Average Error: 5.7 → 0.3
Time: 1.0s
Precision: binary64
\[\frac{1}{e^{\log x - 3}}\]
↓
\[\frac{1}{x} \cdot e^{3}\]
\frac{1}{e^{\log x - 3}}
↓
\frac{1}{x} \cdot e^{3}
double code(double x) {
	return ((double) (1.0 / ((double) exp(((double) (((double) log(x)) - 3.0))))));
}
↓
double code(double x) {
	return ((double) (((double) (1.0 / x)) * ((double) exp(3.0))));
}

Error

Bits error versus x

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 5.7

    \[\frac{1}{e^{\log x - 3}}\]

  2. Simplified0.3

    \[\leadsto \color{blue}{\frac{1}{x} \cdot e^{3}}\]

  3. Final simplification0.3

    \[\leadsto \frac{1}{x} \cdot e^{3}\]

Reproduce

herbie shell --seed 2020153 
(FPCore (x)
  :name "(/ 1 (exp (- (log x) 3)))"
  :precision binary64
  (/ 1.0 (exp (- (log x) 3.0))))