Data.Array.Repa.Algorithms.ColorRamp:rampColorHotToCold from repa-algorithms-3.4.0.1, B

Percentage Accurate: 99.9% → 100.0%

Time: 2.3s

Alternatives: 7

Speedup: 1.3×

• 21.1% of points produce a very large (infinite) output. You may want to add a precondition.

Specification

Math

\[\begin{array}{l} \\ \frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \end{array} \]

FPCore

(FPCore (x y z) :precision binary64 (/ (* 4.0 (- (- x y) (* z 0.5))) z))

C

double code(double x, double y, double z) {
	return (4.0 * ((x - y) - (z * 0.5))) / z;
}

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 = (4.0d0 * ((x - y) - (z * 0.5d0))) / z
end function

Java

public static double code(double x, double y, double z) {
	return (4.0 * ((x - y) - (z * 0.5))) / z;
}

Python

def code(x, y, z):
	return (4.0 * ((x - y) - (z * 0.5))) / z

Julia

function code(x, y, z)
	return Float64(Float64(4.0 * Float64(Float64(x - y) - Float64(z * 0.5))) / z)
end

MATLAB

function tmp = code(x, y, z)
	tmp = (4.0 * ((x - y) - (z * 0.5))) / z;
end

Wolfram

code[x_, y_, z_] := N[(N[(4.0 * N[(N[(x - y), $MachinePrecision] - N[(z * 0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision]

Initial Program: 99.9% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \end{array} \]

FPCore

(FPCore (x y z) :precision binary64 (/ (* 4.0 (- (- x y) (* z 0.5))) z))

C

double code(double x, double y, double z) {
	return (4.0 * ((x - y) - (z * 0.5))) / z;
}

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 = (4.0d0 * ((x - y) - (z * 0.5d0))) / z
end function

Java

public static double code(double x, double y, double z) {
	return (4.0 * ((x - y) - (z * 0.5))) / z;
}

Python

def code(x, y, z):
	return (4.0 * ((x - y) - (z * 0.5))) / z

Julia

function code(x, y, z)
	return Float64(Float64(4.0 * Float64(Float64(x - y) - Float64(z * 0.5))) / z)
end

MATLAB

function tmp = code(x, y, z)
	tmp = (4.0 * ((x - y) - (z * 0.5))) / z;
end

Wolfram

code[x_, y_, z_] := N[(N[(4.0 * N[(N[(x - y), $MachinePrecision] - N[(z * 0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision]

Alternative 1: 100.0% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \end{array} \]

FPCore

(FPCore (x y z) :precision binary64 (fma (/ (- x y) z) 4.0 -2.0))

C

double code(double x, double y, double z) {
	return fma(((x - y) / z), 4.0, -2.0);
}

Julia

function code(x, y, z)
	return fma(Float64(Float64(x - y) / z), 4.0, -2.0)
end

Wolfram

code[x_, y_, z_] := N[(N[(N[(x - y), $MachinePrecision] / z), $MachinePrecision] * 4.0 + -2.0), $MachinePrecision]

Derivation

1. Initial program 99.9%

   \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

2. Taylor expanded in z around inf

   \[\leadsto \color{blue}{4 \cdot \frac{x - y}{z} - 2} \]
3. Step-by-step derivation

   1. sub-flip N/A

      \[\leadsto 4 \cdot \frac{x - y}{z} + \color{blue}{\left(\mathsf{neg}\left(2\right)\right)} \]

   2. *-commutative N/A

      \[\leadsto \frac{x - y}{z} \cdot 4 + \left(\mathsf{neg}\left(\color{blue}{2}\right)\right) \]

   3. metadata-eval N/A

      \[\leadsto \frac{x - y}{z} \cdot 4 + -2 \]

   4. lower-fma.f64 N/A

      \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, \color{blue}{4}, -2\right) \]

   5. lower-/.f64 N/A

      \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

   6. lift--.f64 100.0

      \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

4. Applied rewrites 100.0%

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

Alternative 2: 98.4% accurate, 0.3× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \frac{\left(x - y\right) \cdot 4}{z}\\ t_1 := \frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z}\\ \mathbf{if}\;t\_1 \leq -4 \cdot 10^{+14}:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;t\_1 \leq 5000000000:\\ \;\;\;\;\mathsf{fma}\left(\frac{4}{z}, x, -2\right)\\ \mathbf{else}:\\ \;\;\;\;t\_0\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (/ (* (- x y) 4.0) z)) (t_1 (/ (* 4.0 (- (- x y) (* z 0.5))) z)))
   (if (<= t_1 -4e+14)
     t_0
     (if (<= t_1 5000000000.0) (fma (/ 4.0 z) x -2.0) t_0))))

