Graphics.Rendering.Plot.Render.Plot.Legend:renderLegendInside from plot-0.2.3.4

Percentage Accurate: 99.9% → 99.9%

Time: 3.5s

Alternatives: 8

Speedup: 1.3×

Specification

Math

\[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

FPCore

(FPCore (x y z) :precision binary64 (+ (+ (+ (+ (+ x y) y) x) z) x))

C

double code(double x, double y, double z) {
	return ((((x + y) + y) + x) + z) + x;
}

Fortran

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(8) function code(x, y, z)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    code = ((((x + y) + y) + x) + z) + x
end function

Java

public static double code(double x, double y, double z) {
	return ((((x + y) + y) + x) + z) + x;
}

Python

def code(x, y, z):
	return ((((x + y) + y) + x) + z) + x

Julia

function code(x, y, z)
	return Float64(Float64(Float64(Float64(Float64(x + y) + y) + x) + z) + x)
end

MATLAB

function tmp = code(x, y, z)
	tmp = ((((x + y) + y) + x) + z) + x;
end

Wolfram

code[x_, y_, z_] := N[(N[(N[(N[(N[(x + y), $MachinePrecision] + y), $MachinePrecision] + x), $MachinePrecision] + z), $MachinePrecision] + x), $MachinePrecision]

Initial Program: 99.9% accurate, 1.0× speedup

Math

\[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

FPCore

(FPCore (x y z) :precision binary64 (+ (+ (+ (+ (+ x y) y) x) z) x))

C

double code(double x, double y, double z) {
	return ((((x + y) + y) + x) + z) + x;
}

Fortran

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(8) function code(x, y, z)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    code = ((((x + y) + y) + x) + z) + x
end function

Java

public static double code(double x, double y, double z) {
	return ((((x + y) + y) + x) + z) + x;
}

Python

def code(x, y, z):
	return ((((x + y) + y) + x) + z) + x

Julia

function code(x, y, z)
	return Float64(Float64(Float64(Float64(Float64(x + y) + y) + x) + z) + x)
end

MATLAB

function tmp = code(x, y, z)
	tmp = ((((x + y) + y) + x) + z) + x;
end

Wolfram

code[x_, y_, z_] := N[(N[(N[(N[(N[(x + y), $MachinePrecision] + y), $MachinePrecision] + x), $MachinePrecision] + z), $MachinePrecision] + x), $MachinePrecision]

Alternative 1: 99.9% accurate, 1.3× speedup

Math

\[\mathsf{fma}\left(2, y + x, x\right) + z \]

FPCore

(FPCore (x y z) :precision binary64 (+ (fma 2.0 (+ y x) x) z))

C

double code(double x, double y, double z) {
	return fma(2.0, (y + x), x) + z;
}

Julia

function code(x, y, z)
	return Float64(fma(2.0, Float64(y + x), x) + z)
end

Wolfram

code[x_, y_, z_] := N[(N[(2.0 * N[(y + x), $MachinePrecision] + x), $MachinePrecision] + z), $MachinePrecision]

Derivation

1. Initial program 99.9%

   \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]
2. Step-by-step derivation

   1. lift-+.f64 N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x} \]

   2. lift-+.f64 N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right)} + x \]

   3. associate-+l+ N/A

      \[\leadsto \color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right) + \left(z + x\right)} \]

   4. +-commutative N/A

      \[\leadsto \left(\left(\left(x + y\right) + y\right) + x\right) + \color{blue}{\left(x + z\right)} \]

   5. associate-+r+ N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right) + z} \]

   6. add-flip-rev N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right)} + z \]

   7. lower-+.f64 N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right) + z} \]

   8. add-flip-rev N/A

      \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right)} + z \]

   9. lift-+.f64 N/A

      \[\leadsto \left(\color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right)} + x\right) + z \]

   10. lift-+.f64 N/A

       \[\leadsto \left(\left(\color{blue}{\left(\left(x + y\right) + y\right)} + x\right) + x\right) + z \]

   11. associate-+l+ N/A

       \[\leadsto \left(\color{blue}{\left(\left(x + y\right) + \left(y + x\right)\right)} + x\right) + z \]

   12. +-commutative N/A

       \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

   13. lift-+.f64 N/A

       \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

   14. count-2 N/A

       \[\leadsto \left(\color{blue}{2 \cdot \left(x + y\right)} + x\right) + z \]

