UniformSampleCone, x

Percentage Accurate: 57.1% → 99.0%
Time: 8.7s
Alternatives: 20
Speedup: 7.1×

Specification

?
\[\left(\left(2.328306437 \cdot 10^{-10} \leq ux \land ux \leq 1\right) \land \left(2.328306437 \cdot 10^{-10} \leq uy \land uy \leq 1\right)\right) \land \left(0 \leq maxCos \land maxCos \leq 1\right)\]
\[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0} \end{array} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
   (* (cos (* (* uy 2.0) (PI))) (sqrt (- 1.0 (* t_0 t_0))))))
\begin{array}{l}

\\
\begin{array}{l}
t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0}
\end{array}
\end{array}

Sampling outcomes in binary32 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 20 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: 57.1% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0} \end{array} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
   (* (cos (* (* uy 2.0) (PI))) (sqrt (- 1.0 (* t_0 t_0))))))
\begin{array}{l}

\\
\begin{array}{l}
t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0}
\end{array}
\end{array}

Alternative 1: 99.0% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 - maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt
   (fma
    maxCos
    (* ux (- (* ux (- 2.0 maxCos)) 2.0))
    (fma ux 2.0 (* (- ux) ux))))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 - maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{maxCos \cdot \left(-1 \cdot \left(maxCos \cdot {ux}^{2}\right) + ux \cdot \left(2 \cdot ux - 2\right)\right) + \color{blue}{ux \cdot \left(2 + -1 \cdot ux\right)}} \]
  7. Step-by-step derivation
    1. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -1 \cdot \left(maxCos \cdot {ux}^{2}\right) + \color{blue}{ux \cdot \left(2 \cdot ux - 2\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    2. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \color{blue}{{ux}^{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    3. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot {ux}^{\color{blue}{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    4. pow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    5. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    6. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    7. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    10. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    11. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  8. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \color{blue}{\mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  9. Step-by-step derivation
    1. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    2. lift-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    3. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    4. distribute-lft-inN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot 2 + ux \cdot \left(-1 \cdot ux\right)\right)} \]
    5. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
    6. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
    7. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(\mathsf{neg}\left(ux\right)\right)\right)\right)} \]
    8. lift-neg.f3299.3

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
  10. Applied rewrites99.3%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
  11. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - \color{blue}{2}\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
  12. Step-by-step derivation
    1. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    2. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    3. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    4. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    5. lower-*.f3299.3

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
  13. Applied rewrites99.3%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - \color{blue}{2}\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
  14. Final simplification99.3%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 - maxCos\right) - 2\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)} \]
  15. Add Preprocessing

Alternative 2: 99.0% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2\right), ux \cdot \left(2 - ux\right)\right)} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt
   (fma
    maxCos
    (* ux (- (* ux (fma -1.0 maxCos 2.0)) 2.0))
    (* ux (- 2.0 ux))))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2\right), ux \cdot \left(2 - ux\right)\right)}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{maxCos \cdot \left(-1 \cdot \left(maxCos \cdot {ux}^{2}\right) + ux \cdot \left(2 \cdot ux - 2\right)\right) + \color{blue}{ux \cdot \left(2 + -1 \cdot ux\right)}} \]
  7. Step-by-step derivation
    1. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -1 \cdot \left(maxCos \cdot {ux}^{2}\right) + \color{blue}{ux \cdot \left(2 \cdot ux - 2\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    2. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \color{blue}{{ux}^{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    3. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot {ux}^{\color{blue}{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    4. pow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    5. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    6. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    7. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    10. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    11. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  8. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \color{blue}{\mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  9. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - \color{blue}{2}\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  10. Step-by-step derivation
    1. distribute-rgt-inN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(2 \cdot ux + \left(-1 \cdot maxCos\right) \cdot ux\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    2. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(2 \cdot ux + -1 \cdot \left(maxCos \cdot ux\right)\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    3. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    5. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    6. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(2 \cdot ux + -1 \cdot \left(maxCos \cdot ux\right)\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    7. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(\left(2 \cdot ux + \left(-1 \cdot maxCos\right) \cdot ux\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. distribute-rgt-inN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + -1 \cdot maxCos\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    10. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(2 + \left(\mathsf{neg}\left(maxCos\right)\right)\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    11. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(\left(\mathsf{neg}\left(maxCos\right)\right) + 2\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    12. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \left(-1 \cdot maxCos + 2\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    13. lower-fma.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  11. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - \color{blue}{2}\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  12. Final simplification99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, ux \cdot \left(ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2\right), ux \cdot \left(2 - ux\right)\right)} \]
  13. Add Preprocessing

Alternative 3: 99.0% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt
   (*
    (- (fma (- ux) (* (- maxCos 1.0) (- maxCos 1.0)) 2.0) (* maxCos 2.0))
    ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Step-by-step derivation
    1. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lift-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    3. unpow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    6. lift--.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  7. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  8. Add Preprocessing

Alternative 4: 99.0% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, -ux\right) + 2\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt
   (* (+ (fma maxCos (- (* ux (fma -1.0 maxCos 2.0)) 2.0) (- ux)) 2.0) ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, -ux\right) + 2\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Step-by-step derivation
    1. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lift-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    3. unpow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    6. lift--.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  7. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  8. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \left(-1 \cdot ux + maxCos \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right)\right)\right) \cdot ux} \]
  9. Step-by-step derivation
    1. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot ux + maxCos \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right)\right) + 2\right) \cdot ux} \]
    2. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot ux + maxCos \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right)\right) + 2\right) \cdot ux} \]
    3. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(maxCos \cdot \left(\left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2\right) + -1 \cdot ux\right) + 2\right) \cdot ux} \]
    4. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, \left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    5. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, \left(-1 \cdot \left(maxCos \cdot ux\right) + 2 \cdot ux\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    6. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, \left(2 \cdot ux + -1 \cdot \left(maxCos \cdot ux\right)\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    7. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, \left(2 \cdot ux + \left(-1 \cdot maxCos\right) \cdot ux\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    8. distribute-rgt-inN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \left(2 + -1 \cdot maxCos\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    9. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \left(2 + -1 \cdot maxCos\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    10. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \left(2 + \left(\mathsf{neg}\left(maxCos\right)\right)\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    11. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \left(\left(\mathsf{neg}\left(maxCos\right)\right) + 2\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    12. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \left(-1 \cdot maxCos + 2\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    13. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    14. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, \mathsf{neg}\left(ux\right)\right) + 2\right) \cdot ux} \]
    15. lift-neg.f3299.1

