NMSE Section 6.1 mentioned, B

Percentage Accurate: 78.8% → 99.7%
Time: 8.2s
Alternatives: 9
Speedup: 2.4×

Specification

?
\[\begin{array}{l} \\ \left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \end{array} \]
(FPCore (a b)
 :precision binary64
 (* (* (/ (PI) 2.0) (/ 1.0 (- (* b b) (* a a)))) (- (/ 1.0 a) (/ 1.0 b))))
\begin{array}{l}

\\
\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right)
\end{array}

Sampling outcomes in binary64 precision:

Local Percentage Accuracy vs ?

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

Accuracy vs Speed?

Herbie found 9 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: 78.8% accurate, 1.0× speedup?

\[\begin{array}{l} \\ \left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \end{array} \]
(FPCore (a b)
 :precision binary64
 (* (* (/ (PI) 2.0) (/ 1.0 (- (* b b) (* a a)))) (- (/ 1.0 a) (/ 1.0 b))))
\begin{array}{l}

\\
\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right)
\end{array}

Alternative 1: 99.7% accurate, 2.0× speedup?

\[\begin{array}{l} \\ \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a \cdot b}}{a + b} \end{array} \]
(FPCore (a b) :precision binary64 (/ (/ (* 0.5 (PI)) (* a b)) (+ a b)))
\begin{array}{l}

\\
\frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a \cdot b}}{a + b}
\end{array}
Derivation
  1. Initial program 76.6%

    \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
  2. Add Preprocessing
  3. Applied rewrites88.7%

    \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{\mathsf{PI}\left(\right)}{a + b}, \frac{0.5}{b - a} \cdot {a}^{-1}, \frac{\frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{b - a}}{-b}\right)} \]
  4. Applied rewrites99.7%

    \[\leadsto \color{blue}{\frac{0.5 \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
  5. Step-by-step derivation
    1. lift-*.f64N/A

      \[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}}{a + b} \]
    2. *-commutativeN/A

      \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{1}{2}}}{a + b} \]
    3. lift-*.f64N/A

      \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)} \cdot \frac{1}{2}}{a + b} \]
    4. *-rgt-identityN/A

      \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
    5. lift-/.f64N/A

      \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
    6. associate-*l/N/A

      \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{a \cdot b}}}{a + b} \]
    7. metadata-evalN/A

      \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
    8. div-invN/A

      \[\leadsto \frac{\frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}}{a \cdot b}}{a + b} \]
    9. lift-PI.f64N/A

      \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
    10. lower-/.f64N/A

      \[\leadsto \frac{\color{blue}{\frac{\frac{\mathsf{PI}\left(\right)}{2}}{a \cdot b}}}{a + b} \]
    11. lift-PI.f64N/A

      \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
    12. div-invN/A

      \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{a \cdot b}}{a + b} \]
    13. metadata-evalN/A

      \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
    14. lower-*.f6499.7

      \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot 0.5}}{a \cdot b}}{a + b} \]
    15. lift-*.f64N/A

      \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{a \cdot b}}}{a + b} \]
    16. *-commutativeN/A

      \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{b \cdot a}}}{a + b} \]
    17. lower-*.f6499.7

      \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{\color{blue}{b \cdot a}}}{a + b} \]
  6. Applied rewrites99.7%

    \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{b \cdot a}}}{a + b} \]
  7. Final simplification99.7%

    \[\leadsto \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a \cdot b}}{a + b} \]
  8. Add Preprocessing

Alternative 2: 96.0% accurate, 2.0× speedup?