C

double code(double x, double y, double z) {
	double t_0 = ((x - y) * 4.0) / z;
	double t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	double tmp;
	if (t_1 <= -4e+14) {
		tmp = t_0;
	} else if (t_1 <= 5000000000.0) {
		tmp = fma((4.0 / z), x, -2.0);
	} else {
		tmp = t_0;
	}
	return tmp;
}

Julia

function code(x, y, z)
	t_0 = Float64(Float64(Float64(x - y) * 4.0) / z)
	t_1 = Float64(Float64(4.0 * Float64(Float64(x - y) - Float64(z * 0.5))) / z)
	tmp = 0.0
	if (t_1 <= -4e+14)
		tmp = t_0;
	elseif (t_1 <= 5000000000.0)
		tmp = fma(Float64(4.0 / z), x, -2.0);
	else
		tmp = t_0;
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(N[(x - y), $MachinePrecision] * 4.0), $MachinePrecision] / z), $MachinePrecision]}, Block[{t$95$1 = N[(N[(4.0 * N[(N[(x - y), $MachinePrecision] - N[(z * 0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision]}, If[LessEqual[t$95$1, -4e+14], t$95$0, If[LessEqual[t$95$1, 5000000000.0], N[(N[(4.0 / z), $MachinePrecision] * x + -2.0), $MachinePrecision], t$95$0]]]]

Derivation

1. Split input into 2 regimes

2. if (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < -4e14 or 5e9 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z)

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around 0

      \[\leadsto \frac{\color{blue}{4 \cdot \left(x - y\right)}}{z} \]
   3. Step-by-step derivation

      1. *-commutative N/A

         \[\leadsto \frac{\left(x - y\right) \cdot \color{blue}{4}}{z} \]

      2. lower-*.f64 N/A

         \[\leadsto \frac{\left(x - y\right) \cdot \color{blue}{4}}{z} \]

      3. lift--.f64 67.3

         \[\leadsto \frac{\left(x - y\right) \cdot 4}{z} \]

   4. Applied rewrites 67.3%

      \[\leadsto \frac{\color{blue}{\left(x - y\right) \cdot 4}}{z} \]

   if -4e14 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < 5e9

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{4 \cdot \frac{x - y}{z} - 2} \]
   3. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x - y}{z} + \color{blue}{\left(\mathsf{neg}\left(2\right)\right)} \]

      2. *-commutative N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + \left(\mathsf{neg}\left(\color{blue}{2}\right)\right) \]

      3. metadata-eval N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + -2 \]

      4. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, \color{blue}{4}, -2\right) \]

      5. lower-/.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

      6. lift--.f64 100.0

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

   4. Applied rewrites 100.0%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right)} \]

   5. Taylor expanded in y around 0

      \[\leadsto 4 \cdot \frac{x}{z} - \color{blue}{2} \]
   6. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      2. associate-*r/ N/A

         \[\leadsto \frac{4 \cdot x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      3. associate-*l/ N/A

         \[\leadsto \frac{4}{z} \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      4. mult-flip-rev N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      5. metadata-eval N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + -2 \]

      6. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(4 \cdot \frac{1}{z}, x, -2\right) \]

      7. mult-flip-rev N/A

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

      8. lower-/.f64 67.4

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

   7. Applied rewrites 67.4%

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

Alternative 3: 85.8% accurate, 0.9× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \mathsf{fma}\left(\frac{4}{z}, x, -2\right)\\ \mathbf{if}\;x \leq -1.15 \cdot 10^{+129}:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;x \leq 14000000:\\ \;\;\;\;\mathsf{fma}\left(\frac{y}{z}, -4, -2\right)\\ \mathbf{else}:\\ \;\;\;\;t\_0\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (fma (/ 4.0 z) x -2.0)))
   (if (<= x -1.15e+129)
     t_0
     (if (<= x 14000000.0) (fma (/ y z) -4.0 -2.0) t_0))))

