blaha
The problem in aim comes from the fact, that your struct file contains a z-position which is very "close" to 1.0 (Z=0.8337) and the present default "accuracy check" in aim has problems with it to distinguish it from exactly 1.0. In rotdef1.f change the value of the parameter toler from

Re: [Wien] rotdef problem

2014-10-14 Thread Oliver Albertini
Yes, that seems to be running now with toler=1.d-5, thanks! Excerpts from Peter Blaha's message of 2014-10-13 07:41:21 -0400: > The problem in aim comes from the fact, that your struct file contains a > z-position which is very "close" to 1.0 (Z=0.8337) and the > present default "accurac