Average Error: 1.0 → 1.0
Time: 6.0s
Precision: binary64
\[2 \cdot \cos \left(\frac{2 \cdot \pi}{3} + \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{3}\right) \]
↓
\[2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{1}{\sqrt{3}} \cdot \frac{\cos^{-1} \left(-\frac{g}{h}\right)}{\sqrt{3}}\right)\right) \]
2 \cdot \cos \left(\frac{2 \cdot \pi}{3} + \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{3}\right)
↓
2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{1}{\sqrt{3}} \cdot \frac{\cos^{-1} \left(-\frac{g}{h}\right)}{\sqrt{3}}\right)\right)

(FPCore (g h)
 :precision binary64
 (* 2.0 (cos (+ (/ (* 2.0 PI) 3.0) (/ (acos (/ (- g) h)) 3.0)))))
↓
(FPCore (g h)
 :precision binary64
 (*
  2.0
  (cos
   (fma
    PI
    0.6666666666666666
    (* (/ 1.0 (sqrt 3.0)) (/ (acos (- (/ g h))) (sqrt 3.0)))))))
double code(double g, double h) {
	return 2.0 * cos(((2.0 * ((double) M_PI)) / 3.0) + (acos(-g / h) / 3.0));
}
↓
double code(double g, double h) {
	return 2.0 * cos(fma(((double) M_PI), 0.6666666666666666, ((1.0 / sqrt(3.0)) * (acos(-(g / h)) / sqrt(3.0)))));
}

Error

Bits error versus g

Bits error versus h

Derivation

  1. Initial program 1.0

    \[2 \cdot \cos \left(\frac{2 \cdot \pi}{3} + \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{3}\right) \]

  2. Simplified1.0

    \[\leadsto \color{blue}{2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{3}\right)\right)} \]

  3. Applied add-sqr-sqrt_binary641.0

    \[\leadsto 2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{\color{blue}{\sqrt{3} \cdot \sqrt{3}}}\right)\right) \]

  4. Applied *-un-lft-identity_binary641.0

    \[\leadsto 2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{\color{blue}{1 \cdot \cos^{-1} \left(\frac{-g}{h}\right)}}{\sqrt{3} \cdot \sqrt{3}}\right)\right) \]

  5. Applied times-frac_binary641.0

    \[\leadsto 2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \color{blue}{\frac{1}{\sqrt{3}} \cdot \frac{\cos^{-1} \left(\frac{-g}{h}\right)}{\sqrt{3}}}\right)\right) \]

  6. Simplified1.0

    \[\leadsto 2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{1}{\sqrt{3}} \cdot \color{blue}{\frac{\cos^{-1} \left(-\frac{g}{h}\right)}{\sqrt{3}}}\right)\right) \]

  7. Final simplification1.0

    \[\leadsto 2 \cdot \cos \left(\mathsf{fma}\left(\pi, 0.6666666666666666, \frac{1}{\sqrt{3}} \cdot \frac{\cos^{-1} \left(-\frac{g}{h}\right)}{\sqrt{3}}\right)\right) \]

Reproduce

herbie shell --seed 2021210 
(FPCore (g h)
  :name "2-ancestry mixing, negative discriminant"
  :precision binary64
  (* 2.0 (cos (+ (/ (* 2.0 PI) 3.0) (/ (acos (/ (- g) h)) 3.0)))))