C

double code(double x, double y, double z) {
	double t_0 = fma((4.0 / z), x, -2.0);
	double tmp;
	if (x <= -1.15e+129) {
		tmp = t_0;
	} else if (x <= 14000000.0) {
		tmp = fma((y / z), -4.0, -2.0);
	} else {
		tmp = t_0;
	}
	return tmp;
}

Julia

function code(x, y, z)
	t_0 = fma(Float64(4.0 / z), x, -2.0)
	tmp = 0.0
	if (x <= -1.15e+129)
		tmp = t_0;
	elseif (x <= 14000000.0)
		tmp = fma(Float64(y / z), -4.0, -2.0);
	else
		tmp = t_0;
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(4.0 / z), $MachinePrecision] * x + -2.0), $MachinePrecision]}, If[LessEqual[x, -1.15e+129], t$95$0, If[LessEqual[x, 14000000.0], N[(N[(y / z), $MachinePrecision] * -4.0 + -2.0), $MachinePrecision], t$95$0]]]

Derivation

1. Split input into 2 regimes

2. if x < -1.14999999999999995e129 or 1.4e7 < x

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{4 \cdot \frac{x - y}{z} - 2} \]
   3. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x - y}{z} + \color{blue}{\left(\mathsf{neg}\left(2\right)\right)} \]

      2. *-commutative N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + \left(\mathsf{neg}\left(\color{blue}{2}\right)\right) \]

      3. metadata-eval N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + -2 \]

      4. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, \color{blue}{4}, -2\right) \]

      5. lower-/.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

      6. lift--.f64 100.0

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

   4. Applied rewrites 100.0%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right)} \]

   5. Taylor expanded in y around 0

      \[\leadsto 4 \cdot \frac{x}{z} - \color{blue}{2} \]
   6. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      2. associate-*r/ N/A

         \[\leadsto \frac{4 \cdot x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      3. associate-*l/ N/A

         \[\leadsto \frac{4}{z} \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      4. mult-flip-rev N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      5. metadata-eval N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + -2 \]

      6. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(4 \cdot \frac{1}{z}, x, -2\right) \]

      7. mult-flip-rev N/A

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

      8. lower-/.f64 67.4

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

   7. Applied rewrites 67.4%

      \[\leadsto \mathsf{fma}\left(\frac{4}{z}, \color{blue}{x}, -2\right) \]

   if -1.14999999999999995e129 < x < 1.4e7

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{4 \cdot \frac{x - y}{z} - 2} \]
   3. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x - y}{z} + \color{blue}{\left(\mathsf{neg}\left(2\right)\right)} \]

      2. *-commutative N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + \left(\mathsf{neg}\left(\color{blue}{2}\right)\right) \]

      3. metadata-eval N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + -2 \]

      4. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, \color{blue}{4}, -2\right) \]

      5. lower-/.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

      6. lift--.f64 100.0

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

   4. Applied rewrites 100.0%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right)} \]

   5. Taylor expanded in x around 0

      \[\leadsto -4 \cdot \frac{y}{z} - \color{blue}{2} \]
   6. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto -4 \cdot \frac{y}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      2. *-commutative N/A

         \[\leadsto \frac{y}{z} \cdot -4 + \left(\mathsf{neg}\left(2\right)\right) \]

      3. metadata-eval N/A

         \[\leadsto \frac{y}{z} \cdot -4 + -2 \]

      4. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{y}{z}, -4, -2\right) \]

      5. lower-/.f64 67.5

         \[\leadsto \mathsf{fma}\left(\frac{y}{z}, -4, -2\right) \]

   7. Applied rewrites 67.5%

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

Alternative 4: 80.1% accurate, 0.9× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \frac{-4 \cdot y}{z}\\ \mathbf{if}\;y \leq -1.9 \cdot 10^{+145}:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;y \leq 1.25 \cdot 10^{+95}:\\ \;\;\;\;\mathsf{fma}\left(\frac{4}{z}, x, -2\right)\\ \mathbf{else}:\\ \;\;\;\;t\_0\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (/ (* -4.0 y) z)))
   (if (<= y -1.9e+145) t_0 (if (<= y 1.25e+95) (fma (/ 4.0 z) x -2.0) t_0))))

C

double code(double x, double y, double z) {
	double t_0 = (-4.0 * y) / z;
	double tmp;
	if (y <= -1.9e+145) {
		tmp = t_0;
	} else if (y <= 1.25e+95) {
		tmp = fma((4.0 / z), x, -2.0);
	} else {
		tmp = t_0;
	}
	return tmp;
}

