UniformSampleCone, x

Percentage Accurate: 57.3% → 99.2%
Time: 9.7s
Alternatives: 14
Speedup: 3.5×

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 14 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: 57.3% 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: 99.2% accurate, 0.6× speedup?

\[\begin{array}{l} \\ \sqrt{\mathsf{fma}\left(maxCos, -2, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{PI}\left(\right) \cdot \mathsf{fma}\left(-2, uy, 0.5\right)\right) \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (*
  (sqrt (* (fma maxCos -2.0 (- 2.0 (* (pow (- maxCos 1.0) 2.0) ux))) ux))
  (sin (* (PI) (fma -2.0 uy 0.5)))))
\begin{array}{l}

\\
\sqrt{\mathsf{fma}\left(maxCos, -2, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{PI}\left(\right) \cdot \mathsf{fma}\left(-2, uy, 0.5\right)\right)
\end{array}
Derivation
  1. Initial program 59.1%

    \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
  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

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

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

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

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

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

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

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

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 + \color{blue}{\left(\mathsf{neg}\left(ux\right)\right)} \cdot {\left(maxCos - 1\right)}^{2}\right) \cdot ux} \]
    9. fp-cancel-sub-signN/A

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

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

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

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

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - \color{blue}{{\left(maxCos - 1\right)}^{2}} \cdot ux\right) \cdot ux} \]
    14. lower--.f3298.9

      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\color{blue}{\left(maxCos - 1\right)}}^{2} \cdot ux\right) \cdot ux} \]
  5. Applied rewrites98.9%

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

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

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

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

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

      \[\leadsto \sin \left(\left(\mathsf{neg}\left(\color{blue}{\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)}\right)\right) + \frac{\mathsf{PI}\left(\right)}{2}\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
    6. *-commutativeN/A

      \[\leadsto \sin \left(\left(\mathsf{neg}\left(\color{blue}{\mathsf{PI}\left(\right) \cdot \left(uy \cdot 2\right)}\right)\right) + \frac{\mathsf{PI}\left(\right)}{2}\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
    7. distribute-lft-neg-inN/A

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

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

      \[\leadsto \sin \left(\mathsf{fma}\left(\color{blue}{-\mathsf{PI}\left(\right)}, uy \cdot 2, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
    10. lift-*.f32N/A

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

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

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

      \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
    14. lower-/.f3299.0

      \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
  7. Applied rewrites99.0%

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

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

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

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

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

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

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

      \[\leadsto \sin \left(\mathsf{fma}\left(\mathsf{PI}\left(\right) \cdot uy, -2, \color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
    7. lower-PI.f3299.0

      \[\leadsto \sin \left(\mathsf{fma}\left(\mathsf{PI}\left(\right) \cdot uy, -2, 0.5 \cdot \color{blue}{\mathsf{PI}\left(\right)}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
  10. Applied rewrites99.0%

    \[\leadsto \sin \color{blue}{\left(\mathsf{fma}\left(\mathsf{PI}\left(\right) \cdot uy, -2, 0.5 \cdot \mathsf{PI}\left(\right)\right)\right)} \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
  11. Step-by-step derivation
    1. lift-*.f32N/A

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

      \[\leadsto \color{blue}{\sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{fma}\left(\mathsf{PI}\left(\right) \cdot uy, -2, \frac{1}{2} \cdot \mathsf{PI}\left(\right)\right)\right)} \]
    3. lower-*.f3299.0

      \[\leadsto \color{blue}{\sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{fma}\left(\mathsf{PI}\left(\right) \cdot uy, -2, 0.5 \cdot \mathsf{PI}\left(\right)\right)\right)} \]
  12. Applied rewrites99.1%

    \[\leadsto \color{blue}{\sqrt{\mathsf{fma}\left(maxCos, -2, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{PI}\left(\right) \cdot \mathsf{fma}\left(-2, uy, 0.5\right)\right)} \]
  13. Final simplification99.1%

    \[\leadsto \sqrt{\mathsf{fma}\left(maxCos, -2, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \cdot \sin \left(\mathsf{PI}\left(\right) \cdot \mathsf{fma}\left(-2, uy, 0.5\right)\right) \]
  14. Add Preprocessing

Alternative 2: 75.2% accurate, 0.8× speedup?

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0} \leq 0.017999999225139618:\\ \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(ux - \mathsf{fma}\left(maxCos, ux, 1\right)\right) - -1}\\ \end{array} \end{array} \]
(FPCore (ux uy maxCos)
 :precision binary32
 (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
   (if (<=
        (* (cos (* (* uy 2.0) (PI))) (sqrt (- 1.0 (* t_0 t_0))))
        0.017999999225139618)
     (sqrt (* (- (fma -1.0 (- maxCos 1.0) 1.0) maxCos) ux))
     (sqrt
      (- (* (- (fma maxCos ux 1.0) ux) (- ux (fma maxCos ux 1.0))) -1.0)))))
\begin{array}{l}

\\
\begin{array}{l}
t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
\mathbf{if}\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - t\_0 \cdot t\_0} \leq 0.017999999225139618:\\
\;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\