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;a \leq -5 \cdot 10^{+121}:\\ \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5\\ \mathbf{else}:\\ \;\;\;\;\frac{0.5}{\left(\left(a + b\right) \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right)\\ \end{array} \end{array} \]
(FPCore (a b)
 :precision binary64
 (if (<= a -5e+121)
   (* (/ (PI) (* (* a b) a)) 0.5)
   (* (/ 0.5 (* (* (+ a b) a) b)) (PI))))
\begin{array}{l}

\\
\begin{array}{l}
\mathbf{if}\;a \leq -5 \cdot 10^{+121}:\\
\;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5\\

\mathbf{else}:\\
\;\;\;\;\frac{0.5}{\left(\left(a + b\right) \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right)\\


\end{array}
\end{array}
Derivation
  1. Split input into 2 regimes
  2. if a < -5.00000000000000007e121

    1. Initial program 62.3%

      \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
    2. Add Preprocessing
    3. Taylor expanded in a around inf

      \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \]
    4. Step-by-step derivation
      1. *-commutativeN/A

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
      2. lower-*.f64N/A

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
      3. lower-/.f64N/A

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
      4. lower-PI.f64N/A

        \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{{a}^{2} \cdot b} \cdot \frac{1}{2} \]
      5. lower-*.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
      6. unpow2N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot \frac{1}{2} \]
      7. lower-*.f6475.5

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot 0.5 \]
    5. Applied rewrites75.5%

      \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(a \cdot a\right) \cdot b} \cdot 0.5} \]
    6. Step-by-step derivation
      1. Applied rewrites99.8%

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5 \]

      if -5.00000000000000007e121 < a

      1. Initial program 79.1%

        \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
      2. Add Preprocessing
      3. Applied rewrites90.9%

        \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{\mathsf{PI}\left(\right)}{a + b}, \frac{0.5}{b - a} \cdot {a}^{-1}, \frac{\frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{b - a}}{-b}\right)} \]
      4. Applied rewrites99.7%

        \[\leadsto \color{blue}{\frac{0.5 \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
      5. Step-by-step derivation
        1. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}}{a + b} \]
        2. *-commutativeN/A

          \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{1}{2}}}{a + b} \]
        3. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)} \cdot \frac{1}{2}}{a + b} \]
        4. *-rgt-identityN/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
        5. lift-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
        6. associate-*l/N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{a \cdot b}}}{a + b} \]
        7. metadata-evalN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
        8. div-invN/A

          \[\leadsto \frac{\frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}}{a \cdot b}}{a + b} \]
        9. lift-PI.f64N/A

          \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
        10. lower-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\frac{\mathsf{PI}\left(\right)}{2}}{a \cdot b}}}{a + b} \]
        11. lift-PI.f64N/A

          \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
        12. div-invN/A

          \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{a \cdot b}}{a + b} \]
        13. metadata-evalN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
        14. lower-*.f6499.7

          \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot 0.5}}{a \cdot b}}{a + b} \]
        15. lift-*.f64N/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{a \cdot b}}}{a + b} \]
        16. *-commutativeN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{b \cdot a}}}{a + b} \]
        17. lower-*.f6499.7

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{\color{blue}{b \cdot a}}}{a + b} \]
      6. Applied rewrites99.7%

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{b \cdot a}}}{a + b} \]
      7. Step-by-step derivation
        1. lift-/.f64N/A

          \[\leadsto \color{blue}{\frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}{a + b}} \]
        2. lift-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}}{a + b} \]
        3. associate-/l/N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\left(a + b\right) \cdot \left(b \cdot a\right)}} \]
        4. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{\left(a + b\right) \cdot \left(b \cdot a\right)} \]
        5. associate-/l*N/A

          \[\leadsto \color{blue}{\mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\left(a + b\right) \cdot \left(b \cdot a\right)}} \]
        6. lower-*.f64N/A

