Numeric.Signal.Multichannel:$cput from hsignal-0.2.7.1

Percentage Accurate: 97.1% → 97.4%

Time: 5.1s

Alternatives: 14

Speedup: 0.7×

Specification

Math

\[\begin{array}{l} \\ \frac{x - y}{z - y} \cdot t \end{array} \]

FPCore

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

C

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

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, t)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = ((x - y) / (z - y)) * t
end function

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

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

Initial Program: 97.1% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{x - y}{z - y} \cdot t \end{array} \]

FPCore

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

C

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

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, t)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = ((x - y) / (z - y)) * t
end function

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

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

Alternative 1: 97.4% accurate, 0.7× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_m \leq 3 \cdot 10^{-46}:\\ \;\;\;\;\frac{-1}{y - z} \cdot \left(t\_m \cdot \left(x - y\right)\right)\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m}{z - y} \cdot \left(x - y\right)\\ \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (*
  t_s
  (if (<= t_m 3e-46)
    (* (/ -1.0 (- y z)) (* t_m (- x y)))
    (* (/ t_m (- z y)) (- x y)))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double tmp;
	if (t_m <= 3e-46) {
		tmp = (-1.0 / (y - z)) * (t_m * (x - y));
	} else {
		tmp = (t_m / (z - y)) * (x - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: tmp
    if (t_m <= 3d-46) then
        tmp = ((-1.0d0) / (y - z)) * (t_m * (x - y))
    else
        tmp = (t_m / (z - y)) * (x - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double tmp;
	if (t_m <= 3e-46) {
		tmp = (-1.0 / (y - z)) * (t_m * (x - y));
	} else {
		tmp = (t_m / (z - y)) * (x - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	tmp = 0
	if t_m <= 3e-46:
		tmp = (-1.0 / (y - z)) * (t_m * (x - y))
	else:
		tmp = (t_m / (z - y)) * (x - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	tmp = 0.0
	if (t_m <= 3e-46)
		tmp = Float64(Float64(-1.0 / Float64(y - z)) * Float64(t_m * Float64(x - y)));
	else
		tmp = Float64(Float64(t_m / Float64(z - y)) * Float64(x - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	tmp = 0.0;
	if (t_m <= 3e-46)
		tmp = (-1.0 / (y - z)) * (t_m * (x - y));
	else
		tmp = (t_m / (z - y)) * (x - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := N[(t$95$s * If[LessEqual[t$95$m, 3e-46], N[(N[(-1.0 / N[(y - z), $MachinePrecision]), $MachinePrecision] * N[(t$95$m * N[(x - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], N[(N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision] * N[(x - y), $MachinePrecision]), $MachinePrecision]]), $MachinePrecision]

Derivation

1. Split input into 2 regimes

2. if t < 2.99999999999999987e-46

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. associate-*l/ N/A

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

      4. div-flip N/A

         \[\leadsto \color{blue}{\frac{1}{\frac{z - y}{\left(x - y\right) \cdot t}}} \]

      5. associate-/r/ N/A

         \[\leadsto \color{blue}{\frac{1}{z - y} \cdot \left(\left(x - y\right) \cdot t\right)} \]

      6. lower-*.f64 N/A

         \[\leadsto \color{blue}{\frac{1}{z - y} \cdot \left(\left(x - y\right) \cdot t\right)} \]

      7. frac-2neg N/A

         \[\leadsto \color{blue}{\frac{\mathsf{neg}\left(1\right)}{\mathsf{neg}\left(\left(z - y\right)\right)}} \cdot \left(\left(x - y\right) \cdot t\right) \]

      8. metadata-eval N/A

         \[\leadsto \frac{\color{blue}{-1}}{\mathsf{neg}\left(\left(z - y\right)\right)} \cdot \left(\left(x - y\right) \cdot t\right) \]

      9. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{-1}{\mathsf{neg}\left(\left(z - y\right)\right)}} \cdot \left(\left(x - y\right) \cdot t\right) \]

      10. lift--.f64 N/A

          \[\leadsto \frac{-1}{\mathsf{neg}\left(\color{blue}{\left(z - y\right)}\right)} \cdot \left(\left(x - y\right) \cdot t\right) \]

      11. sub-negate-rev N/A

          \[\leadsto \frac{-1}{\color{blue}{y - z}} \cdot \left(\left(x - y\right) \cdot t\right) \]

      12. lower--.f64 N/A

          \[\leadsto \frac{-1}{\color{blue}{y - z}} \cdot \left(\left(x - y\right) \cdot t\right) \]

      13. *-commutative N/A

          \[\leadsto \frac{-1}{y - z} \cdot \color{blue}{\left(t \cdot \left(x - y\right)\right)} \]

      14. lower-*.f64 84.3

          \[\leadsto \frac{-1}{y - z} \cdot \color{blue}{\left(t \cdot \left(x - y\right)\right)} \]

