ahartmanns / qcomp

qcomp.org – Quantitative Model Repository and Tool Comparison
http://qcomp.org/
6 stars 7 forks source link

MDP Pacman: Fixed a probability in the PRISM file that was slightly off. #100

Closed tquatmann closed 2 years ago

tquatmann commented 4 years ago

The PRISM description was most likely generated with a script that uses floating point arithmetic. As a result, it considered the probability 0.27999999999999997. It is obvious that this should have been 0.28 because otherwise the probabilities doe not sum up to 1. The obtained JANI file was still correct (by accident).