          \[\leadsto \color{blue}{\mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\left(a + b\right) \cdot \left(b \cdot a\right)}} \]
        7. lift-+.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{\left(a + b\right)} \cdot \left(b \cdot a\right)} \]
        8. +-commutativeN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
        9. lift-+.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
        10. *-commutativeN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{\left(b \cdot a\right) \cdot \left(b + a\right)}} \]
        11. lower-/.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \color{blue}{\frac{\frac{1}{2}}{\left(b \cdot a\right) \cdot \left(b + a\right)}} \]
        12. lift-*.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{\left(b \cdot a\right)} \cdot \left(b + a\right)} \]
        13. associate-*l*N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{b \cdot \left(a \cdot \left(b + a\right)\right)}} \]
        14. lift-+.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(b + a\right)}\right)} \]
        15. +-commutativeN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(a + b\right)}\right)} \]
        16. distribute-rgt-outN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \color{blue}{\left(a \cdot a + b \cdot a\right)}} \]
        17. lower-*.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{\color{blue}{b \cdot \left(a \cdot a + b \cdot a\right)}} \]
        18. distribute-rgt-outN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \color{blue}{\left(a \cdot \left(a + b\right)\right)}} \]
        19. +-commutativeN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(b + a\right)}\right)} \]
        20. lift-+.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(b + a\right)}\right)} \]
        21. lower-*.f6495.4

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{0.5}{b \cdot \color{blue}{\left(a \cdot \left(b + a\right)\right)}} \]
        22. lift-+.f64N/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(b + a\right)}\right)} \]
        23. +-commutativeN/A

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{\frac{1}{2}}{b \cdot \left(a \cdot \color{blue}{\left(a + b\right)}\right)} \]
        24. lift-+.f6495.4

          \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{0.5}{b \cdot \left(a \cdot \color{blue}{\left(a + b\right)}\right)} \]
      8. Applied rewrites95.4%

        \[\leadsto \color{blue}{\mathsf{PI}\left(\right) \cdot \frac{0.5}{b \cdot \left(a \cdot \left(a + b\right)\right)}} \]
    7. Recombined 2 regimes into one program.
    8. Final simplification96.0%

      \[\leadsto \begin{array}{l} \mathbf{if}\;a \leq -5 \cdot 10^{+121}:\\ \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5\\ \mathbf{else}:\\ \;\;\;\;\frac{0.5}{\left(\left(a + b\right) \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right)\\ \end{array} \]
    9. Add Preprocessing

    Alternative 3: 99.7% accurate, 2.0× speedup?

    \[\begin{array}{l} \\ \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a + b}}{a \cdot b} \end{array} \]
    (FPCore (a b) :precision binary64 (/ (/ (* 0.5 (PI)) (+ a b)) (* a b)))
    \begin{array}{l}
    
    \\
    \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a + b}}{a \cdot b}
    \end{array}
    
    Derivation
    1. Initial program 76.6%

      \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
    2. Add Preprocessing
    3. Applied rewrites88.7%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{\mathsf{PI}\left(\right)}{a + b}, \frac{0.5}{b - a} \cdot {a}^{-1}, \frac{\frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{b - a}}{-b}\right)} \]
    4. Applied rewrites99.7%

      \[\leadsto \color{blue}{\frac{0.5 \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
    5. Step-by-step derivation
      1. lift-*.f64N/A

        \[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}}{a + b} \]
      2. *-commutativeN/A

        \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{1}{2}}}{a + b} \]
      3. lift-*.f64N/A

        \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)} \cdot \frac{1}{2}}{a + b} \]
      4. *-rgt-identityN/A

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
      5. lift-/.f64N/A

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
      6. associate-*l/N/A

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{a \cdot b}}}{a + b} \]
      7. metadata-evalN/A

        \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
      8. div-invN/A

        \[\leadsto \frac{\frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}}{a \cdot b}}{a + b} \]
      9. lift-PI.f64N/A

        \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
      10. lower-/.f64N/A

        \[\leadsto \frac{\color{blue}{\frac{\frac{\mathsf{PI}\left(\right)}{2}}{a \cdot b}}}{a + b} \]
      11. lift-PI.f64N/A

        \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
      12. div-invN/A

        \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{a \cdot b}}{a + b} \]
      13. metadata-evalN/A

        \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
      14. lower-*.f6499.7

        \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot 0.5}}{a \cdot b}}{a + b} \]
      15. lift-*.f64N/A

        \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{a \cdot b}}}{a + b} \]
      16. *-commutativeN/A

