| Re: band gap accuracy ( No.1 )|
- Date: 2021/12/27 16:06
- Name: Pavel Ondracka <firstname.lastname@example.org>
- This is highly material dependent and could be pretty much anything. My experience is that standard GGAs like PBE underestimate the quasiparticle band gap by ~50%. So in this regard the number of decimal places is IMO irrelevant. If you are instead asking about the precision of the calculations than this depends on your numerical settings, i.e., k-point density, cutoff energy, basis set, etc.
See for example the Figure 1 in "K. Lejaeghere et al., Reproducibility in density functional theory calculations of solids, Science (80-. ). 351, aad3000 (2016)" for the difference between the accuracy and precision and to understand why you question doesn't make much sense.