Falkner and Boettcher, Equation (22+)

Percentage Accurate: 98.5% → 100.0%
Time: 5.7s
Alternatives: 6
Speedup: 1.1×

Specification

?
\[\begin{array}{l} \\ \frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/ 4.0 (* (* (* 3.0 PI) (- 1.0 (* v v))) (sqrt (- 2.0 (* 6.0 (* v v)))))))
double code(double v) {
	return 4.0 / (((3.0 * ((double) M_PI)) * (1.0 - (v * v))) * sqrt((2.0 - (6.0 * (v * v)))));
}
public static double code(double v) {
	return 4.0 / (((3.0 * Math.PI) * (1.0 - (v * v))) * Math.sqrt((2.0 - (6.0 * (v * v)))));
}
def code(v):
	return 4.0 / (((3.0 * math.pi) * (1.0 - (v * v))) * math.sqrt((2.0 - (6.0 * (v * v)))))
function code(v)
	return Float64(4.0 / Float64(Float64(Float64(3.0 * pi) * Float64(1.0 - Float64(v * v))) * sqrt(Float64(2.0 - Float64(6.0 * Float64(v * v))))))
end
function tmp = code(v)
	tmp = 4.0 / (((3.0 * pi) * (1.0 - (v * v))) * sqrt((2.0 - (6.0 * (v * v)))));
end
code[v_] := N[(4.0 / N[(N[(N[(3.0 * Pi), $MachinePrecision] * N[(1.0 - N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[Sqrt[N[(2.0 - N[(6.0 * N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}}
\end{array}

Sampling outcomes in binary64 precision:

Local Percentage Accuracy vs ?

The average percentage accuracy by input value. Horizontal axis shows value of an input variable; the variable is choosen in the title. Vertical axis is accuracy; higher is better. Red represent the original program, while blue represents Herbie's suggestion. These can be toggled with buttons below the plot. The line is an average while dots represent individual samples.

Accuracy vs Speed?

Herbie found 6 alternatives:

AlternativeAccuracySpeedup
The accuracy (vertical axis) and speed (horizontal axis) of each alternatives. Up and to the right is better. The red square shows the initial program, and each blue circle shows an alternative.The line shows the best available speed-accuracy tradeoffs.

Initial Program: 98.5% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/ 4.0 (* (* (* 3.0 PI) (- 1.0 (* v v))) (sqrt (- 2.0 (* 6.0 (* v v)))))))
double code(double v) {
	return 4.0 / (((3.0 * ((double) M_PI)) * (1.0 - (v * v))) * sqrt((2.0 - (6.0 * (v * v)))));
}
public static double code(double v) {
	return 4.0 / (((3.0 * Math.PI) * (1.0 - (v * v))) * Math.sqrt((2.0 - (6.0 * (v * v)))));
}
def code(v):
	return 4.0 / (((3.0 * math.pi) * (1.0 - (v * v))) * math.sqrt((2.0 - (6.0 * (v * v)))))
function code(v)
	return Float64(4.0 / Float64(Float64(Float64(3.0 * pi) * Float64(1.0 - Float64(v * v))) * sqrt(Float64(2.0 - Float64(6.0 * Float64(v * v))))))
end
function tmp = code(v)
	tmp = 4.0 / (((3.0 * pi) * (1.0 - (v * v))) * sqrt((2.0 - (6.0 * (v * v)))));
end
code[v_] := N[(4.0 / N[(N[(N[(3.0 * Pi), $MachinePrecision] * N[(1.0 - N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[Sqrt[N[(2.0 - N[(6.0 * N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}}
\end{array}

Alternative 1: 100.0% accurate, 0.7× speedup?

