Average Error: 58.6 → 0.6
Time: 8.1s
Precision: binary64
\[\log \left(\frac{1 - \varepsilon}{1 + \varepsilon}\right)\]
↓
\[\varepsilon \cdot -2 + \log 1 \cdot \left(1 + {\varepsilon}^{3}\right)\]
\log \left(\frac{1 - \varepsilon}{1 + \varepsilon}\right)
↓
\varepsilon \cdot -2 + \log 1 \cdot \left(1 + {\varepsilon}^{3}\right)
(FPCore (eps) :precision binary64 (log (/ (- 1.0 eps) (+ 1.0 eps))))
↓
(FPCore (eps)
 :precision binary64
 (+ (* eps -2.0) (* (log 1.0) (+ 1.0 (pow eps 3.0)))))
double code(double eps) {
	return log((1.0 - eps) / (1.0 + eps));
}
↓
double code(double eps) {
	return (eps * -2.0) + (log(1.0) * (1.0 + pow(eps, 3.0)));
}

Error

Bits error versus eps

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Target

Original58.6
Target0.2
Herbie0.6
\[-2 \cdot \left(\left(\varepsilon + \frac{{\varepsilon}^{3}}{3}\right) + \frac{{\varepsilon}^{5}}{5}\right)\]

Derivation

  1. Initial program 58.6

    \[\log \left(\frac{1 - \varepsilon}{1 + \varepsilon}\right)\]

  2. Taylor expanded around 0 0.6

    \[\leadsto \color{blue}{-2 \cdot \varepsilon + \log 1 \cdot \left(1 + {\varepsilon}^{3}\right)}\]

  3. Simplified0.6

    \[\leadsto \color{blue}{\varepsilon \cdot -2 + \log 1 \cdot \left(1 + {\varepsilon}^{3}\right)}\]

  4. Final simplification0.6

    \[\leadsto \varepsilon \cdot -2 + \log 1 \cdot \left(1 + {\varepsilon}^{3}\right)\]

Reproduce

herbie shell --seed 2020268 
(FPCore (eps)
  :name "logq (problem 3.4.3)"
  :precision binary64

  :herbie-target
  (* -2.0 (+ (+ eps (/ (pow eps 3.0) 3.0)) (/ (pow eps 5.0) 5.0)))

  (log (/ (- 1.0 eps) (+ 1.0 eps))))