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, -ux\right) + 2\right) \cdot ux} \]
  10. Applied rewrites99.1%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, ux \cdot \mathsf{fma}\left(-1, maxCos, 2\right) - 2, -ux\right) + 2\right) \cdot ux} \]
  11. Add Preprocessing

Alternative 5: 98.4% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \mathsf{fma}\left(-2, maxCos, 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt (* (- (fma (- ux) (fma -2.0 maxCos 1.0) 2.0) (* maxCos 2.0)) ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \mathsf{fma}\left(-2, maxCos, 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Step-by-step derivation
    1. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lift-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    3. unpow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    6. lift--.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  7. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  8. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, 1 + -2 \cdot maxCos, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  9. Step-by-step derivation
    1. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, -2 \cdot maxCos + 1, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lower-fma.f3297.7

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \mathsf{fma}\left(-2, maxCos, 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  10. Applied rewrites97.7%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \mathsf{fma}\left(-2, maxCos, 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  11. Add Preprocessing

Alternative 6: 98.4% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt (* (+ 2.0 (fma -1.0 ux (* maxCos (- (* 2.0 ux) 2.0)))) ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \left(-1 \cdot ux + maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
  7. Step-by-step derivation
    1. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \left(-1 \cdot ux + maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
    2. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
    3. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
    4. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
    5. lower-*.f3297.7

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
  8. Applied rewrites97.7%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
  9. Final simplification97.7%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \mathsf{fma}\left(-1, ux, maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
  10. Add Preprocessing

Alternative 7: 98.4% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -ux\right) + 2\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt (* (+ (fma maxCos (- (* 2.0 ux) 2.0) (- ux)) 2.0) ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -ux\right) + 2\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Step-by-step derivation
    1. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lift-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    3. unpow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    6. lift--.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  7. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  8. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + \left(-1 \cdot ux + maxCos \cdot \left(2 \cdot ux - 2\right)\right)\right) \cdot ux} \]
  9. Step-by-step derivation
    1. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot ux + maxCos \cdot \left(2 \cdot ux - 2\right)\right) + 2\right) \cdot ux} \]
    2. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot ux + maxCos \cdot \left(2 \cdot ux - 2\right)\right) + 2\right) \cdot ux} \]
    3. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(maxCos \cdot \left(2 \cdot ux - 2\right) + -1 \cdot ux\right) + 2\right) \cdot ux} \]
    4. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    5. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    6. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -1 \cdot ux\right) + 2\right) \cdot ux} \]
    7. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, \mathsf{neg}\left(ux\right)\right) + 2\right) \cdot ux} \]
    8. lift-neg.f3297.7

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -ux\right) + 2\right) \cdot ux} \]
  10. Applied rewrites97.7%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos, 2 \cdot ux - 2, -ux\right) + 2\right) \cdot ux} \]
  11. Add Preprocessing

Alternative 8: 97.6% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 - ux\right)\right)} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt (fma maxCos (* -2.0 ux) (* ux (- 2.0 ux))))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 - ux\right)\right)}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{maxCos \cdot \left(-1 \cdot \left(maxCos \cdot {ux}^{2}\right) + ux \cdot \left(2 \cdot ux - 2\right)\right) + \color{blue}{ux \cdot \left(2 + -1 \cdot ux\right)}} \]
  7. Step-by-step derivation
    1. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -1 \cdot \left(maxCos \cdot {ux}^{2}\right) + \color{blue}{ux \cdot \left(2 \cdot ux - 2\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    2. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \color{blue}{{ux}^{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    3. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot {ux}^{\color{blue}{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    4. pow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    5. lift-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    6. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    7. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    10. lower-+.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    11. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  8. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \color{blue}{\mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  9. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  10. Step-by-step derivation
    1. lower-*.f3296.6

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  11. Applied rewrites96.6%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
  12. Final simplification96.6%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -2 \cdot ux, ux \cdot \left(2 - ux\right)\right)} \]
  13. Add Preprocessing

Alternative 9: 97.6% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-1, ux, 2\right) - maxCos \cdot 2\right) \cdot ux} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (cos (* (* uy 2.0) (PI)))
  (sqrt (* (- (fma -1.0 ux 2.0) (* maxCos 2.0)) ux))))
\begin{array}{l}

\\
\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-1, ux, 2\right) - maxCos \cdot 2\right) \cdot ux}
\end{array}
Derivation
  1. Initial program 57.4%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  2. Add Preprocessing
  3. Taylor expanded in ux around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
  4. Step-by-step derivation
    1. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    2. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
    3. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
    4. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    5. associate-*r*N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    6. mul-1-negN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    7. lower-fma.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    8. lower-neg.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    9. lower-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    10. lower--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
    11. *-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    12. lower-*.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  5. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
  6. Step-by-step derivation
    1. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lift-pow.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    3. unpow2N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    4. lower-*.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. lift--.f32N/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    6. lift--.f3299.2

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  7. Applied rewrites99.2%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  8. Taylor expanded in maxCos around 0

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot ux\right) - maxCos \cdot 2\right) \cdot ux} \]
  9. Step-by-step derivation
    1. +-commutativeN/A

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot ux + 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    2. lower-fma.f3296.6

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-1, ux, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  10. Applied rewrites96.6%

    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-1, ux, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
  11. Add Preprocessing

Alternative 10: 96.5% accurate, 1.1× speedup?