\[\begin{array}{l} \\ \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + v \cdot \left(v \cdot -6\right)\right)}^{1.5}\right)}^{0.3333333333333333}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/
  (/ 1.3333333333333333 (* PI (- 1.0 (* v v))))
  (pow (pow (+ 2.0 (* v (* v -6.0))) 1.5) 0.3333333333333333)))
double code(double v) {
	return (1.3333333333333333 / (((double) M_PI) * (1.0 - (v * v)))) / pow(pow((2.0 + (v * (v * -6.0))), 1.5), 0.3333333333333333);
}
public static double code(double v) {
	return (1.3333333333333333 / (Math.PI * (1.0 - (v * v)))) / Math.pow(Math.pow((2.0 + (v * (v * -6.0))), 1.5), 0.3333333333333333);
}
def code(v):
	return (1.3333333333333333 / (math.pi * (1.0 - (v * v)))) / math.pow(math.pow((2.0 + (v * (v * -6.0))), 1.5), 0.3333333333333333)
function code(v)
	return Float64(Float64(1.3333333333333333 / Float64(pi * Float64(1.0 - Float64(v * v)))) / ((Float64(2.0 + Float64(v * Float64(v * -6.0))) ^ 1.5) ^ 0.3333333333333333))
end
function tmp = code(v)
	tmp = (1.3333333333333333 / (pi * (1.0 - (v * v)))) / (((2.0 + (v * (v * -6.0))) ^ 1.5) ^ 0.3333333333333333);
end
code[v_] := N[(N[(1.3333333333333333 / N[(Pi * N[(1.0 - N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / N[Power[N[Power[N[(2.0 + N[(v * N[(v * -6.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 1.5], $MachinePrecision], 0.3333333333333333], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + v \cdot \left(v \cdot -6\right)\right)}^{1.5}\right)}^{0.3333333333333333}}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Step-by-step derivation
    1. associate-/r*100.0%

      \[\leadsto \color{blue}{\frac{\frac{4}{\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
    2. associate-*l*100.0%

      \[\leadsto \frac{\frac{4}{\color{blue}{3 \cdot \left(\pi \cdot \left(1 - v \cdot v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    3. sqr-neg100.0%

      \[\leadsto \frac{\frac{4}{3 \cdot \left(\pi \cdot \left(1 - \color{blue}{\left(-v\right) \cdot \left(-v\right)}\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    4. associate-/r*100.0%

      \[\leadsto \frac{\color{blue}{\frac{\frac{4}{3}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    5. metadata-eval100.0%

      \[\leadsto \frac{\frac{\color{blue}{1.3333333333333333}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    6. sqr-neg100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - \color{blue}{v \cdot v}\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  3. Simplified100.0%

    \[\leadsto \color{blue}{\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
  4. Step-by-step derivation
    1. add-cbrt-cube98.5%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\color{blue}{\sqrt[3]{\left(\sqrt{2 - 6 \cdot \left(v \cdot v\right)} \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}}}} \]
    2. pow1/3100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\color{blue}{{\left(\left(\sqrt{2 - 6 \cdot \left(v \cdot v\right)} \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}\right)}^{0.3333333333333333}}} \]
    3. add-sqr-sqrt100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left(\color{blue}{\left(2 - 6 \cdot \left(v \cdot v\right)\right)} \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}\right)}^{0.3333333333333333}} \]
    4. pow1100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left(\color{blue}{{\left(2 - 6 \cdot \left(v \cdot v\right)\right)}^{1}} \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}\right)}^{0.3333333333333333}} \]
    5. pow1/2100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 - 6 \cdot \left(v \cdot v\right)\right)}^{1} \cdot \color{blue}{{\left(2 - 6 \cdot \left(v \cdot v\right)\right)}^{0.5}}\right)}^{0.3333333333333333}} \]
    6. pow-prod-up100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\color{blue}{\left({\left(2 - 6 \cdot \left(v \cdot v\right)\right)}^{\left(1 + 0.5\right)}\right)}}^{0.3333333333333333}} \]
    7. sub-neg100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\color{blue}{\left(2 + \left(-6 \cdot \left(v \cdot v\right)\right)\right)}}^{\left(1 + 0.5\right)}\right)}^{0.3333333333333333}} \]
    8. *-commutative100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + \left(-\color{blue}{\left(v \cdot v\right) \cdot 6}\right)\right)}^{\left(1 + 0.5\right)}\right)}^{0.3333333333333333}} \]
    9. distribute-rgt-neg-in100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + \color{blue}{\left(v \cdot v\right) \cdot \left(-6\right)}\right)}^{\left(1 + 0.5\right)}\right)}^{0.3333333333333333}} \]
    10. metadata-eval100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + \left(v \cdot v\right) \cdot \color{blue}{-6}\right)}^{\left(1 + 0.5\right)}\right)}^{0.3333333333333333}} \]
    11. associate-*r*100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + \color{blue}{v \cdot \left(v \cdot -6\right)}\right)}^{\left(1 + 0.5\right)}\right)}^{0.3333333333333333}} \]
    12. metadata-eval100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + v \cdot \left(v \cdot -6\right)\right)}^{\color{blue}{1.5}}\right)}^{0.3333333333333333}} \]
  5. Applied egg-rr100.0%