\mathbf{else}:\\
\;\;\;\;\sqrt{\left(\mathsf{fma}\left(maxCos, ux, 1\right) - ux\right) \cdot \left(ux - \mathsf{fma}\left(maxCos, ux, 1\right)\right) - -1}\\


\end{array}
\end{array}
Derivation
  1. Split input into 2 regimes
  2. if (*.f32 (cos.f32 (*.f32 (*.f32 uy #s(literal 2 binary32)) (PI.f32))) (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)))))) < 0.0179999992

    1. Initial program 38.7%

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

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

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
      3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
      12. lower-neg.f3238.8

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

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

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

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

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
      17. lower-fma.f3238.8

        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
    4. Applied rewrites38.8%

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

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

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

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

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

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

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

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

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

        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
      9. lower-fma.f3232.8

        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
    7. Applied rewrites32.8%

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

      \[\leadsto \sqrt{ux \cdot \left(\left(1 + -1 \cdot \left(maxCos - 1\right)\right) - maxCos\right)} \]
    9. Step-by-step derivation
      1. Applied rewrites74.5%

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

      if 0.0179999992 < (*.f32 (cos.f32 (*.f32 (*.f32 uy #s(literal 2 binary32)) (PI.f32))) (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 89.2%

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

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
        3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
        12. lower-neg.f3288.7

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

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

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

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

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
        17. lower-fma.f3288.7

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
      4. Applied rewrites88.7%

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

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

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

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

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

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

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

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

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

          \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
        9. lower-fma.f3275.7

          \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
      7. Applied rewrites75.7%

        \[\leadsto \color{blue}{\sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)}} \]
      8. Step-by-step derivation
        1. Applied rewrites76.2%

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

      Alternative 3: 84.0% accurate, 1.0× speedup?

      \[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \leq 0.9999650120735168:\\ \;\;\;\;\mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \end{array} \end{array} \]
      (FPCore (ux uy maxCos)
       :precision binary32
       (if (<= (cos (* (* uy 2.0) (PI))) 0.9999650120735168)
         (*
          (fma (* (* uy uy) -2.0) (* (PI) (PI)) 1.0)
          (sqrt (* (fma -2.0 maxCos 2.0) ux)))
         (sqrt
          (*
           (-
            (- (fma (* (- 1.0 maxCos) ux) (- maxCos 1.0) (- (- maxCos) -1.0)) -1.0)
            maxCos)
           ux))))
      \begin{array}{l}
      
      \\
      \begin{array}{l}
      \mathbf{if}\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \leq 0.9999650120735168:\\
      \;\;\;\;\mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2\right) \cdot ux}\\
      
      \mathbf{else}:\\
      \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\
      
      
      \end{array}
      \end{array}
      
      Derivation
      1. Split input into 2 regimes
      2. if (cos.f32 (*.f32 (*.f32 uy #s(literal 2 binary32)) (PI.f32))) < 0.999965012

        1. Initial program 58.8%

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

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

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

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

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

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

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

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

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

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

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

            \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \color{blue}{\mathsf{PI}\left(\right)} \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
          11. lower-PI.f3241.3

            \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \color{blue}{\mathsf{PI}\left(\right)}, 1\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
        5. Applied rewrites41.3%

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

          \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\color{blue}{ux \cdot \left(2 - 2 \cdot maxCos\right)}} \]
        7. Step-by-step derivation
          1. metadata-evalN/A

            \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{ux \cdot \left(2 - \color{blue}{\left(\mathsf{neg}\left(-2\right)\right)} \cdot maxCos\right)} \]
          2. fp-cancel-sign-sub-invN/A

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

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

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

            \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\color{blue}{\left(-2 \cdot maxCos + 2\right)} \cdot ux} \]
          6. lower-fma.f3254.2

            \[\leadsto \mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\color{blue}{\mathsf{fma}\left(-2, maxCos, 2\right)} \cdot ux} \]
        8. Applied rewrites54.2%

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

        if 0.999965012 < (cos.f32 (*.f32 (*.f32 uy #s(literal 2 binary32)) (PI.f32)))

        1. Initial program 59.2%

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

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
          3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
          12. lower-neg.f3259.0

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

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

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

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
          17. lower-fma.f3259.0

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
        4. Applied rewrites59.0%

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

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

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

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

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

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

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

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

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

            \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
          9. lower-fma.f3258.0

            \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
        7. Applied rewrites58.0%

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

          \[\leadsto \sqrt{ux \cdot \left(\left(1 + \left(-1 \cdot \left(maxCos - 1\right) + ux \cdot \left(\left(1 - maxCos\right) \cdot \left(maxCos - 1\right)\right)\right)\right) - maxCos\right)} \]
        9. Step-by-step derivation
          1. Applied rewrites96.7%

            \[\leadsto \sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, -\left(maxCos - 1\right)\right) + 1\right) - maxCos\right) \cdot ux} \]
        10. Recombined 2 regimes into one program.
        11. Final simplification84.4%

          \[\leadsto \begin{array}{l} \mathbf{if}\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \leq 0.9999650120735168:\\ \;\;\;\;\mathsf{fma}\left(\left(uy \cdot uy\right) \cdot -2, \mathsf{PI}\left(\right) \cdot \mathsf{PI}\left(\right), 1\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \end{array} \]
        12. Add Preprocessing

