Implicit conversion from float -0.2608695652174049 to int loses precision