rint

float  rint(float n)

<vector> rint(<vector>v)

Returns the closest whole number to n. For vectors, this is done per-component.