        Alternative 4: 99.0% accurate, 1.0× speedup?

        \[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 - \left(\left(maxCos - 1\right) \cdot \left(-1 + maxCos\right)\right) \cdot ux\right) - 2 \cdot maxCos\right) \cdot ux} \end{array} \]
        (FPCore (ux uy maxCos)
         :precision binary32
         (*
          (cos (* (* uy 2.0) (PI)))
          (sqrt
           (*
            (- (- 2.0 (* (* (- maxCos 1.0) (+ -1.0 maxCos)) ux)) (* 2.0 maxCos))
            ux))))
        \begin{array}{l}
        
        \\
        \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(2 - \left(\left(maxCos - 1\right) \cdot \left(-1 + maxCos\right)\right) \cdot ux\right) - 2 \cdot maxCos\right) \cdot ux}
        \end{array}
        
        Derivation
        1. Initial program 59.1%

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(1 - \mathsf{fma}\left(maxCos, ux, 1 - ux\right) \cdot \left(1 - ux\right)\right) - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \color{blue}{\left(maxCos \cdot ux\right)}} \]
        4. Applied rewrites57.6%

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

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

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

            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\left(2 + -1 \cdot \left(ux \cdot \left(-1 \cdot \left(maxCos - 1\right) + maxCos \cdot \left(maxCos - 1\right)\right)\right)\right) - 2 \cdot maxCos\right) \cdot ux}} \]
        7. Applied rewrites98.9%

          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\left(2 - \left(\left(maxCos - 1\right) \cdot \left(-1 + maxCos\right)\right) \cdot ux\right) - 2 \cdot maxCos\right) \cdot ux}} \]
        8. Final simplification98.9%

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

        Alternative 5: 96.6% accurate, 1.0× speedup?

        \[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\ \end{array} \end{array} \]
        (FPCore (ux uy maxCos)
         :precision binary32
         (if (<= uy 7.999999797903001e-5)
           (sqrt
            (*
             (-
              (- (fma (* (- 1.0 maxCos) ux) (- maxCos 1.0) (- (- maxCos) -1.0)) -1.0)
              maxCos)
             ux))
           (* (sin (fma (- (PI)) (* 2.0 uy) (/ (PI) 2.0))) (sqrt (* (- 2.0 ux) ux)))))
        \begin{array}{l}
        
        \\
        \begin{array}{l}
        \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\
        \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\
        
        \mathbf{else}:\\
        \;\;\;\;\sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\
        
        
        \end{array}
        \end{array}
        
        Derivation
        1. Split input into 2 regimes
        2. if uy < 7.9999998e-5

          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. Step-by-step derivation
            1. lift--.f32N/A

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

              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
            3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
            12. lower-neg.f3258.3

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

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

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

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

              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
            17. lower-fma.f3258.3

              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
          4. Applied rewrites58.3%

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

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

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

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

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

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

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

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

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

              \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
            9. lower-fma.f3258.0

              \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
          7. Applied rewrites58.0%

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

            \[\leadsto \sqrt{ux \cdot \left(\left(1 + \left(-1 \cdot \left(maxCos - 1\right) + ux \cdot \left(\left(1 - maxCos\right) \cdot \left(maxCos - 1\right)\right)\right)\right) - maxCos\right)} \]
          9. Step-by-step derivation
            1. Applied rewrites99.1%

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

            if 7.9999998e-5 < uy

            1. Initial program 59.9%

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

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

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

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

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

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

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

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

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 + \color{blue}{\left(\mathsf{neg}\left(ux\right)\right)} \cdot {\left(maxCos - 1\right)}^{2}\right) \cdot ux} \]
              9. fp-cancel-sub-signN/A

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

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

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

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

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - \color{blue}{{\left(maxCos - 1\right)}^{2}} \cdot ux\right) \cdot ux} \]
              14. lower--.f3298.3

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\color{blue}{\left(maxCos - 1\right)}}^{2} \cdot ux\right) \cdot ux} \]
            5. Applied rewrites98.3%

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

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

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

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

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

                \[\leadsto \sin \left(\left(\mathsf{neg}\left(\color{blue}{\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)}\right)\right) + \frac{\mathsf{PI}\left(\right)}{2}\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
              6. *-commutativeN/A

                \[\leadsto \sin \left(\left(\mathsf{neg}\left(\color{blue}{\mathsf{PI}\left(\right) \cdot \left(uy \cdot 2\right)}\right)\right) + \frac{\mathsf{PI}\left(\right)}{2}\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
              7. distribute-lft-neg-inN/A

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

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

                \[\leadsto \sin \left(\mathsf{fma}\left(\color{blue}{-\mathsf{PI}\left(\right)}, uy \cdot 2, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
              10. lift-*.f32N/A

