Implicit conversion from float -11.489361702127653 to int loses precision