[["warning", "casting (double + 0.5) to integer leads to incorrect rounding; consider using lround (#include <cmath>) instead", "bugprone-incorrect-roundings"], {"reliability": "high"}]