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;uy \leq 0.00016500000492669642:\\ \;\;\;\;1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)}\\ \mathbf{else}:\\ \;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-1, ux, 2\right) \cdot ux}\\ \end{array} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (if (<= uy 0.00016500000492669642)
   (*
    1.0
    (sqrt
     (fma
      maxCos
      (fma -1.0 (* maxCos (* ux ux)) (* ux (- (* 2.0 ux) 2.0)))
      (fma ux 2.0 (* (- ux) ux)))))
   (* (cos (* (* uy 2.0) (PI))) (sqrt (* (fma -1.0 ux 2.0) ux)))))
\begin{array}{l}

\\
\begin{array}{l}
\mathbf{if}\;uy \leq 0.00016500000492669642:\\
\;\;\;\;1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)}\\

\mathbf{else}:\\
\;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-1, ux, 2\right) \cdot ux}\\


\end{array}
\end{array}
Derivation
  1. Split input into 2 regimes
  2. if uy < 1.65000005e-4

    1. Initial program 56.6%

      \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
    2. Add Preprocessing
    3. Taylor expanded in ux around 0

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
    4. Step-by-step derivation
      1. *-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
      2. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
      3. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
      4. +-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      5. associate-*r*N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      6. mul-1-negN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      7. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      8. lower-neg.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      9. lower-pow.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      10. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      11. *-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
      12. lower-*.f3299.5

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. Applied rewrites99.5%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
    6. Taylor expanded in maxCos around 0

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{maxCos \cdot \left(-1 \cdot \left(maxCos \cdot {ux}^{2}\right) + ux \cdot \left(2 \cdot ux - 2\right)\right) + \color{blue}{ux \cdot \left(2 + -1 \cdot ux\right)}} \]
    7. Step-by-step derivation
      1. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -1 \cdot \left(maxCos \cdot {ux}^{2}\right) + \color{blue}{ux \cdot \left(2 \cdot ux - 2\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      2. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \color{blue}{{ux}^{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      3. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot {ux}^{\color{blue}{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      4. pow2N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      5. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      6. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      7. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      8. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      9. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      10. lower-+.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      11. lower-*.f3299.6

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. Applied rewrites99.6%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \color{blue}{\mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. Step-by-step derivation
      1. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      2. lift-+.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      3. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      4. distribute-lft-inN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot 2 + ux \cdot \left(-1 \cdot ux\right)\right)} \]
      5. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
      6. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
      7. mul-1-negN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(\mathsf{neg}\left(ux\right)\right)\right)\right)} \]
      8. lift-neg.f3299.6

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    10. Applied rewrites99.6%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    11. Taylor expanded in uy around 0

      \[\leadsto \color{blue}{1} \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    12. Step-by-step derivation
      1. Applied rewrites99.2%

        \[\leadsto \color{blue}{1} \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]

      if 1.65000005e-4 < uy

      1. Initial program 58.5%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Taylor expanded in ux around 0

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
      4. Step-by-step derivation
        1. *-commutativeN/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
        2. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
        3. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
        4. +-commutativeN/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        5. associate-*r*N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        6. mul-1-negN/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        7. lower-fma.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        8. lower-neg.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        9. lower-pow.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        10. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
        11. *-commutativeN/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        12. lower-*.f3298.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
      5. Applied rewrites98.7%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
      6. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        2. lift-pow.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        3. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        4. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        5. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
        6. lift--.f3298.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
      7. Applied rewrites98.7%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, \left(maxCos - 1\right) \cdot \left(maxCos - 1\right), 2\right) - maxCos \cdot 2\right) \cdot ux} \]
      8. Taylor expanded in maxCos around 0

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 + -1 \cdot ux\right) \cdot ux} \]
      9. Step-by-step derivation
        1. +-commutativeN/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(-1 \cdot ux + 2\right) \cdot ux} \]
        2. lower-fma.f3291.4

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-1, ux, 2\right) \cdot ux} \]
      10. Applied rewrites91.4%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-1, ux, 2\right) \cdot ux} \]
    13. Recombined 2 regimes into one program.
    14. Final simplification96.2%

      \[\leadsto \begin{array}{l} \mathbf{if}\;uy \leq 0.00016500000492669642:\\ \;\;\;\;1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)}\\ \mathbf{else}:\\ \;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-1, ux, 2\right) \cdot ux}\\ \end{array} \]
    15. Add Preprocessing

    Alternative 11: 75.2% accurate, 1.8× speedup?

    \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\ \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \mathsf{fma}\left(ux, maxCos - 1, 1\right)}\\ \end{array} \end{array} \]
    (FPCore (ux uy maxCos)
     :precision binary32
     (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
       (if (<= (sqrt (- 1.0 (* t_0 t_0))) 0.016599999740719795)
         (sqrt (* ux (fma -2.0 maxCos 2.0)))
         (sqrt
          (- 1.0 (* (- (+ 1.0 (* maxCos ux)) ux) (fma ux (- maxCos 1.0) 1.0)))))))
    float code(float ux, float uy, float maxCos) {
    	float t_0 = (1.0f - ux) + (ux * maxCos);
    	float tmp;
    	if (sqrtf((1.0f - (t_0 * t_0))) <= 0.016599999740719795f) {
    		tmp = sqrtf((ux * fmaf(-2.0f, maxCos, 2.0f)));
    	} else {
    		tmp = sqrtf((1.0f - (((1.0f + (maxCos * ux)) - ux) * fmaf(ux, (maxCos - 1.0f), 1.0f))));
    	}
    	return tmp;
    }
    
    function code(ux, uy, maxCos)
    	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
    	tmp = Float32(0.0)
    	if (sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) <= Float32(0.016599999740719795))
    		tmp = sqrt(Float32(ux * fma(Float32(-2.0), maxCos, Float32(2.0))));
    	else
    		tmp = sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(1.0) + Float32(maxCos * ux)) - ux) * fma(ux, Float32(maxCos - Float32(1.0)), Float32(1.0)))));
    	end
    	return tmp
    end
    
    \begin{array}{l}
    
    \\
    \begin{array}{l}
    t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
    \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\
    \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\
    
    \mathbf{else}:\\
    \;\;\;\;\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \mathsf{fma}\left(ux, maxCos - 1, 1\right)}\\
    