        \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{b \cdot a}}}{a + b} \]
      17. lower-*.f6499.7

        \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{\color{blue}{b \cdot a}}}{a + b} \]
    6. Applied rewrites99.7%

      \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{b \cdot a}}}{a + b} \]
    7. Step-by-step derivation
      1. lift-/.f64N/A

        \[\leadsto \color{blue}{\frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}{a + b}} \]
      2. lift-/.f64N/A

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}}{a + b} \]
      3. associate-/l/N/A

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\left(a + b\right) \cdot \left(b \cdot a\right)}} \]
      4. lift-+.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(a + b\right)} \cdot \left(b \cdot a\right)} \]
      5. +-commutativeN/A

        \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
      6. lift-+.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
      7. associate-/r*N/A

        \[\leadsto \color{blue}{\frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b + a}}{b \cdot a}} \]
      8. lower-/.f64N/A

        \[\leadsto \color{blue}{\frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b + a}}{b \cdot a}} \]
      9. lower-/.f6499.6

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{b + a}}}{b \cdot a} \]
      10. lift-*.f64N/A

        \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{b + a}}{b \cdot a} \]
      11. *-commutativeN/A

        \[\leadsto \frac{\frac{\color{blue}{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}}{b + a}}{b \cdot a} \]
      12. lower-*.f6499.6

        \[\leadsto \frac{\frac{\color{blue}{0.5 \cdot \mathsf{PI}\left(\right)}}{b + a}}{b \cdot a} \]
      13. lift-+.f64N/A

        \[\leadsto \frac{\frac{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}{\color{blue}{b + a}}}{b \cdot a} \]
      14. +-commutativeN/A

        \[\leadsto \frac{\frac{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}{\color{blue}{a + b}}}{b \cdot a} \]
      15. lift-+.f6499.6

        \[\leadsto \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{\color{blue}{a + b}}}{b \cdot a} \]
      16. lift-*.f64N/A

        \[\leadsto \frac{\frac{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}{a + b}}{\color{blue}{b \cdot a}} \]
      17. *-commutativeN/A

        \[\leadsto \frac{\frac{\frac{1}{2} \cdot \mathsf{PI}\left(\right)}{a + b}}{\color{blue}{a \cdot b}} \]
      18. lower-*.f6499.6

        \[\leadsto \frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a + b}}{\color{blue}{a \cdot b}} \]
    8. Applied rewrites99.6%

      \[\leadsto \color{blue}{\frac{\frac{0.5 \cdot \mathsf{PI}\left(\right)}{a + b}}{a \cdot b}} \]
    9. Add Preprocessing

    Alternative 4: 99.6% accurate, 2.0× speedup?

    \[\begin{array}{l} \\ \frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{a \cdot b} \end{array} \]
    (FPCore (a b) :precision binary64 (* (/ 0.5 (+ a b)) (/ (PI) (* a b))))
    \begin{array}{l}
    
    \\
    \frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{a \cdot b}
    \end{array}
    
    Derivation
    1. Initial program 76.6%

      \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
    2. Add Preprocessing
    3. Applied rewrites88.7%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{\mathsf{PI}\left(\right)}{a + b}, \frac{0.5}{b - a} \cdot {a}^{-1}, \frac{\frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{b - a}}{-b}\right)} \]
    4. Applied rewrites99.7%

      \[\leadsto \color{blue}{\frac{0.5 \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
    5. Step-by-step derivation
      1. lift-/.f64N/A

        \[\leadsto \color{blue}{\frac{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
      2. lift-*.f64N/A

        \[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}}{a + b} \]
      3. *-commutativeN/A

        \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{1}{2}}}{a + b} \]
      4. associate-/l*N/A

        \[\leadsto \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{\frac{1}{2}}{a + b}} \]
      5. lower-*.f64N/A

        \[\leadsto \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{\frac{1}{2}}{a + b}} \]
      6. lift-*.f64N/A

        \[\leadsto \color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)} \cdot \frac{\frac{1}{2}}{a + b} \]
      7. *-rgt-identityN/A