    \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\color{blue}{{\left({\left(2 + v \cdot \left(v \cdot -6\right)\right)}^{1.5}\right)}^{0.3333333333333333}}} \]
  6. Final simplification100.0%

    \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{{\left({\left(2 + v \cdot \left(v \cdot -6\right)\right)}^{1.5}\right)}^{0.3333333333333333}} \]

Alternative 2: 100.0% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/
  (/ 1.3333333333333333 (* PI (- 1.0 (* v v))))
  (sqrt (- 2.0 (* (* v v) 6.0)))))
double code(double v) {
	return (1.3333333333333333 / (((double) M_PI) * (1.0 - (v * v)))) / sqrt((2.0 - ((v * v) * 6.0)));
}
public static double code(double v) {
	return (1.3333333333333333 / (Math.PI * (1.0 - (v * v)))) / Math.sqrt((2.0 - ((v * v) * 6.0)));
}
def code(v):
	return (1.3333333333333333 / (math.pi * (1.0 - (v * v)))) / math.sqrt((2.0 - ((v * v) * 6.0)))
function code(v)
	return Float64(Float64(1.3333333333333333 / Float64(pi * Float64(1.0 - Float64(v * v)))) / sqrt(Float64(2.0 - Float64(Float64(v * v) * 6.0))))
end
function tmp = code(v)
	tmp = (1.3333333333333333 / (pi * (1.0 - (v * v)))) / sqrt((2.0 - ((v * v) * 6.0)));
end
code[v_] := N[(N[(1.3333333333333333 / N[(Pi * N[(1.0 - N[(v * v), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / N[Sqrt[N[(2.0 - N[(N[(v * v), $MachinePrecision] * 6.0), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Step-by-step derivation
    1. associate-/r*100.0%

      \[\leadsto \color{blue}{\frac{\frac{4}{\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
    2. associate-*l*100.0%

      \[\leadsto \frac{\frac{4}{\color{blue}{3 \cdot \left(\pi \cdot \left(1 - v \cdot v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    3. sqr-neg100.0%

      \[\leadsto \frac{\frac{4}{3 \cdot \left(\pi \cdot \left(1 - \color{blue}{\left(-v\right) \cdot \left(-v\right)}\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    4. associate-/r*100.0%

      \[\leadsto \frac{\color{blue}{\frac{\frac{4}{3}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    5. metadata-eval100.0%

      \[\leadsto \frac{\frac{\color{blue}{1.3333333333333333}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    6. sqr-neg100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - \color{blue}{v \cdot v}\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  3. Simplified100.0%

    \[\leadsto \color{blue}{\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
  4. Final simplification100.0%

    \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \]