    
    \end{array}
    \end{array}
    
    Derivation
    1. Split input into 2 regimes
    2. if (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))))) < 0.0165999997

      1. Initial program 36.8%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3236.0

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites36.0%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites32.3%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around 0

        \[\leadsto \sqrt{ux \cdot \left(2 - 2 \cdot maxCos\right)} \]
      9. Step-by-step derivation
        1. metadata-evalN/A

          \[\leadsto \sqrt{ux \cdot \left(2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos\right)} \]
        2. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        4. +-commutativeN/A

          \[\leadsto \sqrt{ux \cdot \left(-2 \cdot maxCos + 2\right)} \]
        5. lift-fma.f3275.8

          \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]
      10. Applied rewrites75.8%

        \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]

      if 0.0165999997 < (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))))

      1. Initial program 88.9%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3288.5

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites88.5%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites73.2%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around 0

        \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 + ux \cdot \left(maxCos - 1\right)\right)} \]
      9. Step-by-step derivation
        1. +-commutativeN/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(ux \cdot \left(maxCos - 1\right) + 1\right)} \]
        2. lower-fma.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \mathsf{fma}\left(ux, maxCos - 1, 1\right)} \]
        3. lift--.f3273.4

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \mathsf{fma}\left(ux, maxCos - 1, 1\right)} \]
      10. Applied rewrites73.4%

        \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \mathsf{fma}\left(ux, maxCos - 1, 1\right)} \]
    3. Recombined 2 regimes into one program.
    4. Add Preprocessing

    Alternative 12: 75.1% accurate, 1.9× speedup?

    \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\ \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \mathsf{fma}\left(maxCos, ux, 1 - ux\right)}\\ \end{array} \end{array} \]
    (FPCore (ux uy maxCos)
     :precision binary32
     (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
       (if (<= (sqrt (- 1.0 (* t_0 t_0))) 0.016599999740719795)
         (sqrt (* ux (fma -2.0 maxCos 2.0)))
         (sqrt
          (- 1.0 (* (- (fma maxCos ux 1.0) ux) (fma maxCos ux (- 1.0 ux))))))))
    float code(float ux, float uy, float maxCos) {
    	float t_0 = (1.0f - ux) + (ux * maxCos);
    	float tmp;
    	if (sqrtf((1.0f - (t_0 * t_0))) <= 0.016599999740719795f) {
    		tmp = sqrtf((ux * fmaf(-2.0f, maxCos, 2.0f)));
    	} else {
    		tmp = sqrtf((1.0f - ((fmaf(maxCos, ux, 1.0f) - ux) * fmaf(maxCos, ux, (1.0f - ux)))));
    	}
    	return tmp;
    }
    
    function code(ux, uy, maxCos)
    	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
    	tmp = Float32(0.0)
    	if (sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) <= Float32(0.016599999740719795))
    		tmp = sqrt(Float32(ux * fma(Float32(-2.0), maxCos, Float32(2.0))));
    	else
    		tmp = sqrt(Float32(Float32(1.0) - Float32(Float32(fma(maxCos, ux, Float32(1.0)) - ux) * fma(maxCos, ux, Float32(Float32(1.0) - ux)))));
    	end
    	return tmp
    end
    
    \begin{array}{l}
    
    \\
    \begin{array}{l}
    t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
    \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\
    \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\
    
    \mathbf{else}:\\
    \;\;\;\;\sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \mathsf{fma}\left(maxCos, ux, 1 - ux\right)}\\
    
    
    \end{array}
    \end{array}
    
    Derivation
    1. Split input into 2 regimes
    2. if (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))))) < 0.0165999997

      1. Initial program 36.8%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3236.0

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites36.0%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites32.3%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around 0

        \[\leadsto \sqrt{ux \cdot \left(2 - 2 \cdot maxCos\right)} \]
      9. Step-by-step derivation
        1. metadata-evalN/A

          \[\leadsto \sqrt{ux \cdot \left(2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos\right)} \]
        2. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        4. +-commutativeN/A

          \[\leadsto \sqrt{ux \cdot \left(-2 \cdot maxCos + 2\right)} \]
        5. lift-fma.f3275.8

          \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]
      10. Applied rewrites75.8%

        \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]

      if 0.0165999997 < (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))))

      1. Initial program 88.9%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3288.5

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites88.5%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites73.2%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Step-by-step derivation
        1. lift-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        2. lift-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        3. +-commutativeN/A

          \[\leadsto \sqrt{1 - \left(\left(maxCos \cdot ux + 1\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        4. lower-fma.f3273.2

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        5. lift--.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        6. lift-fma.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        7. lift-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        8. lift-/.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        9. lift-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        10. lift-/.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        11. lift-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)} \]
        12. pow2N/A

          \[\leadsto \sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      9. Applied rewrites73.4%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \mathsf{fma}\left(maxCos, ux, 1 - ux\right)}} \]
    3. Recombined 2 regimes into one program.
    4. Add Preprocessing

    Alternative 13: 74.2% accurate, 2.0× speedup?

    \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\ \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 - ux\right)}\\ \end{array} \end{array} \]
    (FPCore (ux uy maxCos)
     :precision binary32
     (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
       (if (<= (sqrt (- 1.0 (* t_0 t_0))) 0.016599999740719795)
         (sqrt (* ux (fma -2.0 maxCos 2.0)))
         (sqrt (- 1.0 (* (- (+ 1.0 (* maxCos ux)) ux) (- 1.0 ux)))))))
    float code(float ux, float uy, float maxCos) {
    	float t_0 = (1.0f - ux) + (ux * maxCos);
    	float tmp;
    	if (sqrtf((1.0f - (t_0 * t_0))) <= 0.016599999740719795f) {
    		tmp = sqrtf((ux * fmaf(-2.0f, maxCos, 2.0f)));
    	} else {
    		tmp = sqrtf((1.0f - (((1.0f + (maxCos * ux)) - ux) * (1.0f - ux))));
    	}
    	return tmp;
    }
    
    function code(ux, uy, maxCos)
    	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
    	tmp = Float32(0.0)
    	if (sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) <= Float32(0.016599999740719795))
    		tmp = sqrt(Float32(ux * fma(Float32(-2.0), maxCos, Float32(2.0))));
    	else
    		tmp = sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(1.0) + Float32(maxCos * ux)) - ux) * Float32(Float32(1.0) - ux))));
    	end
    	return tmp
    end
    
    \begin{array}{l}
    
    \\
    \begin{array}{l}
    t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
    \mathbf{if}\;\sqrt{1 - t\_0 \cdot t\_0} \leq 0.016599999740719795:\\
    \;\;\;\;\sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}\\
    