   3. Applied rewrites 84.3%

      \[\leadsto \color{blue}{\frac{-1}{y - z} \cdot \left(t \cdot \left(x - y\right)\right)} \]

   if 2.99999999999999987e-46 < t

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. associate-/r/ N/A

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

      7. lower-*.f64 N/A

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

      8. lower-/.f64 85.1

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

   3. Applied rewrites 85.1%

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

Alternative 2: 97.4% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq 2 \cdot 10^{-312}:\\ \;\;\;\;\frac{t\_m}{z - y} \cdot \left(x - y\right)\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{x - y}{z} \cdot t\_m\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;\frac{t\_m}{\frac{y - z}{y}}\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 2e-312)
      (* (/ t_m (- z y)) (- x y))
      (if (<= t_2 4e-27)
        (* (/ (- x y) z) t_m)
        (if (<= t_2 2.0) (/ t_m (/ (- y z) y)) (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 2e-312) {
		tmp = (t_m / (z - y)) * (x - y);
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = t_m / ((y - z) / y);
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= 2d-312) then
        tmp = (t_m / (z - y)) * (x - y)
    else if (t_2 <= 4d-27) then
        tmp = ((x - y) / z) * t_m
    else if (t_2 <= 2.0d0) then
        tmp = t_m / ((y - z) / y)
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 2e-312) {
		tmp = (t_m / (z - y)) * (x - y);
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = t_m / ((y - z) / y);
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= 2e-312:
		tmp = (t_m / (z - y)) * (x - y)
	elif t_2 <= 4e-27:
		tmp = ((x - y) / z) * t_m
	elif t_2 <= 2.0:
		tmp = t_m / ((y - z) / y)
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= 2e-312)
		tmp = Float64(Float64(t_m / Float64(z - y)) * Float64(x - y));
	elseif (t_2 <= 4e-27)
		tmp = Float64(Float64(Float64(x - y) / z) * t_m);
	elseif (t_2 <= 2.0)
		tmp = Float64(t_m / Float64(Float64(y - z) / y));
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= 2e-312)
		tmp = (t_m / (z - y)) * (x - y);
	elseif (t_2 <= 4e-27)
		tmp = ((x - y) / z) * t_m;
	elseif (t_2 <= 2.0)
		tmp = t_m / ((y - z) / y);
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, 2e-312], N[(N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision] * N[(x - y), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(N[(N[(x - y), $MachinePrecision] / z), $MachinePrecision] * t$95$m), $MachinePrecision], If[LessEqual[t$95$2, 2.0], N[(t$95$m / N[(N[(y - z), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 2.0000000000019e-312

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. associate-/r/ N/A

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

      7. lower-*.f64 N/A

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

      8. lower-/.f64 85.1

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

   3. Applied rewrites 85.1%

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

   if 2.0000000000019e-312 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

   4. Applied rewrites 50.3%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in x around 0

      \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y}}} \]
   5. Step-by-step derivation

   6. Applied rewrites 54.2%

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

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 3: 96.6% accurate, 0.5× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq 2 \cdot 10^{+101}:\\ \;\;\;\;t\_2 \cdot t\_m\\ \mathbf{else}:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (* t_s (if (<= t_2 2e+101) (* t_2 t_m) (* x (/ t_m (- z y)))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 2e+101) {
		tmp = t_2 * t_m;
	} else {
		tmp = x * (t_m / (z - y));
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= 2d+101) then
        tmp = t_2 * t_m
    else
        tmp = x * (t_m / (z - y))
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 2e+101) {
		tmp = t_2 * t_m;
	} else {
		tmp = x * (t_m / (z - y));
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= 2e+101:
		tmp = t_2 * t_m
	else:
		tmp = x * (t_m / (z - y))
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= 2e+101)
		tmp = Float64(t_2 * t_m);
	else
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= 2e+101)
		tmp = t_2 * t_m;
	else
		tmp = x * (t_m / (z - y));
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, 2e+101], N[(t$95$2 * t$95$m), $MachinePrecision], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 2e101

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   if 2e101 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

Alternative 4: 93.2% accurate, 0.4× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y} \cdot t\_m\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq 2 \cdot 10^{+300}:\\ \;\;\;\;t\_2\\ \mathbf{else}:\\ \;\;\;\;\frac{\frac{t\_m}{z - y}}{\frac{1}{x}}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (* (/ (- x y) (- z y)) t_m)))
   (* t_s (if (<= t_2 2e+300) t_2 (/ (/ t_m (- z y)) (/ 1.0 x))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = ((x - y) / (z - y)) * t_m;
	double tmp;
	if (t_2 <= 2e+300) {
		tmp = t_2;
	} else {
		tmp = (t_m / (z - y)) / (1.0 / x);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = ((x - y) / (z - y)) * t_m
    if (t_2 <= 2d+300) then
        tmp = t_2
    else
        tmp = (t_m / (z - y)) / (1.0d0 / x)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = ((x - y) / (z - y)) * t_m;
	double tmp;
	if (t_2 <= 2e+300) {
		tmp = t_2;
	} else {
		tmp = (t_m / (z - y)) / (1.0 / x);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = ((x - y) / (z - y)) * t_m
	tmp = 0
	if t_2 <= 2e+300:
		tmp = t_2
	else:
		tmp = (t_m / (z - y)) / (1.0 / x)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(Float64(x - y) / Float64(z - y)) * t_m)
	tmp = 0.0
	if (t_2 <= 2e+300)
		tmp = t_2;
	else
		tmp = Float64(Float64(t_m / Float64(z - y)) / Float64(1.0 / x));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = ((x - y) / (z - y)) * t_m;
	tmp = 0.0;
	if (t_2 <= 2e+300)
		tmp = t_2;
	else
		tmp = (t_m / (z - y)) / (1.0 / x);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision] * t$95$m), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, 2e+300], t$95$2, N[(N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision] / N[(1.0 / x), $MachinePrecision]), $MachinePrecision]]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (*.f64 (/.f64 (-.f64 x y) (-.f64 z y)) t) < 2.0000000000000001e300

