Ian Simplification

Percentage Accurate: 6.7% → 8.2%
Time: 2.4min
Alternatives: 2
Speedup: 1.1×

Specification

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

\\
\frac{\pi}{2} - 2 \cdot \sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\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 2 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: 6.7% accurate, 1.0× speedup?

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

\\
\frac{\pi}{2} - 2 \cdot \sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)
\end{array}

Alternative 1: 8.2% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\mathsf{fma}\left(-0.5, x, 0.5\right)}\right), -0.5 \cdot \pi\right) \end{array} \]
(FPCore (x)
 :precision binary64
 (fma 2.0 (acos (sqrt (fma -0.5 x 0.5))) (* -0.5 PI)))
double code(double x) {
	return fma(2.0, acos(sqrt(fma(-0.5, x, 0.5))), (-0.5 * ((double) M_PI)));
}
function code(x)
	return fma(2.0, acos(sqrt(fma(-0.5, x, 0.5))), Float64(-0.5 * pi))
end
code[x_] := N[(2.0 * N[ArcCos[N[Sqrt[N[(-0.5 * x + 0.5), $MachinePrecision]], $MachinePrecision]], $MachinePrecision] + N[(-0.5 * Pi), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\mathsf{fma}\left(-0.5, x, 0.5\right)}\right), -0.5 \cdot \pi\right)
\end{array}
Derivation
  1. Initial program 6.4%

    \[\frac{\pi}{2} - 2 \cdot \sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\right) \]
  2. Add Preprocessing
  3. Step-by-step derivation
    1. lift-asin.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)} \]
    2. asin-acosN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{2} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)} \]
    3. lift-PI.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right) \]
    4. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right) \]
    5. sub-negN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{2} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right)} \]
    6. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    7. div-invN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    8. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    9. lower-fma.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)} \]
    10. lower-neg.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \color{blue}{\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)}\right) \]
    11. lower-acos.f647.8

      \[\leadsto \frac{\pi}{2} - 2 \cdot \mathsf{fma}\left(\pi, 0.5, -\color{blue}{\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)}\right) \]
    12. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1 - x}{2}}}\right)\right)\right) \]
    13. lift--.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{\color{blue}{1 - x}}{2}}\right)\right)\right) \]
    14. div-subN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} - \frac{x}{2}}}\right)\right)\right) \]
    15. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2}} - \frac{x}{2}}\right)\right)\right) \]
    16. sub-negN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} + \left(\mathsf{neg}\left(\frac{x}{2}\right)\right)}}\right)\right)\right) \]
    17. +-commutativeN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\left(\mathsf{neg}\left(\frac{x}{2}\right)\right) + \frac{1}{2}}}\right)\right)\right) \]
    18. div-invN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\left(\mathsf{neg}\left(\color{blue}{x \cdot \frac{1}{2}}\right)\right) + \frac{1}{2}}\right)\right)\right) \]
    19. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\left(\mathsf{neg}\left(x \cdot \color{blue}{\frac{1}{2}}\right)\right) + \frac{1}{2}}\right)\right)\right) \]
    20. distribute-rgt-neg-inN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{x \cdot \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} + \frac{1}{2}}\right)\right)\right) \]
    21. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \color{blue}{\frac{-1}{2}} + \frac{1}{2}}\right)\right)\right) \]
    22. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \color{blue}{\frac{1}{-2}} + \frac{1}{2}}\right)\right)\right) \]
    23. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \frac{1}{\color{blue}{\mathsf{neg}\left(2\right)}} + \frac{1}{2}}\right)\right)\right) \]
    24. lower-fma.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\mathsf{fma}\left(x, \frac{1}{\mathsf{neg}\left(2\right)}, \frac{1}{2}\right)}}\right)\right)\right) \]
  4. Applied rewrites7.8%

    \[\leadsto \frac{\pi}{2} - 2 \cdot \color{blue}{\mathsf{fma}\left(\pi, 0.5, -\cos^{-1} \left(\sqrt{\mathsf{fma}\left(x, -0.5, 0.5\right)}\right)\right)} \]
  5. Taylor expanded in x around 0

    \[\leadsto \color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right) - 2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)} \]
  6. Step-by-step derivation
    1. cancel-sign-sub-invN/A