Alternative 3: 98.9% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \frac{1.3333333333333333 \cdot \frac{1}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/ (* 1.3333333333333333 (/ 1.0 PI)) (sqrt (- 2.0 (* (* v v) 6.0)))))
double code(double v) {
	return (1.3333333333333333 * (1.0 / ((double) M_PI))) / sqrt((2.0 - ((v * v) * 6.0)));
}
public static double code(double v) {
	return (1.3333333333333333 * (1.0 / Math.PI)) / Math.sqrt((2.0 - ((v * v) * 6.0)));
}
def code(v):
	return (1.3333333333333333 * (1.0 / math.pi)) / math.sqrt((2.0 - ((v * v) * 6.0)))
function code(v)
	return Float64(Float64(1.3333333333333333 * Float64(1.0 / pi)) / sqrt(Float64(2.0 - Float64(Float64(v * v) * 6.0))))
end
function tmp = code(v)
	tmp = (1.3333333333333333 * (1.0 / pi)) / sqrt((2.0 - ((v * v) * 6.0)));
end
code[v_] := N[(N[(1.3333333333333333 * N[(1.0 / Pi), $MachinePrecision]), $MachinePrecision] / N[Sqrt[N[(2.0 - N[(N[(v * v), $MachinePrecision] * 6.0), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{1.3333333333333333 \cdot \frac{1}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Step-by-step derivation
    1. associate-/r*100.0%

      \[\leadsto \color{blue}{\frac{\frac{4}{\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
    2. associate-*l*100.0%

      \[\leadsto \frac{\frac{4}{\color{blue}{3 \cdot \left(\pi \cdot \left(1 - v \cdot v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    3. sqr-neg100.0%

      \[\leadsto \frac{\frac{4}{3 \cdot \left(\pi \cdot \left(1 - \color{blue}{\left(-v\right) \cdot \left(-v\right)}\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    4. associate-/r*100.0%

      \[\leadsto \frac{\color{blue}{\frac{\frac{4}{3}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    5. metadata-eval100.0%

      \[\leadsto \frac{\frac{\color{blue}{1.3333333333333333}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    6. sqr-neg100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - \color{blue}{v \cdot v}\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  3. Simplified100.0%

    \[\leadsto \color{blue}{\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
  4. Taylor expanded in v around 0 99.2%

    \[\leadsto \frac{\color{blue}{\frac{1.3333333333333333}{\pi}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  5. Step-by-step derivation
    1. div-inv99.2%

      \[\leadsto \frac{\color{blue}{1.3333333333333333 \cdot \frac{1}{\pi}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  6. Applied egg-rr99.2%

    \[\leadsto \frac{\color{blue}{1.3333333333333333 \cdot \frac{1}{\pi}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  7. Final simplification99.2%

    \[\leadsto \frac{1.3333333333333333 \cdot \frac{1}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \]

Alternative 4: 98.9% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \frac{\frac{1.3333333333333333}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \end{array} \]
(FPCore (v)
 :precision binary64
 (/ (/ 1.3333333333333333 PI) (sqrt (- 2.0 (* (* v v) 6.0)))))
double code(double v) {
	return (1.3333333333333333 / ((double) M_PI)) / sqrt((2.0 - ((v * v) * 6.0)));
}
public static double code(double v) {
	return (1.3333333333333333 / Math.PI) / Math.sqrt((2.0 - ((v * v) * 6.0)));
}
def code(v):
	return (1.3333333333333333 / math.pi) / math.sqrt((2.0 - ((v * v) * 6.0)))
function code(v)
	return Float64(Float64(1.3333333333333333 / pi) / sqrt(Float64(2.0 - Float64(Float64(v * v) * 6.0))))
end
function tmp = code(v)
	tmp = (1.3333333333333333 / pi) / sqrt((2.0 - ((v * v) * 6.0)));
end
code[v_] := N[(N[(1.3333333333333333 / Pi), $MachinePrecision] / N[Sqrt[N[(2.0 - N[(N[(v * v), $MachinePrecision] * 6.0), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{\frac{1.3333333333333333}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Step-by-step derivation
    1. associate-/r*100.0%

      \[\leadsto \color{blue}{\frac{\frac{4}{\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
    2. associate-*l*100.0%

      \[\leadsto \frac{\frac{4}{\color{blue}{3 \cdot \left(\pi \cdot \left(1 - v \cdot v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    3. sqr-neg100.0%

      \[\leadsto \frac{\frac{4}{3 \cdot \left(\pi \cdot \left(1 - \color{blue}{\left(-v\right) \cdot \left(-v\right)}\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    4. associate-/r*100.0%