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   if 2.0000000000000001e300 < (*.f64 (/.f64 (-.f64 x y) (-.f64 z y)) t)

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

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

      1. lift-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

      2. lift-/.f64 N/A

         \[\leadsto \frac{t}{\color{blue}{\frac{y - z}{y - x}}} \]

      3. mult-flip N/A

         \[\leadsto \frac{t}{\color{blue}{\left(y - z\right) \cdot \frac{1}{y - x}}} \]

      4. associate-/r* N/A

         \[\leadsto \color{blue}{\frac{\frac{t}{y - z}}{\frac{1}{y - x}}} \]

      5. lift--.f64 N/A

         \[\leadsto \frac{\frac{t}{\color{blue}{y - z}}}{\frac{1}{y - x}} \]

      6. sub-negate-rev N/A

         \[\leadsto \frac{\frac{t}{\color{blue}{\mathsf{neg}\left(\left(z - y\right)\right)}}}{\frac{1}{y - x}} \]

      7. distribute-frac-neg2 N/A

         \[\leadsto \frac{\color{blue}{\mathsf{neg}\left(\frac{t}{z - y}\right)}}{\frac{1}{y - x}} \]

      8. lift--.f64 N/A

         \[\leadsto \frac{\mathsf{neg}\left(\frac{t}{z - y}\right)}{\frac{1}{\color{blue}{y - x}}} \]

      9. sub-negate-rev N/A

         \[\leadsto \frac{\mathsf{neg}\left(\frac{t}{z - y}\right)}{\frac{1}{\color{blue}{\mathsf{neg}\left(\left(x - y\right)\right)}}} \]

      10. lift--.f64 N/A

          \[\leadsto \frac{\mathsf{neg}\left(\frac{t}{z - y}\right)}{\frac{1}{\mathsf{neg}\left(\color{blue}{\left(x - y\right)}\right)}} \]

      11. distribute-frac-neg2 N/A

          \[\leadsto \frac{\mathsf{neg}\left(\frac{t}{z - y}\right)}{\color{blue}{\mathsf{neg}\left(\frac{1}{x - y}\right)}} \]

      12. frac-2neg-rev N/A

          \[\leadsto \color{blue}{\frac{\frac{t}{z - y}}{\frac{1}{x - y}}} \]

      13. lower-/.f64 N/A

          \[\leadsto \color{blue}{\frac{\frac{t}{z - y}}{\frac{1}{x - y}}} \]

      14. lower-/.f64 N/A

          \[\leadsto \frac{\color{blue}{\frac{t}{z - y}}}{\frac{1}{x - y}} \]

      15. lift--.f64 N/A

          \[\leadsto \frac{\frac{t}{\color{blue}{z - y}}}{\frac{1}{x - y}} \]

      16. frac-2neg-rev N/A

          \[\leadsto \frac{\frac{t}{z - y}}{\color{blue}{\frac{\mathsf{neg}\left(1\right)}{\mathsf{neg}\left(\left(x - y\right)\right)}}} \]

      17. metadata-eval N/A

          \[\leadsto \frac{\frac{t}{z - y}}{\frac{\color{blue}{-1}}{\mathsf{neg}\left(\left(x - y\right)\right)}} \]

      18. lift--.f64 N/A

          \[\leadsto \frac{\frac{t}{z - y}}{\frac{-1}{\mathsf{neg}\left(\color{blue}{\left(x - y\right)}\right)}} \]

      19. sub-negate-rev N/A

          \[\leadsto \frac{\frac{t}{z - y}}{\frac{-1}{\color{blue}{y - x}}} \]

      20. lift--.f64 N/A

          \[\leadsto \frac{\frac{t}{z - y}}{\frac{-1}{\color{blue}{y - x}}} \]

      21. lower-/.f64 85.0

          \[\leadsto \frac{\frac{t}{z - y}}{\color{blue}{\frac{-1}{y - x}}} \]

   5. Applied rewrites 85.0%

      \[\leadsto \color{blue}{\frac{\frac{t}{z - y}}{\frac{-1}{y - x}}} \]

   6. Taylor expanded in x around inf

      \[\leadsto \frac{\frac{t}{z - y}}{\color{blue}{\frac{1}{x}}} \]
   7. Step-by-step derivation

      1. lower-/.f64 50.6

         \[\leadsto \frac{\frac{t}{z - y}}{\frac{1}{\color{blue}{x}}} \]

