floor

float|int floor(float n)

<vector> floor(<vector>v)

Returns the largest integer less than or equal to the argument. For vectors, this is done per-component.