Converting to int makes the * 2 + 1 operations done in integer arithmetics, which probably ends up saving a few cpu cycles.

Also prevents warnings about losing precision when implicitly converting from double to int.