Implicit conversion from float -20.06896551724138 to int loses precision