    \mathbf{else}:\\
    \;\;\;\;\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 - ux\right)}\\
    
    
    \end{array}
    \end{array}
    
    Derivation
    1. Split input into 2 regimes
    2. if (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))))) < 0.0165999997

      1. Initial program 36.8%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3236.0

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites36.0%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites32.3%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around 0

        \[\leadsto \sqrt{ux \cdot \left(2 - 2 \cdot maxCos\right)} \]
      9. Step-by-step derivation
        1. metadata-evalN/A

          \[\leadsto \sqrt{ux \cdot \left(2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos\right)} \]
        2. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
        4. +-commutativeN/A

          \[\leadsto \sqrt{ux \cdot \left(-2 \cdot maxCos + 2\right)} \]
        5. lift-fma.f3275.8

          \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]
      10. Applied rewrites75.8%

        \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]

      if 0.0165999997 < (sqrt.f32 (-.f32 #s(literal 1 binary32) (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))))

      1. Initial program 88.9%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3288.5

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites88.5%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites73.2%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in maxCos around 0

        \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\frac{1}{1 + ux} - \frac{{ux}^{2}}{1 + ux}\right)} \]
      9. Step-by-step derivation
        1. sub-divN/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \frac{1 - {ux}^{2}}{1 + ux}} \]
        2. metadata-evalN/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \frac{1 \cdot 1 - {ux}^{2}}{1 + ux}} \]
        3. pow2N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} \]
        4. flip--N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 - ux\right)} \]
        5. lower--.f3269.0

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 - ux\right)} \]
      10. Applied rewrites69.0%

        \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(1 - ux\right)} \]
    3. Recombined 2 regimes into one program.
    4. Add Preprocessing

    Alternative 14: 80.0% accurate, 2.4× speedup?

    \[\begin{array}{l} \\ 1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)} \end{array} \]
    (FPCore (ux uy maxCos)
     :precision binary32
     (*
      1.0
      (sqrt
       (fma
        maxCos
        (fma -1.0 (* maxCos (* ux ux)) (* ux (- (* 2.0 ux) 2.0)))
        (fma ux 2.0 (* (- ux) ux))))))
    float code(float ux, float uy, float maxCos) {
    	return 1.0f * sqrtf(fmaf(maxCos, fmaf(-1.0f, (maxCos * (ux * ux)), (ux * ((2.0f * ux) - 2.0f))), fmaf(ux, 2.0f, (-ux * ux))));
    }
    
    function code(ux, uy, maxCos)
    	return Float32(Float32(1.0) * sqrt(fma(maxCos, fma(Float32(-1.0), Float32(maxCos * Float32(ux * ux)), Float32(ux * Float32(Float32(Float32(2.0) * ux) - Float32(2.0)))), fma(ux, Float32(2.0), Float32(Float32(-ux) * ux)))))
    end
    
    \begin{array}{l}
    
    \\
    1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)}
    \end{array}
    
    Derivation
    1. Initial program 57.4%

      \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
    2. Add Preprocessing
    3. Taylor expanded in ux around 0

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{ux \cdot \left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right)}} \]
    4. Step-by-step derivation
      1. *-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
      2. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot \color{blue}{ux}} \]
      3. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 + -1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right)\right) - 2 \cdot maxCos\right) \cdot ux} \]
      4. +-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(-1 \cdot \left(ux \cdot {\left(maxCos - 1\right)}^{2}\right) + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      5. associate-*r*N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(-1 \cdot ux\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      6. mul-1-negN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\left(\mathsf{neg}\left(ux\right)\right) \cdot {\left(maxCos - 1\right)}^{2} + 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      7. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\mathsf{neg}\left(ux\right), {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      8. lower-neg.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      9. lower-pow.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      10. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - 2 \cdot maxCos\right) \cdot ux} \]
      11. *-commutativeN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
      12. lower-*.f3299.2

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux} \]
    5. Applied rewrites99.2%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\mathsf{fma}\left(-ux, {\left(maxCos - 1\right)}^{2}, 2\right) - maxCos \cdot 2\right) \cdot ux}} \]
    6. Taylor expanded in maxCos around 0

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{maxCos \cdot \left(-1 \cdot \left(maxCos \cdot {ux}^{2}\right) + ux \cdot \left(2 \cdot ux - 2\right)\right) + \color{blue}{ux \cdot \left(2 + -1 \cdot ux\right)}} \]
    7. Step-by-step derivation
      1. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, -1 \cdot \left(maxCos \cdot {ux}^{2}\right) + \color{blue}{ux \cdot \left(2 \cdot ux - 2\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      2. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \color{blue}{{ux}^{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      3. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot {ux}^{\color{blue}{2}}, ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      4. pow2N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      5. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      6. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      7. lower--.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      8. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      9. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      10. lower-+.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      11. lower-*.f3299.2

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    8. Applied rewrites99.2%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \color{blue}{\mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right)}, ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
    9. Step-by-step derivation
      1. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      2. lift-+.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      3. lift-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot \left(2 + -1 \cdot ux\right)\right)} \]
      4. distribute-lft-inN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), ux \cdot 2 + ux \cdot \left(-1 \cdot ux\right)\right)} \]
      5. lower-fma.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
      6. lower-*.f32N/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-1 \cdot ux\right)\right)\right)} \]
      7. mul-1-negN/A

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(\mathsf{neg}\left(ux\right)\right)\right)\right)} \]
      8. lift-neg.f3299.3

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    10. Applied rewrites99.3%