   8. Applied rewrites 50.6%

      \[\leadsto \frac{\frac{t}{z - y}}{\color{blue}{\frac{1}{x}}} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 5: 93.1% accurate, 0.2× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq -2000:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-155}:\\ \;\;\;\;\left(x - y\right) \cdot \frac{t\_m}{z}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{t\_m \cdot \left(x - y\right)}{z}\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;\frac{y}{y - z} \cdot t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 -2000.0)
      (* x (/ t_m (- z y)))
      (if (<= t_2 4e-155)
        (* (- x y) (/ t_m z))
        (if (<= t_2 4e-27)
          (/ (* t_m (- x y)) z)
          (if (<= t_2 2.0) (* (/ y (- y z)) t_m) (/ (* t_m x) (- z y)))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-155) {
		tmp = (x - y) * (t_m / z);
	} else if (t_2 <= 4e-27) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= (-2000.0d0)) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 4d-155) then
        tmp = (x - y) * (t_m / z)
    else if (t_2 <= 4d-27) then
        tmp = (t_m * (x - y)) / z
    else if (t_2 <= 2.0d0) then
        tmp = (y / (y - z)) * t_m
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-155) {
		tmp = (x - y) * (t_m / z);
	} else if (t_2 <= 4e-27) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= -2000.0:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 4e-155:
		tmp = (x - y) * (t_m / z)
	elif t_2 <= 4e-27:
		tmp = (t_m * (x - y)) / z
	elif t_2 <= 2.0:
		tmp = (y / (y - z)) * t_m
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= -2000.0)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 4e-155)
		tmp = Float64(Float64(x - y) * Float64(t_m / z));
	elseif (t_2 <= 4e-27)
		tmp = Float64(Float64(t_m * Float64(x - y)) / z);
	elseif (t_2 <= 2.0)
		tmp = Float64(Float64(y / Float64(y - z)) * t_m);
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= -2000.0)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 4e-155)
		tmp = (x - y) * (t_m / z);
	elseif (t_2 <= 4e-27)
		tmp = (t_m * (x - y)) / z;
	elseif (t_2 <= 2.0)
		tmp = (y / (y - z)) * t_m;
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, -2000.0], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-155], N[(N[(x - y), $MachinePrecision] * N[(t$95$m / z), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(N[(t$95$m * N[(x - y), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision], If[LessEqual[t$95$2, 2.0], N[(N[(y / N[(y - z), $MachinePrecision]), $MachinePrecision] * t$95$m), $MachinePrecision], N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]]), $MachinePrecision]]

Derivation

1. Split input into 5 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < -2e3

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if -2e3 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.00000000000000006e-155

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

   4. Applied rewrites 50.3%

      \[\leadsto \frac{x - y}{\color{blue}{z}} \cdot t \]
   5. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. associate-*l/ N/A

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

      4. associate-/l* N/A

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

      5. lower-*.f64 N/A

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

      6. lower-/.f64 47.7

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

   6. Applied rewrites 47.7%

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

   if 4.00000000000000006e-155 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in z around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 47.3

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

   4. Applied rewrites 47.3%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in x around 0

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 45.3

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

   6. Applied rewrites 45.3%

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

      1. lift-/.f64 N/A

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

      2. lift-*.f64 N/A

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

      3. associate-/l* N/A

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

      4. *-commutative N/A

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

      5. lower-*.f64 N/A

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

      6. lower-/.f64 54.3

         \[\leadsto \frac{y}{y - z} \cdot t \]

   8. Applied rewrites 54.3%

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

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
3. Recombined 5 regimes into one program.
4. Add Preprocessing

Alternative 6: 93.1% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq -2000:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{x - y}{z} \cdot t\_m\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;\frac{t\_m}{\frac{y - z}{y}}\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 -2000.0)
      (* x (/ t_m (- z y)))
      (if (<= t_2 4e-27)
        (* (/ (- x y) z) t_m)
        (if (<= t_2 2.0) (/ t_m (/ (- y z) y)) (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = t_m / ((y - z) / y);
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= (-2000.0d0)) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 4d-27) then
        tmp = ((x - y) / z) * t_m
    else if (t_2 <= 2.0d0) then
        tmp = t_m / ((y - z) / y)
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = t_m / ((y - z) / y);
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= -2000.0:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 4e-27:
		tmp = ((x - y) / z) * t_m
	elif t_2 <= 2.0:
		tmp = t_m / ((y - z) / y)
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= -2000.0)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 4e-27)
		tmp = Float64(Float64(Float64(x - y) / z) * t_m);
	elseif (t_2 <= 2.0)
		tmp = Float64(t_m / Float64(Float64(y - z) / y));
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= -2000.0)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 4e-27)
		tmp = ((x - y) / z) * t_m;
	elseif (t_2 <= 2.0)
		tmp = t_m / ((y - z) / y);
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, -2000.0], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(N[(N[(x - y), $MachinePrecision] / z), $MachinePrecision] * t$95$m), $MachinePrecision], If[LessEqual[t$95$2, 2.0], N[(t$95$m / N[(N[(y - z), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < -2e3

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if -2e3 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

   4. Applied rewrites 50.3%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in x around 0

      \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y}}} \]
   5. Step-by-step derivation

   6. Applied rewrites 54.2%

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

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 7: 91.7% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq -2000:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{x - y}{z} \cdot t\_m\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;\frac{y}{y - z} \cdot t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 -2000.0)
      (* x (/ t_m (- z y)))
      (if (<= t_2 4e-27)
        (* (/ (- x y) z) t_m)
        (if (<= t_2 2.0) (* (/ y (- y z)) t_m) (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= (-2000.0d0)) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 4d-27) then
        tmp = ((x - y) / z) * t_m
    else if (t_2 <= 2.0d0) then
        tmp = (y / (y - z)) * t_m
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -2000.0) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = ((x - y) / z) * t_m;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= -2000.0:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 4e-27:
		tmp = ((x - y) / z) * t_m
	elif t_2 <= 2.0:
		tmp = (y / (y - z)) * t_m
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= -2000.0)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 4e-27)
		tmp = Float64(Float64(Float64(x - y) / z) * t_m);
	elseif (t_2 <= 2.0)
		tmp = Float64(Float64(y / Float64(y - z)) * t_m);
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= -2000.0)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 4e-27)
		tmp = ((x - y) / z) * t_m;
	elseif (t_2 <= 2.0)
		tmp = (y / (y - z)) * t_m;
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, -2000.0], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(N[(N[(x - y), $MachinePrecision] / z), $MachinePrecision] * t$95$m), $MachinePrecision], If[LessEqual[t$95$2, 2.0], N[(N[(y / N[(y - z), $MachinePrecision]), $MachinePrecision] * t$95$m), $MachinePrecision], N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < -2e3