      \[\leadsto \color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(2\right)\right) \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)} \]
    2. metadata-evalN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + \color{blue}{-2} \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right) \]
    3. sub-negN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \color{blue}{\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)\right)\right)} \]
    4. metadata-evalN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} + \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} \cdot x}\right)\right)\right)\right) \]
    5. cancel-sign-sub-invN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} - \frac{1}{2} \cdot x}}\right)\right)\right)\right) \]
    6. distribute-lft-inN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + \color{blue}{\left(-2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right)\right)} \]
    7. associate-+r+N/A

      \[\leadsto \color{blue}{\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right)} \]
    8. distribute-rgt1-inN/A

      \[\leadsto \color{blue}{\left(-2 + 1\right) \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)} + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    9. metadata-evalN/A

      \[\leadsto \color{blue}{-1} \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    10. neg-mul-1N/A

      \[\leadsto \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right)} + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    11. +-commutativeN/A

      \[\leadsto \color{blue}{-2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) + \left(\mathsf{neg}\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right)} \]
  7. Applied rewrites7.8%

    \[\leadsto \color{blue}{\mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\mathsf{fma}\left(-0.5, x, 0.5\right)}\right), \pi \cdot -0.5\right)} \]
  8. Final simplification7.8%

    \[\leadsto \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\mathsf{fma}\left(-0.5, x, 0.5\right)}\right), -0.5 \cdot \pi\right) \]
  9. Add Preprocessing

Alternative 2: 5.4% accurate, 1.2× speedup?

\[\begin{array}{l} \\ \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{0.5}\right), -0.5 \cdot \pi\right) \end{array} \]
(FPCore (x) :precision binary64 (fma 2.0 (acos (sqrt 0.5)) (* -0.5 PI)))
double code(double x) {
	return fma(2.0, acos(sqrt(0.5)), (-0.5 * ((double) M_PI)));
}
function code(x)
	return fma(2.0, acos(sqrt(0.5)), Float64(-0.5 * pi))
end
code[x_] := N[(2.0 * N[ArcCos[N[Sqrt[0.5], $MachinePrecision]], $MachinePrecision] + N[(-0.5 * Pi), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}

\\
\mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{0.5}\right), -0.5 \cdot \pi\right)
\end{array}
Derivation
  1. Initial program 6.7%

    \[\frac{\pi}{2} - 2 \cdot \sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\right) \]
  2. Add Preprocessing
  3. Step-by-step derivation
    1. lift-asin.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\sin^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)} \]
    2. asin-acosN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{2} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)} \]
    3. lift-PI.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right) \]
    4. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}} - \cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right) \]
    5. sub-negN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{2} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right)} \]
    6. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    7. div-invN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    8. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \left(\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}} + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)\right) \]
    9. lower-fma.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \color{blue}{\mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)\right)} \]
    10. lower-neg.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \color{blue}{\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)\right)}\right) \]
    11. lower-acos.f648.2

      \[\leadsto \frac{\pi}{2} - 2 \cdot \mathsf{fma}\left(\pi, 0.5, -\color{blue}{\cos^{-1} \left(\sqrt{\frac{1 - x}{2}}\right)}\right) \]
    12. lift-/.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1 - x}{2}}}\right)\right)\right) \]
    13. lift--.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{\color{blue}{1 - x}}{2}}\right)\right)\right) \]
    14. div-subN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} - \frac{x}{2}}}\right)\right)\right) \]
    15. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2}} - \frac{x}{2}}\right)\right)\right) \]
    16. sub-negN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} + \left(\mathsf{neg}\left(\frac{x}{2}\right)\right)}}\right)\right)\right) \]
    17. +-commutativeN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\left(\mathsf{neg}\left(\frac{x}{2}\right)\right) + \frac{1}{2}}}\right)\right)\right) \]
    18. div-invN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\left(\mathsf{neg}\left(\color{blue}{x \cdot \frac{1}{2}}\right)\right) + \frac{1}{2}}\right)\right)\right) \]
    19. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\left(\mathsf{neg}\left(x \cdot \color{blue}{\frac{1}{2}}\right)\right) + \frac{1}{2}}\right)\right)\right) \]
    20. distribute-rgt-neg-inN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{x \cdot \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} + \frac{1}{2}}\right)\right)\right) \]
    21. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \color{blue}{\frac{-1}{2}} + \frac{1}{2}}\right)\right)\right) \]
    22. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \color{blue}{\frac{1}{-2}} + \frac{1}{2}}\right)\right)\right) \]
    23. metadata-evalN/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{x \cdot \frac{1}{\color{blue}{\mathsf{neg}\left(2\right)}} + \frac{1}{2}}\right)\right)\right) \]
    24. lower-fma.f64N/A

