Implicit conversion from float -36.81638044914135 to int loses precision