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if -2e3 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

   4. Applied rewrites 50.3%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in x around 0

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 45.3

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

   6. Applied rewrites 45.3%

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

      1. lift-/.f64 N/A

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

      2. lift-*.f64 N/A

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

      3. associate-/l* N/A

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

      4. *-commutative N/A

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

      5. lower-*.f64 N/A

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

      6. lower-/.f64 54.3

         \[\leadsto \frac{y}{y - z} \cdot t \]

   8. Applied rewrites 54.3%

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

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 8: 91.4% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq -5 \cdot 10^{-17}:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{t\_m \cdot \left(x - y\right)}{z}\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;\frac{y}{y - z} \cdot t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 -5e-17)
      (* x (/ t_m (- z y)))
      (if (<= t_2 4e-27)
        (/ (* t_m (- x y)) z)
        (if (<= t_2 2.0) (* (/ y (- y z)) t_m) (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -5e-17) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= (-5d-17)) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 4d-27) then
        tmp = (t_m * (x - y)) / z
    else if (t_2 <= 2.0d0) then
        tmp = (y / (y - z)) * t_m
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -5e-17) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = (y / (y - z)) * t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= -5e-17:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 4e-27:
		tmp = (t_m * (x - y)) / z
	elif t_2 <= 2.0:
		tmp = (y / (y - z)) * t_m
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= -5e-17)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 4e-27)
		tmp = Float64(Float64(t_m * Float64(x - y)) / z);
	elseif (t_2 <= 2.0)
		tmp = Float64(Float64(y / Float64(y - z)) * t_m);
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= -5e-17)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 4e-27)
		tmp = (t_m * (x - y)) / z;
	elseif (t_2 <= 2.0)
		tmp = (y / (y - z)) * t_m;
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, -5e-17], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(N[(t$95$m * N[(x - y), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision], If[LessEqual[t$95$2, 2.0], N[(N[(y / N[(y - z), $MachinePrecision]), $MachinePrecision] * t$95$m), $MachinePrecision], N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < -4.9999999999999999e-17

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if -4.9999999999999999e-17 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in z around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 47.3

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

   4. Applied rewrites 47.3%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in x around 0

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 45.3

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

   6. Applied rewrites 45.3%

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

      1. lift-/.f64 N/A

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

      2. lift-*.f64 N/A

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

      3. associate-/l* N/A

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

      4. *-commutative N/A

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

      5. lower-*.f64 N/A

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

      6. lower-/.f64 54.3

         \[\leadsto \frac{y}{y - z} \cdot t \]

   8. Applied rewrites 54.3%

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

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 9: 90.8% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq -5 \cdot 10^{-17}:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 10^{-7}:\\ \;\;\;\;\frac{t\_m \cdot \left(x - y\right)}{z}\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 -5e-17)
      (* x (/ t_m (- z y)))
      (if (<= t_2 1e-7)
        (/ (* t_m (- x y)) z)
        (if (<= t_2 2.0) t_m (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -5e-17) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 1e-7) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= (-5d-17)) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 1d-7) then
        tmp = (t_m * (x - y)) / z
    else if (t_2 <= 2.0d0) then
        tmp = t_m
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= -5e-17) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 1e-7) {
		tmp = (t_m * (x - y)) / z;
	} else if (t_2 <= 2.0) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= -5e-17:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 1e-7:
		tmp = (t_m * (x - y)) / z
	elif t_2 <= 2.0:
		tmp = t_m
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= -5e-17)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 1e-7)
		tmp = Float64(Float64(t_m * Float64(x - y)) / z);
	elseif (t_2 <= 2.0)
		tmp = t_m;
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= -5e-17)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 1e-7)
		tmp = (t_m * (x - y)) / z;
	elseif (t_2 <= 2.0)
		tmp = t_m;
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, -5e-17], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 1e-7], N[(N[(t$95$m * N[(x - y), $MachinePrecision]), $MachinePrecision] / z), $MachinePrecision], If[LessEqual[t$95$2, 2.0], t$95$m, N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < -4.9999999999999999e-17

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if -4.9999999999999999e-17 < (/.f64 (-.f64 x y) (-.f64 z y)) < 9.9999999999999995e-8

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in z around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 47.3

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

   4. Applied rewrites 47.3%

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

   if 9.9999999999999995e-8 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around inf

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

   4. Applied rewrites 35.1%

      \[\leadsto \color{blue}{t} \]