      \[\leadsto \frac{\mathsf{PI}\left(\right)}{2} - 2 \cdot \mathsf{fma}\left(\mathsf{PI}\left(\right), \frac{1}{2}, \mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\mathsf{fma}\left(x, \frac{1}{\mathsf{neg}\left(2\right)}, \frac{1}{2}\right)}}\right)\right)\right) \]
  4. Applied rewrites8.2%

    \[\leadsto \frac{\pi}{2} - 2 \cdot \color{blue}{\mathsf{fma}\left(\pi, 0.5, -\cos^{-1} \left(\sqrt{\mathsf{fma}\left(x, -0.5, 0.5\right)}\right)\right)} \]
  5. Taylor expanded in x around 0

    \[\leadsto \color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right) - 2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)} \]
  6. Step-by-step derivation
    1. cancel-sign-sub-invN/A

      \[\leadsto \color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(2\right)\right) \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)} \]
    2. metadata-evalN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + \color{blue}{-2} \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) - \cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right) \]
    3. sub-negN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \color{blue}{\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} + \frac{-1}{2} \cdot x}\right)\right)\right)\right)} \]
    4. metadata-evalN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} + \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} \cdot x}\right)\right)\right)\right) \]
    5. cancel-sign-sub-invN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\color{blue}{\frac{1}{2} - \frac{1}{2} \cdot x}}\right)\right)\right)\right) \]
    6. distribute-lft-inN/A

      \[\leadsto \frac{1}{2} \cdot \mathsf{PI}\left(\right) + \color{blue}{\left(-2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right)\right)} \]
    7. associate-+r+N/A

      \[\leadsto \color{blue}{\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right) + -2 \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right)} \]
    8. distribute-rgt1-inN/A

      \[\leadsto \color{blue}{\left(-2 + 1\right) \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)} + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    9. metadata-evalN/A

      \[\leadsto \color{blue}{-1} \cdot \left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right) + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    10. neg-mul-1N/A

      \[\leadsto \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right)} + -2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) \]
    11. +-commutativeN/A

      \[\leadsto \color{blue}{-2 \cdot \left(\mathsf{neg}\left(\cos^{-1} \left(\sqrt{\frac{1}{2} - \frac{1}{2} \cdot x}\right)\right)\right) + \left(\mathsf{neg}\left(\frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right)} \]
  7. Applied rewrites8.2%

    \[\leadsto \color{blue}{\mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\mathsf{fma}\left(-0.5, x, 0.5\right)}\right), \pi \cdot -0.5\right)} \]
  8. Taylor expanded in x around 0

    \[\leadsto \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{\frac{1}{2}}\right), \mathsf{PI}\left(\right) \cdot \frac{-1}{2}\right) \]
  9. Step-by-step derivation
    1. Applied rewrites5.4%

      \[\leadsto \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{0.5}\right), \pi \cdot -0.5\right) \]
    2. Final simplification5.4%

      \[\leadsto \mathsf{fma}\left(2, \cos^{-1} \left(\sqrt{0.5}\right), -0.5 \cdot \pi\right) \]
    3. Add Preprocessing

    Developer Target 1: 100.0% accurate, 1.4× speedup?

    \[\begin{array}{l} \\ \sin^{-1} x \end{array} \]
    (FPCore (x) :precision binary64 (asin x))
    double code(double x) {
    	return asin(x);
    }
    
    real(8) function code(x)
        real(8), intent (in) :: x
        code = asin(x)
    end function
    
    public static double code(double x) {
    	return Math.asin(x);
    }
    
    def code(x):
    	return math.asin(x)
    
    function code(x)
    	return asin(x)
    end
    
    function tmp = code(x)
    	tmp = asin(x);
    end
    
    code[x_] := N[ArcSin[x], $MachinePrecision]
    
    \begin{array}{l}
    
    \\
    \sin^{-1} x
    \end{array}
    

    Reproduce

    ?
    herbie shell --seed 2024226 
    (FPCore (x)
      :name "Ian Simplification"
      :precision binary64
    
      :alt
      (! :herbie-platform default (asin x))
    
      (- (/ PI 2.0) (* 2.0 (asin (sqrt (/ (- 1.0 x) 2.0))))))