   15. lower-fma.f64 99.9

       \[\leadsto \color{blue}{\mathsf{fma}\left(2, x + y, x\right)} + z \]

   16. lift-+.f64 N/A

       \[\leadsto \mathsf{fma}\left(2, \color{blue}{x + y}, x\right) + z \]

   17. +-commutative N/A

       \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

   18. lower-+.f64 99.9

       \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

3. Applied rewrites 99.9%

   \[\leadsto \color{blue}{\mathsf{fma}\left(2, y + x, x\right) + z} \]
4. Add Preprocessing

Alternative 2: 86.7% accurate, 0.9× speedup

Math

\[\begin{array}{l} t_0 := \mathsf{fma}\left(2, y, x\right) + z\\ \mathbf{if}\;z \leq -13500:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;z \leq 1.55 \cdot 10^{-12}:\\ \;\;\;\;\mathsf{fma}\left(3, x, y \cdot 2\right)\\ \mathbf{else}:\\ \;\;\;\;t\_0\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (+ (fma 2.0 y x) z)))
   (if (<= z -13500.0) t_0 (if (<= z 1.55e-12) (fma 3.0 x (* y 2.0)) t_0))))

C

double code(double x, double y, double z) {
	double t_0 = fma(2.0, y, x) + z;
	double tmp;
	if (z <= -13500.0) {
		tmp = t_0;
	} else if (z <= 1.55e-12) {
		tmp = fma(3.0, x, (y * 2.0));
	} else {
		tmp = t_0;
	}
	return tmp;
}

Julia

function code(x, y, z)
	t_0 = Float64(fma(2.0, y, x) + z)
	tmp = 0.0
	if (z <= -13500.0)
		tmp = t_0;
	elseif (z <= 1.55e-12)
		tmp = fma(3.0, x, Float64(y * 2.0));
	else
		tmp = t_0;
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(2.0 * y + x), $MachinePrecision] + z), $MachinePrecision]}, If[LessEqual[z, -13500.0], t$95$0, If[LessEqual[z, 1.55e-12], N[(3.0 * x + N[(y * 2.0), $MachinePrecision]), $MachinePrecision], t$95$0]]]

Derivation

1. Split input into 2 regimes

2. if z < -13500 or 1.5500000000000001e-12 < z

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]
   2. Step-by-step derivation

      1. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x} \]

      2. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right)} + x \]

      3. associate-+l+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right) + \left(z + x\right)} \]

      4. +-commutative N/A

         \[\leadsto \left(\left(\left(x + y\right) + y\right) + x\right) + \color{blue}{\left(x + z\right)} \]

      5. associate-+r+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right) + z} \]

      6. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right)} + z \]

      7. lower-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right) + z} \]

      8. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right)} + z \]

      9. lift-+.f64 N/A

         \[\leadsto \left(\color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right)} + x\right) + z \]

      10. lift-+.f64 N/A

          \[\leadsto \left(\left(\color{blue}{\left(\left(x + y\right) + y\right)} + x\right) + x\right) + z \]

      11. associate-+l+ N/A

          \[\leadsto \left(\color{blue}{\left(\left(x + y\right) + \left(y + x\right)\right)} + x\right) + z \]

      12. +-commutative N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      13. lift-+.f64 N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      14. count-2 N/A

          \[\leadsto \left(\color{blue}{2 \cdot \left(x + y\right)} + x\right) + z \]

      15. lower-fma.f64 99.9

          \[\leadsto \color{blue}{\mathsf{fma}\left(2, x + y, x\right)} + z \]

      16. lift-+.f64 N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{x + y}, x\right) + z \]

      17. +-commutative N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

      18. lower-+.f64 99.9

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

   3. Applied rewrites 99.9%

      \[\leadsto \color{blue}{\mathsf{fma}\left(2, y + x, x\right) + z} \]

   4. Taylor expanded in x around 0

      \[\leadsto \mathsf{fma}\left(2, \color{blue}{y}, x\right) + z \]
   5. Step-by-step derivation