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 10: 81.4% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq 4 \cdot 10^{-198}:\\ \;\;\;\;x \cdot \frac{t\_m}{z - y}\\ \mathbf{elif}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{t\_m}{\frac{z}{x}}\\ \mathbf{elif}\;t\_2 \leq 2:\\ \;\;\;\;t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z - y}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 4e-198)
      (* x (/ t_m (- z y)))
      (if (<= t_2 4e-27)
        (/ t_m (/ z x))
        (if (<= t_2 2.0) t_m (/ (* t_m x) (- z y))))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 4e-198) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = t_m / (z / x);
	} else if (t_2 <= 2.0) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= 4d-198) then
        tmp = x * (t_m / (z - y))
    else if (t_2 <= 4d-27) then
        tmp = t_m / (z / x)
    else if (t_2 <= 2.0d0) then
        tmp = t_m
    else
        tmp = (t_m * x) / (z - y)
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 4e-198) {
		tmp = x * (t_m / (z - y));
	} else if (t_2 <= 4e-27) {
		tmp = t_m / (z / x);
	} else if (t_2 <= 2.0) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / (z - y);
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= 4e-198:
		tmp = x * (t_m / (z - y))
	elif t_2 <= 4e-27:
		tmp = t_m / (z / x)
	elif t_2 <= 2.0:
		tmp = t_m
	else:
		tmp = (t_m * x) / (z - y)
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= 4e-198)
		tmp = Float64(x * Float64(t_m / Float64(z - y)));
	elseif (t_2 <= 4e-27)
		tmp = Float64(t_m / Float64(z / x));
	elseif (t_2 <= 2.0)
		tmp = t_m;
	else
		tmp = Float64(Float64(t_m * x) / Float64(z - y));
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= 4e-198)
		tmp = x * (t_m / (z - y));
	elseif (t_2 <= 4e-27)
		tmp = t_m / (z / x);
	elseif (t_2 <= 2.0)
		tmp = t_m;
	else
		tmp = (t_m * x) / (z - y);
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, 4e-198], N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 4e-27], N[(t$95$m / N[(z / x), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$2, 2.0], t$95$m, N[(N[(t$95$m * x), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]]]]), $MachinePrecision]]

Derivation

1. Split input into 4 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 3.9999999999999996e-198

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if 3.9999999999999996e-198 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in y around 0

      \[\leadsto \frac{t}{\color{blue}{\frac{z}{x}}} \]
   5. Step-by-step derivation

      1. lower-/.f64 39.7

         \[\leadsto \frac{t}{\frac{z}{\color{blue}{x}}} \]

   6. Applied rewrites 39.7%

      \[\leadsto \frac{t}{\color{blue}{\frac{z}{x}}} \]

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 2

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around inf

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

   4. Applied rewrites 35.1%

      \[\leadsto \color{blue}{t} \]

   if 2 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

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

Alternative 11: 80.6% accurate, 0.3× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := x \cdot \frac{t\_m}{z - y}\\ t_3 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_3 \leq 4 \cdot 10^{-198}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_3 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{t\_m}{\frac{z}{x}}\\ \mathbf{elif}\;t\_3 \leq 5 \cdot 10^{+22}:\\ \;\;\;\;t\_m\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (* x (/ t_m (- z y)))) (t_3 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_3 4e-198)
      t_2
      (if (<= t_3 4e-27) (/ t_m (/ z x)) (if (<= t_3 5e+22) t_m t_2))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = x * (t_m / (z - y));
	double t_3 = (x - y) / (z - y);
	double tmp;
	if (t_3 <= 4e-198) {
		tmp = t_2;
	} else if (t_3 <= 4e-27) {
		tmp = t_m / (z / x);
	} else if (t_3 <= 5e+22) {
		tmp = t_m;
	} else {
		tmp = t_2;
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: t_3
    real(8) :: tmp
    t_2 = x * (t_m / (z - y))
    t_3 = (x - y) / (z - y)
    if (t_3 <= 4d-198) then
        tmp = t_2
    else if (t_3 <= 4d-27) then
        tmp = t_m / (z / x)
    else if (t_3 <= 5d+22) then
        tmp = t_m
    else
        tmp = t_2
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = x * (t_m / (z - y));
	double t_3 = (x - y) / (z - y);
	double tmp;
	if (t_3 <= 4e-198) {
		tmp = t_2;
	} else if (t_3 <= 4e-27) {
		tmp = t_m / (z / x);
	} else if (t_3 <= 5e+22) {
		tmp = t_m;
	} else {
		tmp = t_2;
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = x * (t_m / (z - y))
	t_3 = (x - y) / (z - y)
	tmp = 0
	if t_3 <= 4e-198:
		tmp = t_2
	elif t_3 <= 4e-27:
		tmp = t_m / (z / x)
	elif t_3 <= 5e+22:
		tmp = t_m
	else:
		tmp = t_2
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(x * Float64(t_m / Float64(z - y)))
	t_3 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_3 <= 4e-198)
		tmp = t_2;
	elseif (t_3 <= 4e-27)
		tmp = Float64(t_m / Float64(z / x));
	elseif (t_3 <= 5e+22)
		tmp = t_m;
	else
		tmp = t_2;
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = x * (t_m / (z - y));
	t_3 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_3 <= 4e-198)
		tmp = t_2;
	elseif (t_3 <= 4e-27)
		tmp = t_m / (z / x);
	elseif (t_3 <= 5e+22)
		tmp = t_m;
	else
		tmp = t_2;
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(x * N[(t$95$m / N[(z - y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]}, Block[{t$95$3 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$3, 4e-198], t$95$2, If[LessEqual[t$95$3, 4e-27], N[(t$95$m / N[(z / x), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$3, 5e+22], t$95$m, t$95$2]]]), $MachinePrecision]]]

Derivation

1. Split input into 3 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 3.9999999999999996e-198 or 4.9999999999999996e22 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in x around inf

