double

float is a 32-bit floating-point number; double is 64-bit. Standard-build vectors use float components, while a floating-point Variant uses double. Follow the C# method declaration and account for precision loss when casting.