Julia

function code(x, y, z)
	t_0 = Float64(Float64(-4.0 * y) / z)
	tmp = 0.0
	if (y <= -1.9e+145)
		tmp = t_0;
	elseif (y <= 1.25e+95)
		tmp = fma(Float64(4.0 / z), x, -2.0);
	else
		tmp = t_0;
	end
	return tmp
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(-4.0 * y), $MachinePrecision] / z), $MachinePrecision]}, If[LessEqual[y, -1.9e+145], t$95$0, If[LessEqual[y, 1.25e+95], N[(N[(4.0 / z), $MachinePrecision] * x + -2.0), $MachinePrecision], t$95$0]]]

Derivation

1. Split input into 2 regimes

2. if y < -1.90000000000000006e145 or 1.25000000000000006e95 < y

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in y around inf

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]
   3. Step-by-step derivation

      1. lower-*.f64 35.8

         \[\leadsto \frac{-4 \cdot \color{blue}{y}}{z} \]

   4. Applied rewrites 35.8%

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]

   if -1.90000000000000006e145 < y < 1.25000000000000006e95

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{4 \cdot \frac{x - y}{z} - 2} \]
   3. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x - y}{z} + \color{blue}{\left(\mathsf{neg}\left(2\right)\right)} \]

      2. *-commutative N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + \left(\mathsf{neg}\left(\color{blue}{2}\right)\right) \]

      3. metadata-eval N/A

         \[\leadsto \frac{x - y}{z} \cdot 4 + -2 \]

      4. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, \color{blue}{4}, -2\right) \]

      5. lower-/.f64 N/A

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

      6. lift--.f64 100.0

         \[\leadsto \mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right) \]

   4. Applied rewrites 100.0%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{x - y}{z}, 4, -2\right)} \]

   5. Taylor expanded in y around 0

      \[\leadsto 4 \cdot \frac{x}{z} - \color{blue}{2} \]
   6. Step-by-step derivation

      1. sub-flip N/A

         \[\leadsto 4 \cdot \frac{x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      2. associate-*r/ N/A

         \[\leadsto \frac{4 \cdot x}{z} + \left(\mathsf{neg}\left(2\right)\right) \]

      3. associate-*l/ N/A

         \[\leadsto \frac{4}{z} \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      4. mult-flip-rev N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + \left(\mathsf{neg}\left(2\right)\right) \]

      5. metadata-eval N/A

         \[\leadsto \left(4 \cdot \frac{1}{z}\right) \cdot x + -2 \]

      6. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(4 \cdot \frac{1}{z}, x, -2\right) \]

      7. mult-flip-rev N/A

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

      8. lower-/.f64 67.4

         \[\leadsto \mathsf{fma}\left(\frac{4}{z}, x, -2\right) \]

   7. Applied rewrites 67.4%

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

Alternative 5: 66.5% accurate, 0.2× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \frac{x \cdot 4}{z}\\ t_1 := \frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z}\\ \mathbf{if}\;t\_1 \leq -1 \cdot 10^{+20}:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;t\_1 \leq -1:\\ \;\;\;\;-2\\ \mathbf{elif}\;t\_1 \leq 5 \cdot 10^{+129}:\\ \;\;\;\;t\_0\\ \mathbf{else}:\\ \;\;\;\;\frac{-4 \cdot y}{z}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (/ (* x 4.0) z)) (t_1 (/ (* 4.0 (- (- x y) (* z 0.5))) z)))
   (if (<= t_1 -1e+20)
     t_0
     (if (<= t_1 -1.0) -2.0 (if (<= t_1 5e+129) t_0 (/ (* -4.0 y) z))))))

C

double code(double x, double y, double z) {
	double t_0 = (x * 4.0) / z;
	double t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	double tmp;
	if (t_1 <= -1e+20) {
		tmp = t_0;
	} else if (t_1 <= -1.0) {
		tmp = -2.0;
	} else if (t_1 <= 5e+129) {
		tmp = t_0;
	} else {
		tmp = (-4.0 * y) / z;
	}
	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) :: t_0
    real(8) :: t_1
    real(8) :: tmp
    t_0 = (x * 4.0d0) / z
    t_1 = (4.0d0 * ((x - y) - (z * 0.5d0))) / z
    if (t_1 <= (-1d+20)) then
        tmp = t_0
    else if (t_1 <= (-1.0d0)) then
        tmp = -2.0d0
    else if (t_1 <= 5d+129) then
        tmp = t_0
    else
        tmp = ((-4.0d0) * y) / z
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z) {
	double t_0 = (x * 4.0) / z;
	double t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	double tmp;
	if (t_1 <= -1e+20) {
		tmp = t_0;
	} else if (t_1 <= -1.0) {
		tmp = -2.0;
	} else if (t_1 <= 5e+129) {
		tmp = t_0;
	} else {
		tmp = (-4.0 * y) / z;
	}
	return tmp;
}