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

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

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

                \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
              14. lower-/.f3298.6

                \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\left(maxCos - 1\right)}^{2} \cdot ux\right) \cdot ux} \]
            7. Applied rewrites98.6%

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

              \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{ux \cdot \color{blue}{\left(2 - ux\right)}} \]
            9. Step-by-step derivation
              1. Applied rewrites93.3%

                \[\leadsto \sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot \color{blue}{ux}} \]
            10. Recombined 2 regimes into one program.
            11. Final simplification96.7%

              \[\leadsto \begin{array}{l} \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\sin \left(\mathsf{fma}\left(-\mathsf{PI}\left(\right), 2 \cdot uy, \frac{\mathsf{PI}\left(\right)}{2}\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\ \end{array} \]
            12. Add Preprocessing

            Alternative 6: 98.4% accurate, 1.0× speedup?

            \[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux - \left(\mathsf{fma}\left(-2, ux, 2\right) \cdot ux\right) \cdot maxCos} \end{array} \]
            (FPCore (ux uy maxCos)
             :precision binary32
             (*
              (cos (* (* uy 2.0) (PI)))
              (sqrt (- (* (- 2.0 ux) ux) (* (* (fma -2.0 ux 2.0) ux) maxCos)))))
            \begin{array}{l}
            
            \\
            \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux - \left(\mathsf{fma}\left(-2, ux, 2\right) \cdot ux\right) \cdot maxCos}
            \end{array}
            
            Derivation
            1. Initial program 59.1%

              \[\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)} \]
            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

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

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

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

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

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

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

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

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 + \color{blue}{\left(\mathsf{neg}\left(ux\right)\right)} \cdot {\left(maxCos - 1\right)}^{2}\right) \cdot ux} \]
              9. fp-cancel-sub-signN/A

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

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

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

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

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - \color{blue}{{\left(maxCos - 1\right)}^{2}} \cdot ux\right) \cdot ux} \]
              14. lower--.f3298.9

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\color{blue}{\left(maxCos - 1\right)}}^{2} \cdot ux\right) \cdot ux} \]
            5. Applied rewrites98.9%

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

              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{-1 \cdot \left(maxCos \cdot \left(ux \cdot \left(2 + -2 \cdot ux\right)\right)\right) + \color{blue}{ux \cdot \left(2 - ux\right)}} \]
            7. Step-by-step derivation
              1. Applied rewrites97.3%

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux - \color{blue}{\left(\mathsf{fma}\left(-2, ux, 2\right) \cdot ux\right) \cdot maxCos}} \]
              2. Final simplification97.3%

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

              Alternative 7: 97.8% accurate, 1.0× speedup?

              \[\begin{array}{l} \\ \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos - 1, ux, 2\right) - maxCos\right) \cdot ux - maxCos \cdot ux} \end{array} \]
              (FPCore (ux uy maxCos)
               :precision binary32
               (*
                (cos (* (* uy 2.0) (PI)))
                (sqrt (- (* (- (fma (- maxCos 1.0) ux 2.0) maxCos) ux) (* maxCos ux)))))
              \begin{array}{l}
              
              \\
              \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(maxCos - 1, ux, 2\right) - maxCos\right) \cdot ux - maxCos \cdot ux}
              \end{array}
              
              Derivation
              1. Initial program 59.1%

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(1 - \mathsf{fma}\left(maxCos, ux, 1 - ux\right) \cdot \left(1 - ux\right)\right) - \left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \color{blue}{\left(maxCos \cdot ux\right)}} \]
              4. Applied rewrites57.6%

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

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

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

                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\color{blue}{\left(\left(-1 \cdot \frac{maxCos}{ux} + 2 \cdot \frac{1}{ux}\right) - \left(1 + -1 \cdot maxCos\right)\right) \cdot {ux}^{2}} - \left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right) \cdot maxCos\right) \cdot ux} \]
              7. Applied rewrites98.8%

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

                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\frac{\left(-maxCos\right) - -2}{ux} - 1\right) - \left(-maxCos\right)\right) \cdot \left(ux \cdot ux\right) - \color{blue}{maxCos \cdot ux}} \]
              9. Step-by-step derivation
                1. lower-*.f3296.7

                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\left(\frac{\left(-maxCos\right) - -2}{ux} - 1\right) - \left(-maxCos\right)\right) \cdot \left(ux \cdot ux\right) - \color{blue}{maxCos \cdot ux}} \]
              10. Applied rewrites96.7%

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

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

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

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

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

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

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

                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\color{blue}{\mathsf{fma}\left(maxCos - 1, ux, 2\right)} - maxCos\right) \cdot ux - maxCos \cdot ux} \]
                7. lower--.f3296.8

                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(\mathsf{fma}\left(\color{blue}{maxCos - 1}, ux, 2\right) - maxCos\right) \cdot ux - maxCos \cdot ux} \]
              13. Applied rewrites96.8%

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

              Alternative 8: 96.4% accurate, 1.1× speedup?