   6. Applied rewrites 71.3%

      \[\leadsto \mathsf{fma}\left(2, \color{blue}{y}, x\right) + z \]

   if -13500 < z < 1.5500000000000001e-12

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in x around 0

      \[\leadsto 3 \cdot x + \color{blue}{y \cdot \left(2 + \frac{z}{y}\right)} \]
   6. Step-by-step derivation

      1. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      2. lower-*.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      3. lower-+.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 88.6

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   7. Applied rewrites 88.6%

      \[\leadsto \mathsf{fma}\left(3, \color{blue}{x}, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   8. Taylor expanded in y around inf

      \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]
   9. Step-by-step derivation

   10. Applied rewrites 66.5%

       \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 3: 85.6% accurate, 0.9× speedup

Math

\[\begin{array}{l} \mathbf{if}\;x \leq -1.06 \cdot 10^{+29}:\\ \;\;\;\;\mathsf{fma}\left(3, x, z\right)\\ \mathbf{elif}\;x \leq 1.15 \cdot 10^{+165}:\\ \;\;\;\;\mathsf{fma}\left(2, y, x\right) + z\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(3, x, z\right)\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (if (<= x -1.06e+29)
   (fma 3.0 x z)
   (if (<= x 1.15e+165) (+ (fma 2.0 y x) z) (fma 3.0 x z))))

C

double code(double x, double y, double z) {
	double tmp;
	if (x <= -1.06e+29) {
		tmp = fma(3.0, x, z);
	} else if (x <= 1.15e+165) {
		tmp = fma(2.0, y, x) + z;
	} else {
		tmp = fma(3.0, x, z);
	}
	return tmp;
}

Julia

function code(x, y, z)
	tmp = 0.0
	if (x <= -1.06e+29)
		tmp = fma(3.0, x, z);
	elseif (x <= 1.15e+165)
		tmp = Float64(fma(2.0, y, x) + z);
	else
		tmp = fma(3.0, x, z);
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := If[LessEqual[x, -1.06e+29], N[(3.0 * x + z), $MachinePrecision], If[LessEqual[x, 1.15e+165], N[(N[(2.0 * y + x), $MachinePrecision] + z), $MachinePrecision], N[(3.0 * x + z), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if x < -1.0600000000000001e29 or 1.15000000000000008e165 < x

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in x around 0

      \[\leadsto 3 \cdot x + \color{blue}{y \cdot \left(2 + \frac{z}{y}\right)} \]
   6. Step-by-step derivation

      1. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      2. lower-*.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      3. lower-+.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 88.6

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   7. Applied rewrites 88.6%

      \[\leadsto \mathsf{fma}\left(3, \color{blue}{x}, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   8. Taylor expanded in y around inf

      \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]
   9. Step-by-step derivation

   10. Applied rewrites 66.5%

       \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]

   11. Taylor expanded in y around 0

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]
   12. Step-by-step derivation

   13. Applied rewrites 67.0%

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]

   if -1.0600000000000001e29 < x < 1.15000000000000008e165

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]
   2. Step-by-step derivation

      1. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x} \]

      2. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right)} + x \]

      3. associate-+l+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right) + \left(z + x\right)} \]

      4. +-commutative N/A

         \[\leadsto \left(\left(\left(x + y\right) + y\right) + x\right) + \color{blue}{\left(x + z\right)} \]

      5. associate-+r+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right) + z} \]

      6. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right)} + z \]

      7. lower-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right) + z} \]

      8. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right)} + z \]

      9. lift-+.f64 N/A

         \[\leadsto \left(\color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right)} + x\right) + z \]

      10. lift-+.f64 N/A

          \[\leadsto \left(\left(\color{blue}{\left(\left(x + y\right) + y\right)} + x\right) + x\right) + z \]

      11. associate-+l+ N/A

          \[\leadsto \left(\color{blue}{\left(\left(x + y\right) + \left(y + x\right)\right)} + x\right) + z \]

      12. +-commutative N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      13. lift-+.f64 N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      14. count-2 N/A

          \[\leadsto \left(\color{blue}{2 \cdot \left(x + y\right)} + x\right) + z \]

      15. lower-fma.f64 99.9

          \[\leadsto \color{blue}{\mathsf{fma}\left(2, x + y, x\right)} + z \]

      16. lift-+.f64 N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{x + y}, x\right) + z \]

      17. +-commutative N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

      18. lower-+.f64 99.9

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

   3. Applied rewrites 99.9%

      \[\leadsto \color{blue}{\mathsf{fma}\left(2, y + x, x\right) + z} \]