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{\frac{1}{2}}{a + b} \]
      8. lift-*.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{a \cdot b}} \cdot \frac{\frac{1}{2}}{a + b} \]
      9. *-commutativeN/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{b \cdot a}} \cdot \frac{\frac{1}{2}}{a + b} \]
      10. lower-*.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{b \cdot a}} \cdot \frac{\frac{1}{2}}{a + b} \]
      11. lower-/.f6499.6

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{b \cdot a} \cdot \color{blue}{\frac{0.5}{a + b}} \]
      12. lift-+.f64N/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{b \cdot a} \cdot \frac{\frac{1}{2}}{\color{blue}{a + b}} \]
      13. +-commutativeN/A

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{b \cdot a} \cdot \frac{\frac{1}{2}}{\color{blue}{b + a}} \]
      14. lower-+.f6499.6

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{b \cdot a} \cdot \frac{0.5}{\color{blue}{b + a}} \]
    6. Applied rewrites99.6%

      \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{b \cdot a} \cdot \frac{0.5}{b + a}} \]
    7. Final simplification99.6%

      \[\leadsto \frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{a \cdot b} \]
    8. Add Preprocessing

    Alternative 5: 69.6% accurate, 2.2× speedup?

    \[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;a \leq -2.9 \cdot 10^{-30}:\\ \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(b \cdot b\right) \cdot a} \cdot 0.5\\ \end{array} \end{array} \]
    (FPCore (a b)
     :precision binary64
     (if (<= a -2.9e-30)
       (* (/ (PI) (* (* a b) a)) 0.5)
       (* (/ (PI) (* (* b b) a)) 0.5)))
    \begin{array}{l}
    
    \\
    \begin{array}{l}
    \mathbf{if}\;a \leq -2.9 \cdot 10^{-30}:\\
    \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5\\
    
    \mathbf{else}:\\
    \;\;\;\;\frac{\mathsf{PI}\left(\right)}{\left(b \cdot b\right) \cdot a} \cdot 0.5\\
    
    
    \end{array}
    \end{array}
    
    Derivation
    1. Split input into 2 regimes
    2. if a < -2.89999999999999989e-30

      1. Initial program 76.3%

        \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
      2. Add Preprocessing
      3. Taylor expanded in a around inf

        \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \]
      4. Step-by-step derivation
        1. *-commutativeN/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
        2. lower-*.f64N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
        3. lower-/.f64N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
        4. lower-PI.f64N/A

          \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{{a}^{2} \cdot b} \cdot \frac{1}{2} \]
        5. lower-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
        6. unpow2N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot \frac{1}{2} \]
        7. lower-*.f6479.2

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot 0.5 \]
      5. Applied rewrites79.2%

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(a \cdot a\right) \cdot b} \cdot 0.5} \]
      6. Step-by-step derivation
        1. Applied rewrites94.3%

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5 \]

        if -2.89999999999999989e-30 < a

        1. Initial program 76.7%

          \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
        2. Add Preprocessing
        3. Taylor expanded in a around 0

          \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{a \cdot {b}^{2}}} \]
        4. Step-by-step derivation
          1. *-commutativeN/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot {b}^{2}} \cdot \frac{1}{2}} \]
          2. lower-*.f64N/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot {b}^{2}} \cdot \frac{1}{2}} \]
          3. lower-/.f64N/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot {b}^{2}}} \cdot \frac{1}{2} \]
          4. lower-PI.f64N/A

            \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{a \cdot {b}^{2}} \cdot \frac{1}{2} \]
          5. *-commutativeN/A

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{b}^{2} \cdot a}} \cdot \frac{1}{2} \]
          6. lower-*.f64N/A

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{b}^{2} \cdot a}} \cdot \frac{1}{2} \]
          7. unpow2N/A

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(b \cdot b\right)} \cdot a} \cdot \frac{1}{2} \]
          8. lower-*.f6460.9

