I would suggest writing this as
p2s(x, y) = (360 + x) % 360, (360 + y) % 360
It will work equally well for Float64 input, but much better for Int or Float32 etc. If you give single precision inputs they will be promoted to double, for example, which is probably undesirable. For integers there is also a large performance loss if you don’t do this.
In general, it’s better to use integer literals everywhere, unless there’s some specific reason not to.