Average Error: 49.1 → 49.1
Time: 2.8s
Precision: binary64
\[\frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}\]
↓
\[\frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}\]
\frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}
↓
\frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}
double code(double k1, double l, double k2) {
	return ((double) (((double) (((double) (k1 / ((double) pow(l, 5.0)))) * 1.0)) / ((double) (((double) exp(((double) (k2 / l)))) - 1.0))));
}
↓
double code(double k1, double l, double k2) {
	return ((double) (((double) (((double) (k1 / ((double) pow(l, 5.0)))) * 1.0)) / ((double) (((double) exp(((double) (k2 / l)))) - 1.0))));
}

Error

Bits error versus k1

Bits error versus l

Bits error versus k2

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 49.1

    \[\frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}\]

  2. Final simplification49.1

    \[\leadsto \frac{\frac{k1}{{\ell}^{5}} \cdot 1}{e^{\frac{k2}{\ell}} - 1}\]

Reproduce

herbie shell --seed 2020152 
(FPCore (k1 l k2)
  :name "(/ (* (/ k1 (pow l 5)) 1) (- (exp (/ k2 l)) 1))"
  :precision binary64
  (/ (* (/ k1 (pow l 5.0)) 1.0) (- (exp (/ k2 l)) 1.0)))