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(b \cdot b\right)} \cdot a} \cdot 0.5 \]
        5. Applied rewrites60.9%

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(b \cdot b\right) \cdot a} \cdot 0.5} \]
      7. Recombined 2 regimes into one program.
      8. Add Preprocessing

      Alternative 6: 98.9% accurate, 2.4× speedup?

      \[\begin{array}{l} \\ \frac{\mathsf{PI}\left(\right)}{\left(2 \cdot \left(a + b\right)\right) \cdot \left(a \cdot b\right)} \end{array} \]
      (FPCore (a b) :precision binary64 (/ (PI) (* (* 2.0 (+ a b)) (* a b))))
      \begin{array}{l}
      
      \\
      \frac{\mathsf{PI}\left(\right)}{\left(2 \cdot \left(a + b\right)\right) \cdot \left(a \cdot b\right)}
      \end{array}
      
      Derivation
      1. Initial program 76.6%

        \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
      2. Add Preprocessing
      3. Applied rewrites88.7%

        \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{\mathsf{PI}\left(\right)}{a + b}, \frac{0.5}{b - a} \cdot {a}^{-1}, \frac{\frac{0.5}{a + b} \cdot \frac{\mathsf{PI}\left(\right)}{b - a}}{-b}\right)} \]
      4. Applied rewrites99.7%

        \[\leadsto \color{blue}{\frac{0.5 \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}{a + b}} \]
      5. Step-by-step derivation
        1. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{1}{2} \cdot \left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)}}{a + b} \]
        2. *-commutativeN/A

          \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right) \cdot \frac{1}{2}}}{a + b} \]
        3. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\left(\frac{\mathsf{PI}\left(\right)}{a \cdot b} \cdot 1\right)} \cdot \frac{1}{2}}{a + b} \]
        4. *-rgt-identityN/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
        5. lift-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{a \cdot b}} \cdot \frac{1}{2}}{a + b} \]
        6. associate-*l/N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{a \cdot b}}}{a + b} \]
        7. metadata-evalN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
        8. div-invN/A

          \[\leadsto \frac{\frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}}{a \cdot b}}{a + b} \]
        9. lift-PI.f64N/A

          \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
        10. lower-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\frac{\mathsf{PI}\left(\right)}{2}}{a \cdot b}}}{a + b} \]
        11. lift-PI.f64N/A

          \[\leadsto \frac{\frac{\frac{\color{blue}{\mathsf{PI}\left(\right)}}{2}}{a \cdot b}}{a + b} \]
        12. div-invN/A

          \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{a \cdot b}}{a + b} \]
        13. metadata-evalN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{a \cdot b}}{a + b} \]
        14. lower-*.f6499.7

          \[\leadsto \frac{\frac{\color{blue}{\mathsf{PI}\left(\right) \cdot 0.5}}{a \cdot b}}{a + b} \]
        15. lift-*.f64N/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{a \cdot b}}}{a + b} \]
        16. *-commutativeN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{b \cdot a}}}{a + b} \]
        17. lower-*.f6499.7

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{\color{blue}{b \cdot a}}}{a + b} \]
      6. Applied rewrites99.7%

        \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot 0.5}{b \cdot a}}}{a + b} \]
      7. Step-by-step derivation
        1. lift-/.f64N/A

          \[\leadsto \color{blue}{\frac{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}{a + b}} \]
        2. lift-/.f64N/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{b \cdot a}}}{a + b} \]
        3. associate-/l/N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\left(a + b\right) \cdot \left(b \cdot a\right)}} \]
        4. lift-*.f64N/A

          \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}}{\left(a + b\right) \cdot \left(b \cdot a\right)} \]
        5. lift-+.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(a + b\right)} \cdot \left(b \cdot a\right)} \]
        6. +-commutativeN/A