   4. Taylor expanded in x around 0

      \[\leadsto \mathsf{fma}\left(2, \color{blue}{y}, x\right) + z \]
   5. Step-by-step derivation

   6. Applied rewrites 71.3%

      \[\leadsto \mathsf{fma}\left(2, \color{blue}{y}, x\right) + z \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 4: 84.2% accurate, 1.0× speedup

Math

\[\begin{array}{l} \mathbf{if}\;x \leq -1.06 \cdot 10^{+29}:\\ \;\;\;\;\mathsf{fma}\left(3, x, z\right)\\ \mathbf{elif}\;x \leq 1.15 \cdot 10^{+165}:\\ \;\;\;\;\left(y + y\right) + z\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(3, x, z\right)\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (if (<= x -1.06e+29)
   (fma 3.0 x z)
   (if (<= x 1.15e+165) (+ (+ y y) z) (fma 3.0 x z))))

C

double code(double x, double y, double z) {
	double tmp;
	if (x <= -1.06e+29) {
		tmp = fma(3.0, x, z);
	} else if (x <= 1.15e+165) {
		tmp = (y + y) + z;
	} else {
		tmp = fma(3.0, x, z);
	}
	return tmp;
}

Julia

function code(x, y, z)
	tmp = 0.0
	if (x <= -1.06e+29)
		tmp = fma(3.0, x, z);
	elseif (x <= 1.15e+165)
		tmp = Float64(Float64(y + y) + z);
	else
		tmp = fma(3.0, x, z);
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := If[LessEqual[x, -1.06e+29], N[(3.0 * x + z), $MachinePrecision], If[LessEqual[x, 1.15e+165], N[(N[(y + y), $MachinePrecision] + z), $MachinePrecision], N[(3.0 * x + z), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if x < -1.0600000000000001e29 or 1.15000000000000008e165 < x

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in x around 0

      \[\leadsto 3 \cdot x + \color{blue}{y \cdot \left(2 + \frac{z}{y}\right)} \]
   6. Step-by-step derivation

      1. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      2. lower-*.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      3. lower-+.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 88.6

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   7. Applied rewrites 88.6%

      \[\leadsto \mathsf{fma}\left(3, \color{blue}{x}, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   8. Taylor expanded in y around inf

      \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]
   9. Step-by-step derivation

   10. Applied rewrites 66.5%

       \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]

   11. Taylor expanded in y around 0

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]
   12. Step-by-step derivation

   13. Applied rewrites 67.0%

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]

   if -1.0600000000000001e29 < x < 1.15000000000000008e165

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]
   2. Step-by-step derivation

      1. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x} \]

      2. lift-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right)} + x \]

      3. associate-+l+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right) + \left(z + x\right)} \]

      4. +-commutative N/A

         \[\leadsto \left(\left(\left(x + y\right) + y\right) + x\right) + \color{blue}{\left(x + z\right)} \]

      5. associate-+r+ N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right) + z} \]

      6. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right)} + z \]

      7. lower-+.f64 N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) - \left(\mathsf{neg}\left(x\right)\right)\right) + z} \]

      8. add-flip-rev N/A

         \[\leadsto \color{blue}{\left(\left(\left(\left(x + y\right) + y\right) + x\right) + x\right)} + z \]

      9. lift-+.f64 N/A

         \[\leadsto \left(\color{blue}{\left(\left(\left(x + y\right) + y\right) + x\right)} + x\right) + z \]

      10. lift-+.f64 N/A

          \[\leadsto \left(\left(\color{blue}{\left(\left(x + y\right) + y\right)} + x\right) + x\right) + z \]

      11. associate-+l+ N/A

          \[\leadsto \left(\color{blue}{\left(\left(x + y\right) + \left(y + x\right)\right)} + x\right) + z \]

      12. +-commutative N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      13. lift-+.f64 N/A

          \[\leadsto \left(\left(\left(x + y\right) + \color{blue}{\left(x + y\right)}\right) + x\right) + z \]

      14. count-2 N/A

          \[\leadsto \left(\color{blue}{2 \cdot \left(x + y\right)} + x\right) + z \]

