pub fn tanfull(x: f32) -> f32
Tangent in radians.
The range reduction technique used here will be hopelessly inaccurate for |x| >> 1000.