              \[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\ \end{array} \end{array} \]
              (FPCore (ux uy maxCos)
               :precision binary32
               (if (<= uy 7.999999797903001e-5)
                 (sqrt
                  (*
                   (-
                    (- (fma (* (- 1.0 maxCos) ux) (- maxCos 1.0) (- (- maxCos) -1.0)) -1.0)
                    maxCos)
                   ux))
                 (* (cos (* (* uy 2.0) (PI))) (sqrt (* (- 2.0 ux) ux)))))
              \begin{array}{l}
              
              \\
              \begin{array}{l}
              \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\
              \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\
              
              \mathbf{else}:\\
              \;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\
              
              
              \end{array}
              \end{array}
              
              Derivation
              1. Split input into 2 regimes
              2. if uy < 7.9999998e-5

                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. Step-by-step derivation
                  1. lift--.f32N/A

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

                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                  3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                  12. lower-neg.f3258.3

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

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

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

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

                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                  17. lower-fma.f3258.3

                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                4. Applied rewrites58.3%

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

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

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

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

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

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

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

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

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

                    \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                  9. lower-fma.f3258.0

                    \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                7. Applied rewrites58.0%

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

                  \[\leadsto \sqrt{ux \cdot \left(\left(1 + \left(-1 \cdot \left(maxCos - 1\right) + ux \cdot \left(\left(1 - maxCos\right) \cdot \left(maxCos - 1\right)\right)\right)\right) - maxCos\right)} \]
                9. Step-by-step derivation
                  1. Applied rewrites99.1%

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

                  if 7.9999998e-5 < uy

                  1. Initial program 59.9%

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

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

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

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

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

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

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

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

                      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 + \color{blue}{\left(\mathsf{neg}\left(ux\right)\right)} \cdot {\left(maxCos - 1\right)}^{2}\right) \cdot ux} \]
                    9. fp-cancel-sub-signN/A

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

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

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

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

                      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - \color{blue}{{\left(maxCos - 1\right)}^{2}} \cdot ux\right) \cdot ux} \]
                    14. lower--.f3298.3

                      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(-2, maxCos, 2 - {\color{blue}{\left(maxCos - 1\right)}}^{2} \cdot ux\right) \cdot ux} \]
                  5. Applied rewrites98.3%

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

                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{ux \cdot \color{blue}{\left(2 - ux\right)}} \]
                  7. Step-by-step derivation
                    1. Applied rewrites93.0%

                      \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot \color{blue}{ux}} \]
                  8. Recombined 2 regimes into one program.
                  9. Final simplification96.6%

                    \[\leadsto \begin{array}{l} \mathbf{if}\;uy \leq 7.999999797903001 \cdot 10^{-5}:\\ \;\;\;\;\sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}\\ \mathbf{else}:\\ \;\;\;\;\cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\left(2 - ux\right) \cdot ux}\\ \end{array} \]
                  10. Add Preprocessing

                  Alternative 9: 75.2% accurate, 2.3× speedup?

                  \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\ \;\;\;\;\sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\ \end{array} \end{array} \]
                  (FPCore (ux uy maxCos)
                   :precision binary32
                   (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
                     (if (<= (* t_0 t_0) 0.9997000098228455)
                       (sqrt (fma (- ux (fma maxCos ux 1.0)) (- (fma maxCos ux 1.0) ux) 1.0))
                       (sqrt (* (- (fma -1.0 (- maxCos 1.0) 1.0) maxCos) ux)))))
                  float code(float ux, float uy, float maxCos) {
                  	float t_0 = (1.0f - ux) + (ux * maxCos);
                  	float tmp;
                  	if ((t_0 * t_0) <= 0.9997000098228455f) {
                  		tmp = sqrtf(fmaf((ux - fmaf(maxCos, ux, 1.0f)), (fmaf(maxCos, ux, 1.0f) - ux), 1.0f));
                  	} else {
                  		tmp = sqrtf(((fmaf(-1.0f, (maxCos - 1.0f), 1.0f) - maxCos) * ux));
                  	}
                  	return tmp;
                  }
                  
                  function code(ux, uy, maxCos)
                  	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
                  	tmp = Float32(0.0)
                  	if (Float32(t_0 * t_0) <= Float32(0.9997000098228455))
                  		tmp = sqrt(fma(Float32(ux - fma(maxCos, ux, Float32(1.0))), Float32(fma(maxCos, ux, Float32(1.0)) - ux), Float32(1.0)));
                  	else
                  		tmp = sqrt(Float32(Float32(fma(Float32(-1.0), Float32(maxCos - Float32(1.0)), Float32(1.0)) - maxCos) * ux));
                  	end
                  	return tmp
                  end
                  
                  \begin{array}{l}
                  
                  \\
                  \begin{array}{l}
                  t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
                  \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\
                  \;\;\;\;\sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)}\\
                  
                  \mathbf{else}:\\
                  \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\
                  
                  
                  \end{array}
                  \end{array}
                  
                  Derivation
                  1. Split input into 2 regimes
                  2. if (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))) < 0.99970001

