Average Error: 35.3 → 35.3
Time: 2.6s
Precision: binary64
\[1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}\]
↓
\[1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}\]
1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}
↓
1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}
double code(double j, double n) {
	return ((double) (1.0 - ((double) pow(((double) (((double) (j / ((double) (((double) (n - 1.0)) / 2.0)))) - 1.0)), 2.0))));
}
↓
double code(double j, double n) {
	return ((double) (1.0 - ((double) pow(((double) (((double) (j / ((double) (((double) (n - 1.0)) / 2.0)))) - 1.0)), 2.0))));
}

Error

Bits error versus j

Bits error versus n

Try it out

Your Program's Arguments

Results

Enter valid numbers for all inputs

Derivation

  1. Initial program 35.3

    \[1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}\]

  2. Final simplification35.3

    \[\leadsto 1 - {\left(\frac{j}{\frac{n - 1}{2}} - 1\right)}^{2}\]

Reproduce

herbie shell --seed 2020152 
(FPCore (j n)
  :name "(- 1 (pow (- (/ j (/ (- n 1) 2)) 1) 2))"
  :precision binary64
  (- 1.0 (pow (- (/ j (/ (- n 1.0) 2.0)) 1.0) 2.0)))