      15. lower-fma.f64 99.9

          \[\leadsto \color{blue}{\mathsf{fma}\left(2, x + y, x\right)} + z \]

      16. lift-+.f64 N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{x + y}, x\right) + z \]

      17. +-commutative N/A

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

      18. lower-+.f64 99.9

          \[\leadsto \mathsf{fma}\left(2, \color{blue}{y + x}, x\right) + z \]

   3. Applied rewrites 99.9%

      \[\leadsto \color{blue}{\mathsf{fma}\left(2, y + x, x\right) + z} \]

   4. Taylor expanded in x around 0

      \[\leadsto \color{blue}{2 \cdot y} + z \]
   5. Step-by-step derivation

      1. lower-*.f64 66.5

         \[\leadsto 2 \cdot \color{blue}{y} + z \]

   6. Applied rewrites 66.5%

      \[\leadsto \color{blue}{2 \cdot y} + z \]
   7. Step-by-step derivation

      1. lift-*.f64 N/A

         \[\leadsto 2 \cdot \color{blue}{y} + z \]

      2. count-2-rev N/A

         \[\leadsto \left(y + \color{blue}{y}\right) + z \]

      3. lower-+.f64 66.5

         \[\leadsto \left(y + \color{blue}{y}\right) + z \]

   8. Applied rewrites 66.5%

      \[\leadsto \color{blue}{\left(y + y\right) + z} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 5: 79.7% accurate, 1.1× speedup

Math

\[\begin{array}{l} \mathbf{if}\;y \leq -5.2 \cdot 10^{+116}:\\ \;\;\;\;y \cdot 2\\ \mathbf{elif}\;y \leq 4.6 \cdot 10^{+157}:\\ \;\;\;\;\mathsf{fma}\left(3, x, z\right)\\ \mathbf{else}:\\ \;\;\;\;y \cdot 2\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (if (<= y -5.2e+116) (* y 2.0) (if (<= y 4.6e+157) (fma 3.0 x z) (* y 2.0))))

C

double code(double x, double y, double z) {
	double tmp;
	if (y <= -5.2e+116) {
		tmp = y * 2.0;
	} else if (y <= 4.6e+157) {
		tmp = fma(3.0, x, z);
	} else {
		tmp = y * 2.0;
	}
	return tmp;
}

Julia

function code(x, y, z)
	tmp = 0.0
	if (y <= -5.2e+116)
		tmp = Float64(y * 2.0);
	elseif (y <= 4.6e+157)
		tmp = fma(3.0, x, z);
	else
		tmp = Float64(y * 2.0);
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := If[LessEqual[y, -5.2e+116], N[(y * 2.0), $MachinePrecision], If[LessEqual[y, 4.6e+157], N[(3.0 * x + z), $MachinePrecision], N[(y * 2.0), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if y < -5.19999999999999973e116 or 4.60000000000000008e157 < y

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in y around inf

      \[\leadsto y \cdot 2 \]
   6. Step-by-step derivation

   7. Applied rewrites 33.9%

      \[\leadsto y \cdot 2 \]

   if -5.19999999999999973e116 < y < 4.60000000000000008e157

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in x around 0

      \[\leadsto 3 \cdot x + \color{blue}{y \cdot \left(2 + \frac{z}{y}\right)} \]
   6. Step-by-step derivation

      1. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      2. lower-*.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      3. lower-+.f64 N/A

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 88.6

         \[\leadsto \mathsf{fma}\left(3, x, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   7. Applied rewrites 88.6%

      \[\leadsto \mathsf{fma}\left(3, \color{blue}{x}, y \cdot \left(2 + \frac{z}{y}\right)\right) \]

   8. Taylor expanded in y around inf

      \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]
   9. Step-by-step derivation

   10. Applied rewrites 66.5%

       \[\leadsto \mathsf{fma}\left(3, x, y \cdot 2\right) \]

   11. Taylor expanded in y around 0

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]
   12. Step-by-step derivation

   13. Applied rewrites 67.0%

       \[\leadsto \mathsf{fma}\left(3, x, z\right) \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 6: 54.9% accurate, 1.3× speedup