                    1. Initial program 88.6%

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

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

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                      3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                      12. lower-neg.f3288.2

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

                        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                      9. lower-fma.f3272.2

                        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                    7. Applied rewrites72.2%

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

                    if 0.99970001 < (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))

                    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

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

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                      3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                      12. lower-neg.f3236.8

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

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

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

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

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                      17. lower-fma.f3236.8

                        \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                    4. Applied rewrites36.8%

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

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

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

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

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

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

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

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

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

                        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                      9. lower-fma.f3233.3

                        \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                    7. Applied rewrites33.3%

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

                      \[\leadsto \sqrt{ux \cdot \left(\left(1 + -1 \cdot \left(maxCos - 1\right)\right) - maxCos\right)} \]
                    9. Step-by-step derivation
                      1. Applied rewrites77.2%

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

                    Alternative 10: 74.2% accurate, 2.5× speedup?

                    \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\ \;\;\;\;\sqrt{\mathsf{fma}\left(ux - 1, \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\ \end{array} \end{array} \]
                    (FPCore (ux uy maxCos)
                     :precision binary32
                     (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
                       (if (<= (* t_0 t_0) 0.9997000098228455)
                         (sqrt (fma (- ux 1.0) (- (fma maxCos ux 1.0) ux) 1.0))
                         (sqrt (* (- (fma -1.0 (- maxCos 1.0) 1.0) maxCos) ux)))))
                    float code(float ux, float uy, float maxCos) {
                    	float t_0 = (1.0f - ux) + (ux * maxCos);
                    	float tmp;
                    	if ((t_0 * t_0) <= 0.9997000098228455f) {
                    		tmp = sqrtf(fmaf((ux - 1.0f), (fmaf(maxCos, ux, 1.0f) - ux), 1.0f));
                    	} else {
                    		tmp = sqrtf(((fmaf(-1.0f, (maxCos - 1.0f), 1.0f) - maxCos) * ux));
                    	}
                    	return tmp;
                    }
                    
                    function code(ux, uy, maxCos)
                    	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
                    	tmp = Float32(0.0)
                    	if (Float32(t_0 * t_0) <= Float32(0.9997000098228455))
                    		tmp = sqrt(fma(Float32(ux - Float32(1.0)), Float32(fma(maxCos, ux, Float32(1.0)) - ux), Float32(1.0)));
                    	else
                    		tmp = sqrt(Float32(Float32(fma(Float32(-1.0), Float32(maxCos - Float32(1.0)), Float32(1.0)) - maxCos) * ux));
                    	end
                    	return tmp
                    end
                    
                    \begin{array}{l}
                    
                    \\
                    \begin{array}{l}
                    t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
                    \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\
                    \;\;\;\;\sqrt{\mathsf{fma}\left(ux - 1, \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)}\\
                    
                    \mathbf{else}:\\
                    \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\
                    
                    
                    \end{array}
                    \end{array}
                    
                    Derivation
                    1. Split input into 2 regimes
                    2. if (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))) < 0.99970001

                      1. Initial program 88.6%

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

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

                          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                        3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                          \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                        12. lower-neg.f3288.2

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

                          \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                        9. lower-fma.f3272.2

                          \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                      7. Applied rewrites72.2%

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

                        \[\leadsto \sqrt{\mathsf{fma}\left(ux - 1, \mathsf{fma}\left(maxCos, ux, 1\right) - ux, 1\right)} \]
                      9. Step-by-step derivation
                        1. Applied rewrites68.2%

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

                        if 0.99970001 < (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))

                        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

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

                            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                          3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                          12. lower-neg.f3236.8

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

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

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

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

                            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                          17. lower-fma.f3236.8

                            \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                        4. Applied rewrites36.8%

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

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

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

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

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

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

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

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

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

                            \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                          9. lower-fma.f3233.3

                            \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                        7. Applied rewrites33.3%

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

                          \[\leadsto \sqrt{ux \cdot \left(\left(1 + -1 \cdot \left(maxCos - 1\right)\right) - maxCos\right)} \]
                        9. Step-by-step derivation
                          1. Applied rewrites77.2%

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

                        Alternative 11: 74.1% accurate, 2.6× speedup?

                        \[\begin{array}{l} \\ \begin{array}{l} t_0 := \left(1 - ux\right) + ux \cdot maxCos\\ \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\ \;\;\;\;\sqrt{\mathsf{fma}\left(ux - 1, 1 - ux, 1\right)}\\ \mathbf{else}:\\ \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\ \end{array} \end{array} \]
                        (FPCore (ux uy maxCos)
                         :precision binary32
                         (let* ((t_0 (+ (- 1.0 ux) (* ux maxCos))))
                           (if (<= (* t_0 t_0) 0.9997000098228455)
                             (sqrt (fma (- ux 1.0) (- 1.0 ux) 1.0))
                             (sqrt (* (- (fma -1.0 (- maxCos 1.0) 1.0) maxCos) ux)))))
                        float code(float ux, float uy, float maxCos) {
                        	float t_0 = (1.0f - ux) + (ux * maxCos);
                        	float tmp;
                        	if ((t_0 * t_0) <= 0.9997000098228455f) {
                        		tmp = sqrtf(fmaf((ux - 1.0f), (1.0f - ux), 1.0f));
                        	} else {
                        		tmp = sqrtf(((fmaf(-1.0f, (maxCos - 1.0f), 1.0f) - maxCos) * ux));
                        	}
                        	return tmp;
                        }
                        