      \[\leadsto \frac{\color{blue}{\frac{\frac{4}{3}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    5. metadata-eval100.0%

      \[\leadsto \frac{\frac{\color{blue}{1.3333333333333333}}{\pi \cdot \left(1 - \left(-v\right) \cdot \left(-v\right)\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
    6. sqr-neg100.0%

      \[\leadsto \frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - \color{blue}{v \cdot v}\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  3. Simplified100.0%

    \[\leadsto \color{blue}{\frac{\frac{1.3333333333333333}{\pi \cdot \left(1 - v \cdot v\right)}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}}} \]
  4. Taylor expanded in v around 0 99.2%

    \[\leadsto \frac{\color{blue}{\frac{1.3333333333333333}{\pi}}}{\sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  5. Final simplification99.2%

    \[\leadsto \frac{\frac{1.3333333333333333}{\pi}}{\sqrt{2 - \left(v \cdot v\right) \cdot 6}} \]

Alternative 5: 98.8% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \frac{1.3333333333333333}{\pi} \cdot \sqrt{0.5} \end{array} \]
(FPCore (v) :precision binary64 (* (/ 1.3333333333333333 PI) (sqrt 0.5)))
double code(double v) {
	return (1.3333333333333333 / ((double) M_PI)) * sqrt(0.5);
}
public static double code(double v) {
	return (1.3333333333333333 / Math.PI) * Math.sqrt(0.5);
}
def code(v):
	return (1.3333333333333333 / math.pi) * math.sqrt(0.5)
function code(v)
	return Float64(Float64(1.3333333333333333 / pi) * sqrt(0.5))
end
function tmp = code(v)
	tmp = (1.3333333333333333 / pi) * sqrt(0.5);
end
code[v_] := N[(N[(1.3333333333333333 / Pi), $MachinePrecision] * N[Sqrt[0.5], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\frac{1.3333333333333333}{\pi} \cdot \sqrt{0.5}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Simplified100.0%

    \[\leadsto \color{blue}{\frac{-1.3333333333333333}{\mathsf{fma}\left(v, v, -1\right) \cdot \left(\pi \cdot \sqrt{\mathsf{fma}\left(v, v \cdot -6, 2\right)}\right)}} \]
  3. Taylor expanded in v around 0 97.7%

    \[\leadsto \color{blue}{1.3333333333333333 \cdot \frac{\sqrt{0.5}}{\pi}} \]
  4. Step-by-step derivation
    1. associate-*r/97.7%

      \[\leadsto \color{blue}{\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}} \]
  5. Simplified97.7%

    \[\leadsto \color{blue}{\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}} \]
  6. Step-by-step derivation
    1. associate-/l*99.2%

      \[\leadsto \color{blue}{\frac{1.3333333333333333}{\frac{\pi}{\sqrt{0.5}}}} \]
    2. associate-/r/99.2%

      \[\leadsto \color{blue}{\frac{1.3333333333333333}{\pi} \cdot \sqrt{0.5}} \]
  7. Applied egg-rr99.2%

    \[\leadsto \color{blue}{\frac{1.3333333333333333}{\pi} \cdot \sqrt{0.5}} \]
  8. Final simplification99.2%

    \[\leadsto \frac{1.3333333333333333}{\pi} \cdot \sqrt{0.5} \]

Alternative 6: 97.3% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \frac{\sqrt{0.8888888888888888}}{\pi} \end{array} \]
(FPCore (v) :precision binary64 (/ (sqrt 0.8888888888888888) PI))
double code(double v) {
	return sqrt(0.8888888888888888) / ((double) M_PI);
}
public static double code(double v) {
	return Math.sqrt(0.8888888888888888) / Math.PI;
}
def code(v):
	return math.sqrt(0.8888888888888888) / math.pi
function code(v)
	return Float64(sqrt(0.8888888888888888) / pi)
end
function tmp = code(v)
	tmp = sqrt(0.8888888888888888) / pi;
end
code[v_] := N[(N[Sqrt[0.8888888888888888], $MachinePrecision] / Pi), $MachinePrecision]
\begin{array}{l}