Math

\[\begin{array}{l} \mathbf{if}\;y \leq -6.2 \cdot 10^{+57}:\\ \;\;\;\;y \cdot 2\\ \mathbf{elif}\;y \leq 4.6 \cdot 10^{+157}:\\ \;\;\;\;z + x\\ \mathbf{else}:\\ \;\;\;\;y \cdot 2\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (if (<= y -6.2e+57) (* y 2.0) (if (<= y 4.6e+157) (+ z x) (* y 2.0))))

C

double code(double x, double y, double z) {
	double tmp;
	if (y <= -6.2e+57) {
		tmp = y * 2.0;
	} else if (y <= 4.6e+157) {
		tmp = z + x;
	} else {
		tmp = y * 2.0;
	}
	return tmp;
}

Fortran

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(8) function code(x, y, z)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8) :: tmp
    if (y <= (-6.2d+57)) then
        tmp = y * 2.0d0
    else if (y <= 4.6d+157) then
        tmp = z + x
    else
        tmp = y * 2.0d0
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z) {
	double tmp;
	if (y <= -6.2e+57) {
		tmp = y * 2.0;
	} else if (y <= 4.6e+157) {
		tmp = z + x;
	} else {
		tmp = y * 2.0;
	}
	return tmp;
}

Python

def code(x, y, z):
	tmp = 0
	if y <= -6.2e+57:
		tmp = y * 2.0
	elif y <= 4.6e+157:
		tmp = z + x
	else:
		tmp = y * 2.0
	return tmp

Julia

function code(x, y, z)
	tmp = 0.0
	if (y <= -6.2e+57)
		tmp = Float64(y * 2.0);
	elseif (y <= 4.6e+157)
		tmp = Float64(z + x);
	else
		tmp = Float64(y * 2.0);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z)
	tmp = 0.0;
	if (y <= -6.2e+57)
		tmp = y * 2.0;
	elseif (y <= 4.6e+157)
		tmp = z + x;
	else
		tmp = y * 2.0;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_] := If[LessEqual[y, -6.2e+57], N[(y * 2.0), $MachinePrecision], If[LessEqual[y, 4.6e+157], N[(z + x), $MachinePrecision], N[(y * 2.0), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if y < -6.20000000000000026e57 or 4.60000000000000008e157 < y

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around inf

      \[\leadsto \color{blue}{y \cdot \left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto y \cdot \color{blue}{\left(2 + \left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)\right)} \]

      2. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \color{blue}{\left(2 \cdot \frac{x}{y} + \left(\frac{x}{y} + \frac{z}{y}\right)\right)}\right) \]

      3. lower-fma.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \color{blue}{\frac{x}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      4. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{\color{blue}{y}}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      5. lower-+.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      6. lower-/.f64 N/A

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

      7. lower-/.f64 80.3

         \[\leadsto y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right) \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{y \cdot \left(2 + \mathsf{fma}\left(2, \frac{x}{y}, \frac{x}{y} + \frac{z}{y}\right)\right)} \]

   5. Taylor expanded in y around inf

      \[\leadsto y \cdot 2 \]
   6. Step-by-step derivation

   7. Applied rewrites 33.9%

      \[\leadsto y \cdot 2 \]

   if -6.20000000000000026e57 < y < 4.60000000000000008e157

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around 0

      \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]
   3. Step-by-step derivation

      1. lower-+.f64 N/A

         \[\leadsto \left(z + \color{blue}{2 \cdot x}\right) + x \]

      2. lower-*.f64 66.9

         \[\leadsto \left(z + 2 \cdot \color{blue}{x}\right) + x \]

   4. Applied rewrites 66.9%

      \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]

   5. Taylor expanded in x around 0

      \[\leadsto z + x \]
   6. Step-by-step derivation

   7. Applied rewrites 39.4%

      \[\leadsto z + x \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 7: 54.2% accurate, 1.3× speedup

Math

\[\begin{array}{l} \mathbf{if}\;z \leq -13500:\\ \;\;\;\;z + x\\ \mathbf{elif}\;z \leq 1.55 \cdot 10^{-12}:\\ \;\;\;\;3 \cdot x\\ \mathbf{else}:\\ \;\;\;\;z + x\\ \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (if (<= z -13500.0) (+ z x) (if (<= z 1.55e-12) (* 3.0 x) (+ z x))))

C

double code(double x, double y, double z) {
	double tmp;
	if (z <= -13500.0) {
		tmp = z + x;
	} else if (z <= 1.55e-12) {
		tmp = 3.0 * x;
	} else {
		tmp = z + x;
	}
	return tmp;
}

