Average Error: 51.1 → 51.1
Time: 2.6s
Precision: binary64
\[\frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}\]
↓
\[\frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}\]
\frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}
↓
\frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}
double code(double x) {
	return ((double) (((double) (2.0 * ((double) (((double) (((double) exp(x)) - 1.0)) - x)))) / ((double) (x * x))));
}
↓
double code(double x) {
	return ((double) (((double) (2.0 * ((double) (((double) (((double) exp(x)) - 1.0)) - x)))) / ((double) (x * x))));
}

Error

Bits error versus x

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 51.1

    \[\frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}\]

  2. Final simplification51.1

    \[\leadsto \frac{2 \cdot \left(\left(e^{x} - 1\right) - x\right)}{x \cdot x}\]

Reproduce

herbie shell --seed 2020152 
(FPCore (x)
  :name "(/ (* 2.0 (- (- (exp x) 1.0) x)) (* x x))"
  :precision binary64
  (/ (* 2.0 (- (- (exp x) 1.0) x)) (* x x)))