\\
\frac{\sqrt{0.8888888888888888}}{\pi}
\end{array}
Derivation
  1. Initial program 98.5%

    \[\frac{4}{\left(\left(3 \cdot \pi\right) \cdot \left(1 - v \cdot v\right)\right) \cdot \sqrt{2 - 6 \cdot \left(v \cdot v\right)}} \]
  2. Simplified100.0%

    \[\leadsto \color{blue}{\frac{-1.3333333333333333}{\mathsf{fma}\left(v, v, -1\right) \cdot \left(\pi \cdot \sqrt{\mathsf{fma}\left(v, v \cdot -6, 2\right)}\right)}} \]
  3. Taylor expanded in v around 0 97.7%

    \[\leadsto \color{blue}{1.3333333333333333 \cdot \frac{\sqrt{0.5}}{\pi}} \]
  4. Step-by-step derivation
    1. associate-*r/97.7%

      \[\leadsto \color{blue}{\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}} \]
  5. Simplified97.7%

    \[\leadsto \color{blue}{\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}} \]
  6. Step-by-step derivation
    1. expm1-log1p-u97.7%

      \[\leadsto \color{blue}{\mathsf{expm1}\left(\mathsf{log1p}\left(\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}\right)\right)} \]
    2. expm1-udef97.7%

      \[\leadsto \color{blue}{e^{\mathsf{log1p}\left(\frac{1.3333333333333333 \cdot \sqrt{0.5}}{\pi}\right)} - 1} \]
    3. add-sqr-sqrt97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\color{blue}{\sqrt{1.3333333333333333 \cdot \sqrt{0.5}} \cdot \sqrt{1.3333333333333333 \cdot \sqrt{0.5}}}}{\pi}\right)} - 1 \]
    4. sqrt-unprod97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\color{blue}{\sqrt{\left(1.3333333333333333 \cdot \sqrt{0.5}\right) \cdot \left(1.3333333333333333 \cdot \sqrt{0.5}\right)}}}{\pi}\right)} - 1 \]
    5. swap-sqr97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\sqrt{\color{blue}{\left(1.3333333333333333 \cdot 1.3333333333333333\right) \cdot \left(\sqrt{0.5} \cdot \sqrt{0.5}\right)}}}{\pi}\right)} - 1 \]
    6. metadata-eval97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\sqrt{\color{blue}{1.7777777777777777} \cdot \left(\sqrt{0.5} \cdot \sqrt{0.5}\right)}}{\pi}\right)} - 1 \]
    7. add-sqr-sqrt97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\sqrt{1.7777777777777777 \cdot \color{blue}{0.5}}}{\pi}\right)} - 1 \]
    8. metadata-eval97.7%

      \[\leadsto e^{\mathsf{log1p}\left(\frac{\sqrt{\color{blue}{0.8888888888888888}}}{\pi}\right)} - 1 \]
  7. Applied egg-rr97.7%

    \[\leadsto \color{blue}{e^{\mathsf{log1p}\left(\frac{\sqrt{0.8888888888888888}}{\pi}\right)} - 1} \]
  8. Step-by-step derivation
    1. expm1-def97.7%

      \[\leadsto \color{blue}{\mathsf{expm1}\left(\mathsf{log1p}\left(\frac{\sqrt{0.8888888888888888}}{\pi}\right)\right)} \]
    2. expm1-log1p97.7%

      \[\leadsto \color{blue}{\frac{\sqrt{0.8888888888888888}}{\pi}} \]
  9. Simplified97.7%

    \[\leadsto \color{blue}{\frac{\sqrt{0.8888888888888888}}{\pi}} \]
  10. Final simplification97.7%

    \[\leadsto \frac{\sqrt{0.8888888888888888}}{\pi} \]

Reproduce

?
herbie shell --seed 2023291 
(FPCore (v)
  :name "Falkner and Boettcher, Equation (22+)"
  :precision binary64
  (/ 4.0 (* (* (* 3.0 PI) (- 1.0 (* v v))) (sqrt (- 2.0 (* 6.0 (* v v)))))))