Implicit conversion from float -138.05232558139534 to int loses precision