Average Error: 39.6 → 0.3
Time: 55.6s
Precision: 64
Internal Precision: 1344
\[\frac{e^{x} - 1}{x}\]
↓
\[\begin{array}{l}
\mathbf{if}\;x \le -0.0001149485681894146:\\
\;\;\;\;\frac{\sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}} \cdot \sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}}{\frac{x}{\sqrt[3]{e^{x} - 1}}}\\
\mathbf{else}:\\
\;\;\;\;\frac{\frac{1}{2} \cdot {x}^{2} + \left(\frac{1}{6} \cdot {x}^{3} + x\right)}{x}\\
\end{array}\]
Target
| Original | 39.6 |
|---|
| Target | 38.8 |
|---|
| Herbie | 0.3 |
|---|
\[\begin{array}{l}
\mathbf{if}\;x \lt 1 \land x \gt -1:\\
\;\;\;\;\frac{e^{x} - 1}{\log \left(e^{x}\right)}\\
\mathbf{else}:\\
\;\;\;\;\frac{e^{x} - 1}{x}\\
\end{array}\]
Derivation
- Split input into 2 regimes
if x < -0.0001149485681894146
Initial program 0.1
\[\frac{e^{x} - 1}{x}\]
- Using strategy
rm Applied flip--0.1
\[\leadsto \frac{\color{blue}{\frac{e^{x} \cdot e^{x} - 1 \cdot 1}{e^{x} + 1}}}{x}\]
Applied simplify0.1
\[\leadsto \frac{\frac{\color{blue}{e^{x + x} - 1}}{e^{x} + 1}}{x}\]
- Using strategy
rm Applied add-cube-cbrt0.1
\[\leadsto \frac{\color{blue}{\left(\sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}} \cdot \sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}\right) \cdot \sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}}}{x}\]
Applied associate-/l*0.1
\[\leadsto \color{blue}{\frac{\sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}} \cdot \sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}}{\frac{x}{\sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}}}}\]
Applied simplify0.1
\[\leadsto \frac{\sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}} \cdot \sqrt[3]{\frac{e^{x + x} - 1}{e^{x} + 1}}}{\color{blue}{\frac{x}{\sqrt[3]{e^{x} - 1}}}}\]
if -0.0001149485681894146 < x
Initial program 60.4
\[\frac{e^{x} - 1}{x}\]
Taylor expanded around 0 0.4
\[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot {x}^{2} + \left(\frac{1}{6} \cdot {x}^{3} + x\right)}}{x}\]
- Recombined 2 regimes into one program.
Runtime
herbie shell --seed '#(1071373924 2949776965 1885069702 3247780810 90874544 2263903749)'
(FPCore (x)
:name "Kahan's exp quotient"
:herbie-target
(if (and (< x 1) (> x -1)) (/ (- (exp x) 1) (log (exp x))) (/ (- (exp x) 1) x))
(/ (- (exp x) 1) x))