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 N/A

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

      3. lower--.f64 50.4

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

   4. Applied rewrites 50.4%

      \[\leadsto \color{blue}{\frac{t \cdot x}{z - y}} \]
   5. Step-by-step derivation

      1. lift--.f64 N/A

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

      2. lift-/.f64 N/A

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

      3. lift-*.f64 N/A

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

      4. *-commutative N/A

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

      5. associate-/l* N/A

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

      6. lower-*.f64 N/A

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

      7. lower-/.f64 N/A

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

      8. lift--.f64 50.7

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

   6. Applied rewrites 50.7%

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

   if 3.9999999999999996e-198 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]
   2. Step-by-step derivation

      1. lift-*.f64 N/A

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

      2. *-commutative N/A

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

      3. lift-/.f64 N/A

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

      4. div-flip N/A

         \[\leadsto t \cdot \color{blue}{\frac{1}{\frac{z - y}{x - y}}} \]

      5. mult-flip-rev N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      6. lower-/.f64 N/A

         \[\leadsto \color{blue}{\frac{t}{\frac{z - y}{x - y}}} \]

      7. frac-2neg N/A

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

      8. lower-/.f64 N/A

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

      9. lift--.f64 N/A

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

      10. sub-negate-rev N/A

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

      11. lower--.f64 N/A

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

      12. lift--.f64 N/A

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

      13. sub-negate-rev N/A

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

      14. lower--.f64 97.0

          \[\leadsto \frac{t}{\frac{y - z}{\color{blue}{y - x}}} \]

   3. Applied rewrites 97.0%

      \[\leadsto \color{blue}{\frac{t}{\frac{y - z}{y - x}}} \]

   4. Taylor expanded in y around 0

      \[\leadsto \frac{t}{\color{blue}{\frac{z}{x}}} \]
   5. Step-by-step derivation

      1. lower-/.f64 39.7

         \[\leadsto \frac{t}{\frac{z}{\color{blue}{x}}} \]

   6. Applied rewrites 39.7%

      \[\leadsto \frac{t}{\color{blue}{\frac{z}{x}}} \]

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 4.9999999999999996e22

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around inf

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

   4. Applied rewrites 35.1%

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

Alternative 12: 68.8% accurate, 0.4× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_2 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;\frac{x}{z} \cdot t\_m\\ \mathbf{elif}\;t\_2 \leq 10^{+23}:\\ \;\;\;\;t\_m\\ \mathbf{else}:\\ \;\;\;\;\frac{t\_m \cdot x}{z}\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (- x y) (- z y))))
   (*
    t_s
    (if (<= t_2 4e-27)
      (* (/ x z) t_m)
      (if (<= t_2 1e+23) t_m (/ (* t_m x) z))))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 4e-27) {
		tmp = (x / z) * t_m;
	} else if (t_2 <= 1e+23) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / z;
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: tmp
    t_2 = (x - y) / (z - y)
    if (t_2 <= 4d-27) then
        tmp = (x / z) * t_m
    else if (t_2 <= 1d+23) then
        tmp = t_m
    else
        tmp = (t_m * x) / z
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (x - y) / (z - y);
	double tmp;
	if (t_2 <= 4e-27) {
		tmp = (x / z) * t_m;
	} else if (t_2 <= 1e+23) {
		tmp = t_m;
	} else {
		tmp = (t_m * x) / z;
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (x - y) / (z - y)
	tmp = 0
	if t_2 <= 4e-27:
		tmp = (x / z) * t_m
	elif t_2 <= 1e+23:
		tmp = t_m
	else:
		tmp = (t_m * x) / z
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_2 <= 4e-27)
		tmp = Float64(Float64(x / z) * t_m);
	elseif (t_2 <= 1e+23)
		tmp = t_m;
	else
		tmp = Float64(Float64(t_m * x) / z);
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_2 <= 4e-27)
		tmp = (x / z) * t_m;
	elseif (t_2 <= 1e+23)
		tmp = t_m;
	else
		tmp = (t_m * x) / z;
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$2, 4e-27], N[(N[(x / z), $MachinePrecision] * t$95$m), $MachinePrecision], If[LessEqual[t$95$2, 1e+23], t$95$m, N[(N[(t$95$m * x), $MachinePrecision] / z), $MachinePrecision]]]), $MachinePrecision]]

Derivation

1. Split input into 3 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

      1. lower-/.f64 39.7

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

   4. Applied rewrites 39.7%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 9.9999999999999992e22

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around inf

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

   4. Applied rewrites 35.1%

      \[\leadsto \color{blue}{t} \]

   if 9.9999999999999992e22 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 37.8

         \[\leadsto \frac{t \cdot x}{z} \]