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    11. Taylor expanded in uy around 0

      \[\leadsto \color{blue}{1} \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
    12. Step-by-step derivation
      1. Applied rewrites80.0%

        \[\leadsto \color{blue}{1} \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, ux \cdot \left(-ux\right)\right)\right)} \]
      2. Final simplification80.0%

        \[\leadsto 1 \cdot \sqrt{\mathsf{fma}\left(maxCos, \mathsf{fma}\left(-1, maxCos \cdot \left(ux \cdot ux\right), ux \cdot \left(2 \cdot ux - 2\right)\right), \mathsf{fma}\left(ux, 2, \left(-ux\right) \cdot ux\right)\right)} \]
      3. Add Preprocessing

      Alternative 15: 79.8% accurate, 3.0× speedup?

      \[\begin{array}{l} \\ \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \end{array} \]
      (FPCore (ux uy maxCos)
       :precision binary32
       (sqrt
        (*
         (* ux ux)
         (- (/ (fma -2.0 maxCos 2.0) ux) (* (- maxCos 1.0) (- maxCos 1.0))))))
      float code(float ux, float uy, float maxCos) {
      	return sqrtf(((ux * ux) * ((fmaf(-2.0f, maxCos, 2.0f) / ux) - ((maxCos - 1.0f) * (maxCos - 1.0f)))));
      }
      
      function code(ux, uy, maxCos)
      	return sqrt(Float32(Float32(ux * ux) * Float32(Float32(fma(Float32(-2.0), maxCos, Float32(2.0)) / ux) - Float32(Float32(maxCos - Float32(1.0)) * Float32(maxCos - Float32(1.0))))))
      end
      
      \begin{array}{l}
      
      \\
      \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)}
      \end{array}
      
      Derivation
      1. Initial program 57.4%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3256.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites56.7%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites48.4%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around inf

        \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
      9. Step-by-step derivation
        1. lower-*.f32N/A

          \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        2. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        3. lift-*.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        4. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)\right)} \]
        5. associate--r+N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(2 \cdot \frac{1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        6. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2 \cdot 1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        7. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        8. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - \frac{2 \cdot maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        9. div-subN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - 2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        10. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        11. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 + -2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        12. +-commutativeN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        13. lower--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
      10. Applied rewrites79.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
      11. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
        2. lift-pow.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
        3. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        4. lower-*.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        5. lift--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        6. lift--.f3279.8

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
      12. Applied rewrites79.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
      13. Add Preprocessing

      Alternative 16: 79.5% accurate, 3.2× speedup?

      \[\begin{array}{l} \\ \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \end{array} \]
      (FPCore (ux uy maxCos)
       :precision binary32
       (sqrt (* (* ux ux) (- (/ (fma -2.0 maxCos 2.0) ux) (- 1.0 (* 2.0 maxCos))))))
      float code(float ux, float uy, float maxCos) {
      	return sqrtf(((ux * ux) * ((fmaf(-2.0f, maxCos, 2.0f) / ux) - (1.0f - (2.0f * maxCos)))));
      }
      
      function code(ux, uy, maxCos)
      	return sqrt(Float32(Float32(ux * ux) * Float32(Float32(fma(Float32(-2.0), maxCos, Float32(2.0)) / ux) - Float32(Float32(1.0) - Float32(Float32(2.0) * maxCos)))))
      end
      
      \begin{array}{l}
      
      \\
      \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)}
      \end{array}
      
      Derivation
      1. Initial program 57.4%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3256.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites56.7%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites48.4%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around inf

        \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
      9. Step-by-step derivation
        1. lower-*.f32N/A

          \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        2. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        3. lift-*.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        4. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)\right)} \]
        5. associate--r+N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(2 \cdot \frac{1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        6. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2 \cdot 1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        7. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        8. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - \frac{2 \cdot maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        9. div-subN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - 2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        10. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        11. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 + -2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        12. +-commutativeN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        13. lower--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
      10. Applied rewrites79.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
      11. Taylor expanded in maxCos around 0

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 + -2 \cdot maxCos\right)\right)} \]
      12. Step-by-step derivation
        1. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos\right)\right)} \]
        2. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \]
        3. lower--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \]
        4. lower-*.f3278.8

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \]
      13. Applied rewrites78.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \]
      14. Final simplification78.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - \left(1 - 2 \cdot maxCos\right)\right)} \]
      15. Add Preprocessing

      Alternative 17: 79.0% accurate, 3.8× speedup?

      \[\begin{array}{l} \\ \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - 1\right)} \end{array} \]
      (FPCore (ux uy maxCos)
       :precision binary32
       (sqrt (* (* ux ux) (- (/ (fma -2.0 maxCos 2.0) ux) 1.0))))
      float code(float ux, float uy, float maxCos) {
      	return sqrtf(((ux * ux) * ((fmaf(-2.0f, maxCos, 2.0f) / ux) - 1.0f)));
      }
      
      function code(ux, uy, maxCos)
      	return sqrt(Float32(Float32(ux * ux) * Float32(Float32(fma(Float32(-2.0), maxCos, Float32(2.0)) / ux) - Float32(1.0))))
      end
      
      \begin{array}{l}
      
      \\
      \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - 1\right)}
      \end{array}
      
      Derivation
      1. Initial program 57.4%

        \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      2. Add Preprocessing
      3. Step-by-step derivation
        1. lift--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. flip--N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        3. lower-/.f32N/A

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        6. lower--.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        7. unpow2N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        8. lower-*.f32N/A

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        9. lower-+.f3256.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      4. Applied rewrites56.7%

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
      5. Taylor expanded in uy around 0

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
      6. Step-by-step derivation
        1. lower-sqrt.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        2. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        3. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        4. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        5. lower-+.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        6. lower-*.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. lower--.f32N/A

          \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
      7. Applied rewrites48.4%

        \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
      8. Taylor expanded in ux around inf

        \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
      9. Step-by-step derivation
        1. lower-*.f32N/A

