Implicit conversion from float -114.11392405063293 to int loses precision