Python

def code(x, y, z):
	t_0 = (x * 4.0) / z
	t_1 = (4.0 * ((x - y) - (z * 0.5))) / z
	tmp = 0
	if t_1 <= -1e+20:
		tmp = t_0
	elif t_1 <= -1.0:
		tmp = -2.0
	elif t_1 <= 5e+129:
		tmp = t_0
	else:
		tmp = (-4.0 * y) / z
	return tmp

Julia

function code(x, y, z)
	t_0 = Float64(Float64(x * 4.0) / z)
	t_1 = Float64(Float64(4.0 * Float64(Float64(x - y) - Float64(z * 0.5))) / z)
	tmp = 0.0
	if (t_1 <= -1e+20)
		tmp = t_0;
	elseif (t_1 <= -1.0)
		tmp = -2.0;
	elseif (t_1 <= 5e+129)
		tmp = t_0;
	else
		tmp = Float64(Float64(-4.0 * y) / z);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z)
	t_0 = (x * 4.0) / z;
	t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	tmp = 0.0;
	if (t_1 <= -1e+20)
		tmp = t_0;
	elseif (t_1 <= -1.0)
		tmp = -2.0;
	elseif (t_1 <= 5e+129)
		tmp = t_0;
	else
		tmp = (-4.0 * y) / z;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(x * 4.0), $MachinePrecision] / z), $MachinePrecision]}, Block[{t$95$1 = N[(N[(4.0 * N[(N[(x - y), $MachinePrecision] - N[(z * 0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision]}, If[LessEqual[t$95$1, -1e+20], t$95$0, If[LessEqual[t$95$1, -1.0], -2.0, If[LessEqual[t$95$1, 5e+129], t$95$0, N[(N[(-4.0 * y), $MachinePrecision] / z), $MachinePrecision]]]]]]

Derivation

1. Split input into 3 regimes

2. if (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < -1e20 or -1 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < 5.0000000000000003e129

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in x around inf

      \[\leadsto \frac{\color{blue}{4 \cdot x}}{z} \]
   3. Step-by-step derivation

      1. *-commutative N/A

         \[\leadsto \frac{x \cdot \color{blue}{4}}{z} \]

      2. lower-*.f64 35.7

         \[\leadsto \frac{x \cdot \color{blue}{4}}{z} \]

   4. Applied rewrites 35.7%

      \[\leadsto \frac{\color{blue}{x \cdot 4}}{z} \]

   if -1e20 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < -1

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{-2} \]
   3. Step-by-step derivation

   4. Applied rewrites 33.7%

      \[\leadsto \color{blue}{-2} \]

   if 5.0000000000000003e129 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z)

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in y around inf

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]
   3. Step-by-step derivation

      1. lower-*.f64 35.8

         \[\leadsto \frac{-4 \cdot \color{blue}{y}}{z} \]

   4. Applied rewrites 35.8%

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Alternative 6: 66.1% accurate, 0.4× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := \frac{-4 \cdot y}{z}\\ t_1 := \frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z}\\ \mathbf{if}\;t\_1 \leq -4 \cdot 10^{+14}:\\ \;\;\;\;t\_0\\ \mathbf{elif}\;t\_1 \leq -1:\\ \;\;\;\;-2\\ \mathbf{else}:\\ \;\;\;\;t\_0\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z)
 :precision binary64
 (let* ((t_0 (/ (* -4.0 y) z)) (t_1 (/ (* 4.0 (- (- x y) (* z 0.5))) z)))
   (if (<= t_1 -4e+14) t_0 (if (<= t_1 -1.0) -2.0 t_0))))