Fortran

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(8) function code(x, y, z)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8) :: tmp
    if (z <= (-13500.0d0)) then
        tmp = z + x
    else if (z <= 1.55d-12) then
        tmp = 3.0d0 * x
    else
        tmp = z + x
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z) {
	double tmp;
	if (z <= -13500.0) {
		tmp = z + x;
	} else if (z <= 1.55e-12) {
		tmp = 3.0 * x;
	} else {
		tmp = z + x;
	}
	return tmp;
}

Python

def code(x, y, z):
	tmp = 0
	if z <= -13500.0:
		tmp = z + x
	elif z <= 1.55e-12:
		tmp = 3.0 * x
	else:
		tmp = z + x
	return tmp

Julia

function code(x, y, z)
	tmp = 0.0
	if (z <= -13500.0)
		tmp = Float64(z + x);
	elseif (z <= 1.55e-12)
		tmp = Float64(3.0 * x);
	else
		tmp = Float64(z + x);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z)
	tmp = 0.0;
	if (z <= -13500.0)
		tmp = z + x;
	elseif (z <= 1.55e-12)
		tmp = 3.0 * x;
	else
		tmp = z + x;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_] := If[LessEqual[z, -13500.0], N[(z + x), $MachinePrecision], If[LessEqual[z, 1.55e-12], N[(3.0 * x), $MachinePrecision], N[(z + x), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if z < -13500 or 1.5500000000000001e-12 < z

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in y around 0

      \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]
   3. Step-by-step derivation

      1. lower-+.f64 N/A

         \[\leadsto \left(z + \color{blue}{2 \cdot x}\right) + x \]

      2. lower-*.f64 66.9

         \[\leadsto \left(z + 2 \cdot \color{blue}{x}\right) + x \]

   4. Applied rewrites 66.9%

      \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]

   5. Taylor expanded in x around 0

      \[\leadsto z + x \]
   6. Step-by-step derivation

   7. Applied rewrites 39.4%

      \[\leadsto z + x \]

   if -13500 < z < 1.5500000000000001e-12

   1. Initial program 99.9%

      \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

   2. Taylor expanded in x around inf

      \[\leadsto \color{blue}{3 \cdot x} \]
   3. Step-by-step derivation

      1. lower-*.f64 34.4

         \[\leadsto 3 \cdot \color{blue}{x} \]

   4. Applied rewrites 34.4%

      \[\leadsto \color{blue}{3 \cdot x} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 8: 39.4% accurate, 3.9× speedup

Math

\[z + x \]

FPCore

(FPCore (x y z) :precision binary64 (+ z x))

C

double code(double x, double y, double z) {
	return z + x;
}

Fortran

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(8) function code(x, y, z)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    code = z + x
end function

Java

public static double code(double x, double y, double z) {
	return z + x;
}

Python

def code(x, y, z):
	return z + x

Julia

function code(x, y, z)
	return Float64(z + x)
end

MATLAB

function tmp = code(x, y, z)
	tmp = z + x;
end

Wolfram

code[x_, y_, z_] := N[(z + x), $MachinePrecision]

Derivation

1. Initial program 99.9%

   \[\left(\left(\left(\left(x + y\right) + y\right) + x\right) + z\right) + x \]

2. Taylor expanded in y around 0

   \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]
3. Step-by-step derivation

   1. lower-+.f64 N/A

      \[\leadsto \left(z + \color{blue}{2 \cdot x}\right) + x \]

   2. lower-*.f64 66.9

      \[\leadsto \left(z + 2 \cdot \color{blue}{x}\right) + x \]

4. Applied rewrites 66.9%

   \[\leadsto \color{blue}{\left(z + 2 \cdot x\right)} + x \]

5. Taylor expanded in x around 0

   \[\leadsto z + x \]
6. Step-by-step derivation

7. Applied rewrites 39.4%

   \[\leadsto z + x \]
8. Add Preprocessing

Reproduce

herbie shell --seed 2025172 
(FPCore (x y z)
  :name "Graphics.Rendering.Plot.Render.Plot.Legend:renderLegendInside from plot-0.2.3.4"
  :precision binary64
  (+ (+ (+ (+ (+ x y) y) x) z) x))