          \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
        7. lift-+.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \frac{1}{2}}{\color{blue}{\left(b + a\right)} \cdot \left(b \cdot a\right)} \]
        8. metadata-evalN/A

          \[\leadsto \frac{\mathsf{PI}\left(\right) \cdot \color{blue}{\frac{1}{2}}}{\left(b + a\right) \cdot \left(b \cdot a\right)} \]
        9. div-invN/A

          \[\leadsto \frac{\color{blue}{\frac{\mathsf{PI}\left(\right)}{2}}}{\left(b + a\right) \cdot \left(b \cdot a\right)} \]
        10. *-commutativeN/A

          \[\leadsto \frac{\frac{\mathsf{PI}\left(\right)}{2}}{\color{blue}{\left(b \cdot a\right) \cdot \left(b + a\right)}} \]
        11. associate-/r*N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{2 \cdot \left(\left(b \cdot a\right) \cdot \left(b + a\right)\right)}} \]
        12. associate-*l*N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(2 \cdot \left(b \cdot a\right)\right) \cdot \left(b + a\right)}} \]
        13. lift-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(2 \cdot \left(b \cdot a\right)\right)} \cdot \left(b + a\right)} \]
        14. lift-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(2 \cdot \left(b \cdot a\right)\right) \cdot \left(b + a\right)}} \]
        15. lower-/.f6499.4

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(2 \cdot \left(b \cdot a\right)\right) \cdot \left(b + a\right)}} \]
        16. lift-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(2 \cdot \left(b \cdot a\right)\right) \cdot \left(b + a\right)}} \]
        17. *-commutativeN/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(b + a\right) \cdot \left(2 \cdot \left(b \cdot a\right)\right)}} \]
        18. lift-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\left(b + a\right) \cdot \color{blue}{\left(2 \cdot \left(b \cdot a\right)\right)}} \]
        19. associate-*r*N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(\left(b + a\right) \cdot 2\right) \cdot \left(b \cdot a\right)}} \]
        20. lower-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(\left(b + a\right) \cdot 2\right) \cdot \left(b \cdot a\right)}} \]
      8. Applied rewrites99.4%

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(\left(a + b\right) \cdot 2\right) \cdot \left(a \cdot b\right)}} \]
      9. Final simplification99.4%

        \[\leadsto \frac{\mathsf{PI}\left(\right)}{\left(2 \cdot \left(a + b\right)\right) \cdot \left(a \cdot b\right)} \]
      10. Add Preprocessing

      Alternative 7: 62.6% accurate, 2.6× speedup?

      \[\begin{array}{l} \\ \frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5 \end{array} \]
      (FPCore (a b) :precision binary64 (* (/ (PI) (* (* a b) a)) 0.5))
      \begin{array}{l}
      
      \\
      \frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5
      \end{array}
      
      Derivation
      1. Initial program 76.6%

        \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
      2. Add Preprocessing
      3. Taylor expanded in a around inf

        \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \]
      4. Step-by-step derivation
        1. *-commutativeN/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
        2. lower-*.f64N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
        3. lower-/.f64N/A

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
        4. lower-PI.f64N/A

          \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{{a}^{2} \cdot b} \cdot \frac{1}{2} \]
        5. lower-*.f64N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
        6. unpow2N/A

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot \frac{1}{2} \]
        7. lower-*.f6458.8

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot 0.5 \]
      5. Applied rewrites58.8%

        \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(a \cdot a\right) \cdot b} \cdot 0.5} \]
      6. Step-by-step derivation
        1. Applied rewrites65.7%

          \[\leadsto \frac{\mathsf{PI}\left(\right)}{\left(a \cdot b\right) \cdot a} \cdot 0.5 \]
        2. Add Preprocessing

        Alternative 8: 62.5% accurate, 2.6× speedup?

        \[\begin{array}{l} \\ \frac{0.5}{\left(a \cdot b\right) \cdot a} \cdot \mathsf{PI}\left(\right) \end{array} \]
        (FPCore (a b) :precision binary64 (* (/ 0.5 (* (* a b) a)) (PI)))
        \begin{array}{l}
        
        \\
        \frac{0.5}{\left(a \cdot b\right) \cdot a} \cdot \mathsf{PI}\left(\right)
        \end{array}
        