          \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        2. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        3. lift-*.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        4. pow2N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)\right)} \]
        5. associate--r+N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(2 \cdot \frac{1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        6. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2 \cdot 1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        7. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        8. associate-*r/N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - \frac{2 \cdot maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        9. div-subN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - 2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        10. metadata-evalN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        11. fp-cancel-sign-sub-invN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 + -2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        12. +-commutativeN/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        13. lower--.f32N/A

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
      10. Applied rewrites79.8%

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
      11. Taylor expanded in maxCos around 0

        \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - 1\right)} \]
      12. Step-by-step derivation
        1. Applied rewrites78.3%

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - 1\right)} \]
        2. Final simplification78.3%

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - 1\right)} \]
        3. Add Preprocessing

        Alternative 18: 75.6% accurate, 3.9× speedup?

        \[\begin{array}{l} \\ \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \end{array} \]
        (FPCore (ux uy maxCos)
         :precision binary32
         (sqrt (* (* ux ux) (- (* 2.0 (/ 1.0 ux)) 1.0))))
        float code(float ux, float uy, float maxCos) {
        	return sqrtf(((ux * ux) * ((2.0f * (1.0f / ux)) - 1.0f)));
        }
        
        module fmin_fmax_functions
            implicit none
            private
            public fmax
            public fmin
        
            interface fmax
                module procedure fmax88
                module procedure fmax44
                module procedure fmax84
                module procedure fmax48
            end interface
            interface fmin
                module procedure fmin88
                module procedure fmin44
                module procedure fmin84
                module procedure fmin48
            end interface
        contains
            real(8) function fmax88(x, y) result (res)
                real(8), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(x, max(x, y), y /= y), x /= x)
            end function
            real(4) function fmax44(x, y) result (res)
                real(4), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(y, merge(x, max(x, y), y /= y), x /= x)
            end function
            real(8) function fmax84(x, y) result(res)
                real(8), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
            end function
            real(8) function fmax48(x, y) result(res)
                real(4), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
            end function
            real(8) function fmin88(x, y) result (res)
                real(8), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(x, min(x, y), y /= y), x /= x)
            end function
            real(4) function fmin44(x, y) result (res)
                real(4), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(y, merge(x, min(x, y), y /= y), x /= x)
            end function
            real(8) function fmin84(x, y) result(res)
                real(8), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
            end function
            real(8) function fmin48(x, y) result(res)
                real(4), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
            end function
        end module
        
        real(4) function code(ux, uy, maxcos)
        use fmin_fmax_functions
            real(4), intent (in) :: ux
            real(4), intent (in) :: uy
            real(4), intent (in) :: maxcos
            code = sqrt(((ux * ux) * ((2.0e0 * (1.0e0 / ux)) - 1.0e0)))
        end function
        
        function code(ux, uy, maxCos)
        	return sqrt(Float32(Float32(ux * ux) * Float32(Float32(Float32(2.0) * Float32(Float32(1.0) / ux)) - Float32(1.0))))
        end
        
        function tmp = code(ux, uy, maxCos)
        	tmp = sqrt(((ux * ux) * ((single(2.0) * (single(1.0) / ux)) - single(1.0))));
        end
        
        \begin{array}{l}
        
        \\
        \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)}
        \end{array}
        
        Derivation
        1. Initial program 57.4%

          \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. Add Preprocessing
        3. Step-by-step derivation
          1. lift--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          2. flip--N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          3. lower-/.f32N/A

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          5. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          6. lower--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          7. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          8. lower-*.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          9. lower-+.f3256.7

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        4. Applied rewrites56.7%

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. Taylor expanded in uy around 0

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
        6. Step-by-step derivation
          1. lower-sqrt.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          2. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          3. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          4. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          5. lower-+.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          6. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          7. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. Applied rewrites48.4%

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
        8. Taylor expanded in ux around inf

          \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
        9. Step-by-step derivation
          1. lower-*.f32N/A

            \[\leadsto \sqrt{{ux}^{2} \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
          2. pow2N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
          3. lift-*.f32N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + {\left(maxCos - 1\right)}^{2}\right)\right)} \]
          4. pow2N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - \left(2 \cdot \frac{maxCos}{ux} + \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)\right)} \]
          5. associate--r+N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(2 \cdot \frac{1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          6. associate-*r/N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2 \cdot 1}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          7. metadata-evalN/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - 2 \cdot \frac{maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          8. associate-*r/N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\left(\frac{2}{ux} - \frac{2 \cdot maxCos}{ux}\right) - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          9. div-subN/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - 2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          10. metadata-evalN/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          11. fp-cancel-sign-sub-invN/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{2 + -2 \cdot maxCos}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          12. +-commutativeN/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
          13. lower--.f32N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{-2 \cdot maxCos + 2}{ux} - \left(maxCos - 1\right) \cdot \left(maxCos - 1\right)\right)} \]
        10. Applied rewrites79.8%

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(\frac{\mathsf{fma}\left(-2, maxCos, 2\right)}{ux} - {\left(maxCos - 1\right)}^{2}\right)} \]
        11. Taylor expanded in maxCos around 0

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \]
        12. Step-by-step derivation
          1. lower--.f32N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \]
          2. lower-*.f32N/A

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \]
          3. lower-/.f3274.3

            \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \]
        13. Applied rewrites74.3%

          \[\leadsto \sqrt{\left(ux \cdot ux\right) \cdot \left(2 \cdot \frac{1}{ux} - 1\right)} \]
        14. Add Preprocessing

        Alternative 19: 65.1% accurate, 7.1× speedup?

        \[\begin{array}{l} \\ \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \end{array} \]
        (FPCore (ux uy maxCos) :precision binary32 (sqrt (* ux (fma -2.0 maxCos 2.0))))
        float code(float ux, float uy, float maxCos) {
        	return sqrtf((ux * fmaf(-2.0f, maxCos, 2.0f)));
        }
        
        function code(ux, uy, maxCos)
        	return sqrt(Float32(ux * fma(Float32(-2.0), maxCos, Float32(2.0))))
        end
        
        \begin{array}{l}
        
        \\
        \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)}
        \end{array}
        
        Derivation
        1. Initial program 57.4%

          \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. Add Preprocessing
        3. Step-by-step derivation
          1. lift--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          2. flip--N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          3. lower-/.f32N/A