C

double code(double x, double y, double z) {
	double t_0 = (-4.0 * y) / z;
	double t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	double tmp;
	if (t_1 <= -4e+14) {
		tmp = t_0;
	} else if (t_1 <= -1.0) {
		tmp = -2.0;
	} else {
		tmp = t_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) :: t_0
    real(8) :: t_1
    real(8) :: tmp
    t_0 = ((-4.0d0) * y) / z
    t_1 = (4.0d0 * ((x - y) - (z * 0.5d0))) / z
    if (t_1 <= (-4d+14)) then
        tmp = t_0
    else if (t_1 <= (-1.0d0)) then
        tmp = -2.0d0
    else
        tmp = t_0
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z) {
	double t_0 = (-4.0 * y) / z;
	double t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	double tmp;
	if (t_1 <= -4e+14) {
		tmp = t_0;
	} else if (t_1 <= -1.0) {
		tmp = -2.0;
	} else {
		tmp = t_0;
	}
	return tmp;
}

Python

def code(x, y, z):
	t_0 = (-4.0 * y) / z
	t_1 = (4.0 * ((x - y) - (z * 0.5))) / z
	tmp = 0
	if t_1 <= -4e+14:
		tmp = t_0
	elif t_1 <= -1.0:
		tmp = -2.0
	else:
		tmp = t_0
	return tmp

Julia

function code(x, y, z)
	t_0 = Float64(Float64(-4.0 * y) / z)
	t_1 = Float64(Float64(4.0 * Float64(Float64(x - y) - Float64(z * 0.5))) / z)
	tmp = 0.0
	if (t_1 <= -4e+14)
		tmp = t_0;
	elseif (t_1 <= -1.0)
		tmp = -2.0;
	else
		tmp = t_0;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z)
	t_0 = (-4.0 * y) / z;
	t_1 = (4.0 * ((x - y) - (z * 0.5))) / z;
	tmp = 0.0;
	if (t_1 <= -4e+14)
		tmp = t_0;
	elseif (t_1 <= -1.0)
		tmp = -2.0;
	else
		tmp = t_0;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_] := Block[{t$95$0 = N[(N[(-4.0 * y), $MachinePrecision] / z), $MachinePrecision]}, Block[{t$95$1 = N[(N[(4.0 * N[(N[(x - y), $MachinePrecision] - N[(z * 0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision]}, If[LessEqual[t$95$1, -4e+14], t$95$0, If[LessEqual[t$95$1, -1.0], -2.0, t$95$0]]]]

Derivation

1. Split input into 2 regimes

2. if (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < -4e14 or -1 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z)

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in y around inf

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]
   3. Step-by-step derivation

      1. lower-*.f64 35.8

         \[\leadsto \frac{-4 \cdot \color{blue}{y}}{z} \]

   4. Applied rewrites 35.8%

      \[\leadsto \frac{\color{blue}{-4 \cdot y}}{z} \]

   if -4e14 < (/.f64 (*.f64 #s(literal 4 binary64) (-.f64 (-.f64 x y) (*.f64 z #s(literal 1/2 binary64)))) z) < -1

   1. Initial program 99.9%

      \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

   2. Taylor expanded in z around inf

      \[\leadsto \color{blue}{-2} \]
   3. Step-by-step derivation

   4. Applied rewrites 33.7%

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

Alternative 7: 33.7% accurate, 15.5× speedup

Math

\[\begin{array}{l} \\ -2 \end{array} \]

FPCore

(FPCore (x y z) :precision binary64 -2.0)

C

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

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 = -2.0d0
end function

Java

public static double code(double x, double y, double z) {
	return -2.0;
}

Python

def code(x, y, z):
	return -2.0

Julia

function code(x, y, z)
	return -2.0
end

MATLAB

function tmp = code(x, y, z)
	tmp = -2.0;
end

Wolfram

code[x_, y_, z_] := -2.0

Derivation

1. Initial program 99.9%

   \[\frac{4 \cdot \left(\left(x - y\right) - z \cdot 0.5\right)}{z} \]

2. Taylor expanded in z around inf

   \[\leadsto \color{blue}{-2} \]
3. Step-by-step derivation

4. Applied rewrites 33.7%

   \[\leadsto \color{blue}{-2} \]
5. Add Preprocessing

Reproduce

herbie shell --seed 2025136 
(FPCore (x y z)
  :name "Data.Array.Repa.Algorithms.ColorRamp:rampColorHotToCold from repa-algorithms-3.4.0.1, B"
  :precision binary64
  (/ (* 4.0 (- (- x y) (* z 0.5))) z))