        Derivation
        1. Initial program 76.6%

          \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
        2. Add Preprocessing
        3. Taylor expanded in a around inf

          \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \]
        4. Step-by-step derivation
          1. *-commutativeN/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
          2. lower-*.f64N/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
          3. lower-/.f64N/A

            \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
          4. lower-PI.f64N/A

            \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{{a}^{2} \cdot b} \cdot \frac{1}{2} \]
          5. lower-*.f64N/A

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
          6. unpow2N/A

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot \frac{1}{2} \]
          7. lower-*.f6458.8

            \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot 0.5 \]
        5. Applied rewrites58.8%

          \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(a \cdot a\right) \cdot b} \cdot 0.5} \]
        6. Step-by-step derivation
          1. Applied rewrites58.8%

            \[\leadsto \mathsf{PI}\left(\right) \cdot \color{blue}{\frac{0.5}{\left(a \cdot a\right) \cdot b}} \]
          2. Step-by-step derivation
            1. Applied rewrites65.7%

              \[\leadsto \mathsf{PI}\left(\right) \cdot \frac{0.5}{\left(b \cdot a\right) \cdot \color{blue}{a}} \]
            2. Final simplification65.7%

              \[\leadsto \frac{0.5}{\left(a \cdot b\right) \cdot a} \cdot \mathsf{PI}\left(\right) \]
            3. Add Preprocessing

            Alternative 9: 56.8% accurate, 2.6× speedup?

            \[\begin{array}{l} \\ \frac{0.5}{\left(a \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right) \end{array} \]
            (FPCore (a b) :precision binary64 (* (/ 0.5 (* (* a a) b)) (PI)))
            \begin{array}{l}
            
            \\
            \frac{0.5}{\left(a \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right)
            \end{array}
            
            Derivation
            1. Initial program 76.6%

              \[\left(\frac{\mathsf{PI}\left(\right)}{2} \cdot \frac{1}{b \cdot b - a \cdot a}\right) \cdot \left(\frac{1}{a} - \frac{1}{b}\right) \]
            2. Add Preprocessing
            3. Taylor expanded in a around inf

              \[\leadsto \color{blue}{\frac{1}{2} \cdot \frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \]
            4. Step-by-step derivation
              1. *-commutativeN/A

                \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
              2. lower-*.f64N/A

                \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b} \cdot \frac{1}{2}} \]
              3. lower-/.f64N/A

                \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
              4. lower-PI.f64N/A

                \[\leadsto \frac{\color{blue}{\mathsf{PI}\left(\right)}}{{a}^{2} \cdot b} \cdot \frac{1}{2} \]
              5. lower-*.f64N/A

                \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{{a}^{2} \cdot b}} \cdot \frac{1}{2} \]
              6. unpow2N/A

                \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot \frac{1}{2} \]
              7. lower-*.f6458.8

                \[\leadsto \frac{\mathsf{PI}\left(\right)}{\color{blue}{\left(a \cdot a\right)} \cdot b} \cdot 0.5 \]
            5. Applied rewrites58.8%

              \[\leadsto \color{blue}{\frac{\mathsf{PI}\left(\right)}{\left(a \cdot a\right) \cdot b} \cdot 0.5} \]
            6. Step-by-step derivation
              1. Applied rewrites58.8%

                \[\leadsto \mathsf{PI}\left(\right) \cdot \color{blue}{\frac{0.5}{\left(a \cdot a\right) \cdot b}} \]
              2. Final simplification58.8%

                \[\leadsto \frac{0.5}{\left(a \cdot a\right) \cdot b} \cdot \mathsf{PI}\left(\right) \]
              3. Add Preprocessing

              Reproduce

              ?
              herbie shell --seed 2024295 
              (FPCore (a b)
                :name "NMSE Section 6.1 mentioned, B"
                :precision binary64
                (* (* (/ (PI) 2.0) (/ 1.0 (- (* b b) (* a a)))) (- (/ 1.0 a) (/ 1.0 b))))