                        function code(ux, uy, maxCos)
                        	t_0 = Float32(Float32(Float32(1.0) - ux) + Float32(ux * maxCos))
                        	tmp = Float32(0.0)
                        	if (Float32(t_0 * t_0) <= Float32(0.9997000098228455))
                        		tmp = sqrt(fma(Float32(ux - Float32(1.0)), Float32(Float32(1.0) - ux), Float32(1.0)));
                        	else
                        		tmp = sqrt(Float32(Float32(fma(Float32(-1.0), Float32(maxCos - Float32(1.0)), Float32(1.0)) - maxCos) * ux));
                        	end
                        	return tmp
                        end
                        
                        \begin{array}{l}
                        
                        \\
                        \begin{array}{l}
                        t_0 := \left(1 - ux\right) + ux \cdot maxCos\\
                        \mathbf{if}\;t\_0 \cdot t\_0 \leq 0.9997000098228455:\\
                        \;\;\;\;\sqrt{\mathsf{fma}\left(ux - 1, 1 - ux, 1\right)}\\
                        
                        \mathbf{else}:\\
                        \;\;\;\;\sqrt{\left(\mathsf{fma}\left(-1, maxCos - 1, 1\right) - maxCos\right) \cdot ux}\\
                        
                        
                        \end{array}
                        \end{array}
                        
                        Derivation
                        1. Split input into 2 regimes
                        2. if (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos))) < 0.99970001

                          1. Initial program 88.6%

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

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

                              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                            3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                              \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                            12. lower-neg.f3288.2

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

                              \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                            9. lower-fma.f3272.2

                              \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                          7. Applied rewrites72.2%

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

                            \[\leadsto \sqrt{1 + \left(1 - ux\right) \cdot \left(ux - 1\right)} \]
                          9. Step-by-step derivation
                            1. Applied rewrites67.9%

                              \[\leadsto \sqrt{\mathsf{fma}\left(ux - 1, 1 - ux, 1\right)} \]

                            if 0.99970001 < (*.f32 (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)) (+.f32 (-.f32 #s(literal 1 binary32) ux) (*.f32 ux maxCos)))

                            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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                              3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                              12. lower-neg.f3236.8

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

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

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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                              17. lower-fma.f3236.8

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                            4. Applied rewrites36.8%

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

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

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

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

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

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

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

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

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

                                \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                              9. lower-fma.f3233.3

                                \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                            7. Applied rewrites33.3%

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

                              \[\leadsto \sqrt{ux \cdot \left(\left(1 + -1 \cdot \left(maxCos - 1\right)\right) - maxCos\right)} \]
                            9. Step-by-step derivation
                              1. Applied rewrites77.2%

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

                            Alternative 12: 80.0% accurate, 3.5× speedup?

                            \[\begin{array}{l} \\ \sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux} \end{array} \]
                            (FPCore (ux uy maxCos)
                             :precision binary32
                             (sqrt
                              (*
                               (-
                                (- (fma (* (- 1.0 maxCos) ux) (- maxCos 1.0) (- (- maxCos) -1.0)) -1.0)
                                maxCos)
                               ux)))
                            float code(float ux, float uy, float maxCos) {
                            	return sqrtf((((fmaf(((1.0f - maxCos) * ux), (maxCos - 1.0f), (-maxCos - -1.0f)) - -1.0f) - maxCos) * ux));
                            }
                            
                            function code(ux, uy, maxCos)
                            	return sqrt(Float32(Float32(Float32(fma(Float32(Float32(Float32(1.0) - maxCos) * ux), Float32(maxCos - Float32(1.0)), Float32(Float32(-maxCos) - Float32(-1.0))) - Float32(-1.0)) - maxCos) * ux))
                            end
                            
                            \begin{array}{l}
                            
                            \\
                            \sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, \left(-maxCos\right) - -1\right) - -1\right) - maxCos\right) \cdot ux}
                            \end{array}
                            
                            Derivation
                            1. Initial program 59.1%

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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                              3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                              12. lower-neg.f3258.9

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

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

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

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

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                              17. lower-fma.f3258.9

                                \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                            4. Applied rewrites58.9%

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

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

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

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

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

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

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

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

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

                                \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                              9. lower-fma.f3250.0

                                \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                            7. Applied rewrites50.0%

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

                              \[\leadsto \sqrt{ux \cdot \left(\left(1 + \left(-1 \cdot \left(maxCos - 1\right) + ux \cdot \left(\left(1 - maxCos\right) \cdot \left(maxCos - 1\right)\right)\right)\right) - maxCos\right)} \]
                            9. Step-by-step derivation
                              1. Applied rewrites80.5%

                                \[\leadsto \sqrt{\left(\left(\mathsf{fma}\left(\left(1 - maxCos\right) \cdot ux, maxCos - 1, -\left(maxCos - 1\right)\right) + 1\right) - maxCos\right) \cdot ux} \]
                              2. Final simplification80.5%

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

                              Alternative 13: 48.0% accurate, 6.8× speedup?