   4. Applied rewrites 37.8%

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

Alternative 13: 67.6% accurate, 0.4× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ \begin{array}{l} t_2 := \frac{t\_m \cdot x}{z}\\ t_3 := \frac{x - y}{z - y}\\ t\_s \cdot \begin{array}{l} \mathbf{if}\;t\_3 \leq 4 \cdot 10^{-27}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_3 \leq 10^{+23}:\\ \;\;\;\;t\_m\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m)
 :precision binary64
 (let* ((t_2 (/ (* t_m x) z)) (t_3 (/ (- x y) (- z y))))
   (* t_s (if (<= t_3 4e-27) t_2 (if (<= t_3 1e+23) t_m t_2)))))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (t_m * x) / z;
	double t_3 = (x - y) / (z - y);
	double tmp;
	if (t_3 <= 4e-27) {
		tmp = t_2;
	} else if (t_3 <= 1e+23) {
		tmp = t_m;
	} else {
		tmp = t_2;
	}
	return t_s * tmp;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    real(8) :: t_2
    real(8) :: t_3
    real(8) :: tmp
    t_2 = (t_m * x) / z
    t_3 = (x - y) / (z - y)
    if (t_3 <= 4d-27) then
        tmp = t_2
    else if (t_3 <= 1d+23) then
        tmp = t_m
    else
        tmp = t_2
    end if
    code = t_s * tmp
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	double t_2 = (t_m * x) / z;
	double t_3 = (x - y) / (z - y);
	double tmp;
	if (t_3 <= 4e-27) {
		tmp = t_2;
	} else if (t_3 <= 1e+23) {
		tmp = t_m;
	} else {
		tmp = t_2;
	}
	return t_s * tmp;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	t_2 = (t_m * x) / z
	t_3 = (x - y) / (z - y)
	tmp = 0
	if t_3 <= 4e-27:
		tmp = t_2
	elif t_3 <= 1e+23:
		tmp = t_m
	else:
		tmp = t_2
	return t_s * tmp

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	t_2 = Float64(Float64(t_m * x) / z)
	t_3 = Float64(Float64(x - y) / Float64(z - y))
	tmp = 0.0
	if (t_3 <= 4e-27)
		tmp = t_2;
	elseif (t_3 <= 1e+23)
		tmp = t_m;
	else
		tmp = t_2;
	end
	return Float64(t_s * tmp)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp_2 = code(t_s, x, y, z, t_m)
	t_2 = (t_m * x) / z;
	t_3 = (x - y) / (z - y);
	tmp = 0.0;
	if (t_3 <= 4e-27)
		tmp = t_2;
	elseif (t_3 <= 1e+23)
		tmp = t_m;
	else
		tmp = t_2;
	end
	tmp_2 = t_s * tmp;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := Block[{t$95$2 = N[(N[(t$95$m * x), $MachinePrecision] / z), $MachinePrecision]}, Block[{t$95$3 = N[(N[(x - y), $MachinePrecision] / N[(z - y), $MachinePrecision]), $MachinePrecision]}, N[(t$95$s * If[LessEqual[t$95$3, 4e-27], t$95$2, If[LessEqual[t$95$3, 1e+23], t$95$m, t$95$2]]), $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if (/.f64 (-.f64 x y) (-.f64 z y)) < 4.0000000000000002e-27 or 9.9999999999999992e22 < (/.f64 (-.f64 x y) (-.f64 z y))

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around 0

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

      1. lower-/.f64 N/A

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

      2. lower-*.f64 37.8

         \[\leadsto \frac{t \cdot x}{z} \]

   4. Applied rewrites 37.8%

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

   if 4.0000000000000002e-27 < (/.f64 (-.f64 x y) (-.f64 z y)) < 9.9999999999999992e22

   1. Initial program 97.1%

      \[\frac{x - y}{z - y} \cdot t \]

   2. Taylor expanded in y around inf

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

   4. Applied rewrites 35.1%

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

Alternative 14: 35.1% accurate, 12.6× speedup

Math

\[\begin{array}{l} t\_m = \left|t\right| \\ t\_s = \mathsf{copysign}\left(1, t\right) \\ t\_s \cdot t\_m \end{array} \]

FPCore

t\_m = (fabs.f64 t)
t\_s = (copysign.f64 #s(literal 1 binary64) t)
(FPCore (t_s x y z t_m) :precision binary64 (* t_s t_m))

C

t\_m = fabs(t);
t\_s = copysign(1.0, t);
double code(double t_s, double x, double y, double z, double t_m) {
	return t_s * t_m;
}

Fortran

t\_m =     private
t\_s =     private
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(t_s, x, y, z, t_m)
use fmin_fmax_functions
    real(8), intent (in) :: t_s
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t_m
    code = t_s * t_m
end function

Java

t\_m = Math.abs(t);
t\_s = Math.copySign(1.0, t);
public static double code(double t_s, double x, double y, double z, double t_m) {
	return t_s * t_m;
}

Python

t\_m = math.fabs(t)
t\_s = math.copysign(1.0, t)
def code(t_s, x, y, z, t_m):
	return t_s * t_m

Julia

t\_m = abs(t)
t\_s = copysign(1.0, t)
function code(t_s, x, y, z, t_m)
	return Float64(t_s * t_m)
end

MATLAB

t\_m = abs(t);
t\_s = sign(t) * abs(1.0);
function tmp = code(t_s, x, y, z, t_m)
	tmp = t_s * t_m;
end

Wolfram

t\_m = N[Abs[t], $MachinePrecision]
t\_s = N[With[{TMP1 = Abs[1.0], TMP2 = Sign[t]}, TMP1 * If[TMP2 == 0, 1, TMP2]], $MachinePrecision]
code[t$95$s_, x_, y_, z_, t$95$m_] := N[(t$95$s * t$95$m), $MachinePrecision]

Derivation

1. Initial program 97.1%

   \[\frac{x - y}{z - y} \cdot t \]

2. Taylor expanded in y around inf

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

4. Applied rewrites 35.1%

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

Reproduce

herbie shell --seed 2025142 
(FPCore (x y z t)
  :name "Numeric.Signal.Multichannel:$cput from hsignal-0.2.7.1"
  :precision binary64
  (* (/ (- x y) (- z y)) t))