Implicit conversion from float -22.25 to int loses precision