pub fn exp_err_e(x: f64, dx: f64) -> (Value, Result)
This function exponentiates x with an associated absolute error dx.