lbfgs::vec_ops
pub fn scalar_mult<T>(a: &mut [T], s: T) where T: Float,
Calculates a scalar times slice: out = s * out