                              \[\begin{array}{l} \\ \sqrt{\mathsf{fma}\left(ux - 1, 1 - ux, 1\right)} \end{array} \]
                              (FPCore (ux uy maxCos)
                               :precision binary32
                               (sqrt (fma (- ux 1.0) (- 1.0 ux) 1.0)))
                              float code(float ux, float uy, float maxCos) {
                              	return sqrtf(fmaf((ux - 1.0f), (1.0f - ux), 1.0f));
                              }
                              
                              function code(ux, uy, maxCos)
                              	return sqrt(fma(Float32(ux - Float32(1.0)), Float32(Float32(1.0) - ux), Float32(1.0)))
                              end
                              
                              \begin{array}{l}
                              
                              \\
                              \sqrt{\mathsf{fma}\left(ux - 1, 1 - ux, 1\right)}
                              \end{array}
                              
                              Derivation
                              1. Initial program 59.1%

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

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

                                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                                3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                                12. lower-neg.f3258.9

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

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

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

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

                                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                                17. lower-fma.f3258.9

                                  \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                              4. Applied rewrites58.9%

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

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

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

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

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

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

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

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

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

                                  \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                                9. lower-fma.f3250.0

                                  \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                              7. Applied rewrites50.0%

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

                                \[\leadsto \sqrt{1 + \left(1 - ux\right) \cdot \left(ux - 1\right)} \]
                              9. Step-by-step derivation
                                1. Applied rewrites47.8%

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

                                Alternative 14: -0.0% accurate, 7.4× speedup?

                                \[\begin{array}{l} \\ \left(\sqrt{-1} \cdot ux\right) \cdot maxCos \end{array} \]
                                (FPCore (ux uy maxCos) :precision binary32 (* (* (sqrt -1.0) ux) maxCos))
                                float code(float ux, float uy, float maxCos) {
                                	return (sqrtf(-1.0f) * ux) * maxCos;
                                }
                                
                                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)) * ux) * maxcos
                                end function
                                
                                function code(ux, uy, maxCos)
                                	return Float32(Float32(sqrt(Float32(-1.0)) * ux) * maxCos)
                                end
                                
                                function tmp = code(ux, uy, maxCos)
                                	tmp = (sqrt(single(-1.0)) * ux) * maxCos;
                                end
                                
                                \begin{array}{l}
                                
                                \\
                                \left(\sqrt{-1} \cdot ux\right) \cdot maxCos
                                \end{array}
                                
                                Derivation
                                1. Initial program 59.1%

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

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

                                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{1 - \color{blue}{\left(\left(1 - ux\right) + ux \cdot maxCos\right) \cdot \left(\left(1 - ux\right) + ux \cdot maxCos\right)}} \]
                                  3. fp-cancel-sub-sign-invN/A

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

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

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

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

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

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

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

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

                                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, \mathsf{neg}\left(\left(\left(1 - ux\right) + ux \cdot maxCos\right)\right), 1\right)} \]
                                  12. lower-neg.f3258.9

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

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

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

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

                                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\left(\color{blue}{maxCos \cdot ux} + \left(1 - ux\right)\right), 1\right)} \]
                                  17. lower-fma.f3258.9

                                    \[\leadsto \cos \left(\left(uy \cdot 2\right) \cdot \mathsf{PI}\left(\right)\right) \cdot \sqrt{\mathsf{fma}\left(\mathsf{fma}\left(maxCos, ux, 1 - ux\right), -\color{blue}{\mathsf{fma}\left(maxCos, ux, 1 - ux\right)}, 1\right)} \]
                                4. Applied rewrites58.9%

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

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

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

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

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

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

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

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

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

                                    \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\left(maxCos \cdot ux + 1\right)} - ux, 1\right)} \]
                                  9. lower-fma.f3250.0

                                    \[\leadsto \sqrt{\mathsf{fma}\left(ux - \mathsf{fma}\left(maxCos, ux, 1\right), \color{blue}{\mathsf{fma}\left(maxCos, ux, 1\right)} - ux, 1\right)} \]
                                7. Applied rewrites50.0%

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

                                  \[\leadsto maxCos \cdot \color{blue}{\left(ux \cdot \sqrt{-1}\right)} \]
                                9. Step-by-step derivation
                                  1. Applied rewrites-0.0%

                                    \[\leadsto \left(\sqrt{-1} \cdot ux\right) \cdot \color{blue}{maxCos} \]
                                  2. Add Preprocessing

                                  Reproduce

                                  ?
                                  herbie shell --seed 2025010 
                                  (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)))))))