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          5. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          6. lower--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          7. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          8. lower-*.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          9. lower-+.f3256.7

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        4. Applied rewrites56.7%

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. Taylor expanded in uy around 0

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
        6. Step-by-step derivation
          1. lower-sqrt.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          2. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          3. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          4. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          5. lower-+.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          6. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          7. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. Applied rewrites48.4%

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
        8. Taylor expanded in ux around 0

          \[\leadsto \sqrt{ux \cdot \left(2 - 2 \cdot maxCos\right)} \]
        9. Step-by-step derivation
          1. metadata-evalN/A

            \[\leadsto \sqrt{ux \cdot \left(2 - \left(\mathsf{neg}\left(-2\right)\right) \cdot maxCos\right)} \]
          2. fp-cancel-sign-sub-invN/A

            \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
          3. lower-*.f32N/A

            \[\leadsto \sqrt{ux \cdot \left(2 + -2 \cdot maxCos\right)} \]
          4. +-commutativeN/A

            \[\leadsto \sqrt{ux \cdot \left(-2 \cdot maxCos + 2\right)} \]
          5. lift-fma.f3265.3

            \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]
        10. Applied rewrites65.3%

          \[\leadsto \sqrt{ux \cdot \mathsf{fma}\left(-2, maxCos, 2\right)} \]
        11. Add Preprocessing

        Alternative 20: 6.6% accurate, 11.1× speedup?

        \[\begin{array}{l} \\ \sqrt{1 - 1} \end{array} \]
        (FPCore (ux uy maxCos) :precision binary32 (sqrt (- 1.0 1.0)))
        float code(float ux, float uy, float maxCos) {
        	return sqrtf((1.0f - 1.0f));
        }
        
        module fmin_fmax_functions
            implicit none
            private
            public fmax
            public fmin
        
            interface fmax
                module procedure fmax88
                module procedure fmax44
                module procedure fmax84
                module procedure fmax48
            end interface
            interface fmin
                module procedure fmin88
                module procedure fmin44
                module procedure fmin84
                module procedure fmin48
            end interface
        contains
            real(8) function fmax88(x, y) result (res)
                real(8), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(x, max(x, y), y /= y), x /= x)
            end function
            real(4) function fmax44(x, y) result (res)
                real(4), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(y, merge(x, max(x, y), y /= y), x /= x)
            end function
            real(8) function fmax84(x, y) result(res)
                real(8), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
            end function
            real(8) function fmax48(x, y) result(res)
                real(4), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
            end function
            real(8) function fmin88(x, y) result (res)
                real(8), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(x, min(x, y), y /= y), x /= x)
            end function
            real(4) function fmin44(x, y) result (res)
                real(4), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(y, merge(x, min(x, y), y /= y), x /= x)
            end function
            real(8) function fmin84(x, y) result(res)
                real(8), intent (in) :: x
                real(4), intent (in) :: y
                res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
            end function
            real(8) function fmin48(x, y) result(res)
                real(4), intent (in) :: x
                real(8), intent (in) :: y
                res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
            end function
        end module
        
        real(4) function code(ux, uy, maxcos)
        use fmin_fmax_functions
            real(4), intent (in) :: ux
            real(4), intent (in) :: uy
            real(4), intent (in) :: maxcos
            code = sqrt((1.0e0 - 1.0e0))
        end function
        
        function code(ux, uy, maxCos)
        	return sqrt(Float32(Float32(1.0) - Float32(1.0)))
        end
        
        function tmp = code(ux, uy, maxCos)
        	tmp = sqrt((single(1.0) - single(1.0)));
        end
        
        \begin{array}{l}
        
        \\
        \sqrt{1 - 1}
        \end{array}
        
        Derivation
        1. Initial program 57.4%

          \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        2. Add Preprocessing
        3. Step-by-step derivation
          1. lift--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\left(1 - ux\right)} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          2. flip--N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 \cdot 1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          3. lower-/.f32N/A

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1} - ux \cdot ux}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          5. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{{ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          6. lower--.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{\color{blue}{1 - {ux}^{2}}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          7. unpow2N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          8. lower-*.f32N/A

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - \color{blue}{ux \cdot ux}}{1 + ux} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          9. lower-+.f3256.7

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\frac{1 - ux \cdot ux}{\color{blue}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        4. Applied rewrites56.7%

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\color{blue}{\frac{1 - ux \cdot ux}{1 + ux}} + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. Taylor expanded in uy around 0

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)}} \]
        6. Step-by-step derivation
          1. lower-sqrt.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          2. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          3. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          4. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          5. lower-+.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          6. lower-*.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
          7. lower--.f32N/A

            \[\leadsto \sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\left(maxCos \cdot ux + \frac{1}{1 + ux}\right) - \frac{{ux}^{2}}{1 + ux}\right)} \]
        7. Applied rewrites48.4%

          \[\leadsto \color{blue}{\sqrt{1 - \left(\left(1 + maxCos \cdot ux\right) - ux\right) \cdot \left(\mathsf{fma}\left(maxCos, ux, \frac{1}{1 + ux}\right) - \frac{ux \cdot ux}{1 + ux}\right)}} \]
        8. Taylor expanded in ux around 0

          \[\leadsto \sqrt{1 - 1} \]
        9. Step-by-step derivation
          1. Applied rewrites6.6%

            \[\leadsto \sqrt{1 - 1} \]
          2. Add Preprocessing

          Reproduce

          ?
          herbie shell --seed 2025033 
          (FPCore (ux uy maxCos)
            :name "UniformSampleCone, x"
            :precision binary32
            :pre (and (and (and (<= 2.328306437e-10 ux) (<= ux 1.0)) (and (<= 2.328306437e-10 uy) (<= uy 1.0))) (and (<= 0.0 maxCos) (<= maxCos 1.0)))
            (* (cos (* (* uy 2.0) (PI))) (sqrt (- 1.0 (* (+ (- 1.0 ux) (* ux maxCos)) (+ (- 1.0 ux) (* ux maxCos)))))))