Numeric.SpecFunctions:incompleteBetaWorker from math-functions-0.1.5.2, A

Percentage Accurate: 98.2% → 98.2%

Time: 6.8s

Alternatives: 17

Speedup: 1.0×

• 46.6% of points produce a very large (infinite) output. You may want to add a precondition.

Specification

Math

\[\begin{array}{l} \\ \frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (/ (* x (exp (- (+ (* y (log z)) (* (- t 1.0) (log a))) b))) y))

C

double code(double x, double y, double z, double t, double a, double b) {
	return (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
}

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, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    code = (x * exp((((y * log(z)) + ((t - 1.0d0) * log(a))) - b))) / y
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	return (x * Math.exp((((y * Math.log(z)) + ((t - 1.0) * Math.log(a))) - b))) / y;
}

Python

def code(x, y, z, t, a, b):
	return (x * math.exp((((y * math.log(z)) + ((t - 1.0) * math.log(a))) - b))) / y

Julia

function code(x, y, z, t, a, b)
	return Float64(Float64(x * exp(Float64(Float64(Float64(y * log(z)) + Float64(Float64(t - 1.0) * log(a))) - b))) / y)
end

MATLAB

function tmp = code(x, y, z, t, a, b)
	tmp = (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := N[(N[(x * N[Exp[N[(N[(N[(y * N[Log[z], $MachinePrecision]), $MachinePrecision] + N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]

Initial Program: 98.2% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (/ (* x (exp (- (+ (* y (log z)) (* (- t 1.0) (log a))) b))) y))

C

double code(double x, double y, double z, double t, double a, double b) {
	return (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
}

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, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    code = (x * exp((((y * log(z)) + ((t - 1.0d0) * log(a))) - b))) / y
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	return (x * Math.exp((((y * Math.log(z)) + ((t - 1.0) * Math.log(a))) - b))) / y;
}

Python

def code(x, y, z, t, a, b):
	return (x * math.exp((((y * math.log(z)) + ((t - 1.0) * math.log(a))) - b))) / y

Julia

function code(x, y, z, t, a, b)
	return Float64(Float64(x * exp(Float64(Float64(Float64(y * log(z)) + Float64(Float64(t - 1.0) * log(a))) - b))) / y)
end

MATLAB

function tmp = code(x, y, z, t, a, b)
	tmp = (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := N[(N[(x * N[Exp[N[(N[(N[(y * N[Log[z], $MachinePrecision]), $MachinePrecision] + N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]

Alternative 1: 98.2% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (/ (* x (exp (- (+ (* y (log z)) (* (- t 1.0) (log a))) b))) y))

C

double code(double x, double y, double z, double t, double a, double b) {
	return (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
}

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, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    code = (x * exp((((y * log(z)) + ((t - 1.0d0) * log(a))) - b))) / y
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	return (x * Math.exp((((y * Math.log(z)) + ((t - 1.0) * Math.log(a))) - b))) / y;
}

Python

def code(x, y, z, t, a, b):
	return (x * math.exp((((y * math.log(z)) + ((t - 1.0) * math.log(a))) - b))) / y

Julia

function code(x, y, z, t, a, b)
	return Float64(Float64(x * exp(Float64(Float64(Float64(y * log(z)) + Float64(Float64(t - 1.0) * log(a))) - b))) / y)
end

MATLAB

function tmp = code(x, y, z, t, a, b)
	tmp = (x * exp((((y * log(z)) + ((t - 1.0) * log(a))) - b))) / y;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := N[(N[(x * N[Exp[N[(N[(N[(y * N[Log[z], $MachinePrecision]), $MachinePrecision] + N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]

Derivation

1. Initial program 98.2%

   \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]
2. Add Preprocessing

Alternative 2: 93.5% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \left(t - 1\right) \cdot \log a\\ t_2 := \frac{x \cdot e^{\log a \cdot \left(t - 1\right) - b}}{y}\\ \mathbf{if}\;t\_1 \leq -2 \cdot 10^{+27}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_1 \leq 4 \cdot 10^{+65}:\\ \;\;\;\;x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* (- t 1.0) (log a)))
        (t_2 (/ (* x (exp (- (* (log a) (- t 1.0)) b))) y)))
   (if (<= t_1 -2e+27)
     t_2
     (if (<= t_1 4e+65)
       (* x (/ (* (/ 1.0 a) (exp (- (* (log z) y) b))) y))
       t_2))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * log(a);
	double t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = x * (((1.0 / a) * exp(((log(z) * y) - b))) / y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: t_2
    real(8) :: tmp
    t_1 = (t - 1.0d0) * log(a)
    t_2 = (x * exp(((log(a) * (t - 1.0d0)) - b))) / y
    if (t_1 <= (-2d+27)) then
        tmp = t_2
    else if (t_1 <= 4d+65) then
        tmp = x * (((1.0d0 / a) * exp(((log(z) * y) - b))) / y)
    else
        tmp = t_2
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * Math.log(a);
	double t_2 = (x * Math.exp(((Math.log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = x * (((1.0 / a) * Math.exp(((Math.log(z) * y) - b))) / y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (t - 1.0) * math.log(a)
	t_2 = (x * math.exp(((math.log(a) * (t - 1.0)) - b))) / y
	tmp = 0
	if t_1 <= -2e+27:
		tmp = t_2
	elif t_1 <= 4e+65:
		tmp = x * (((1.0 / a) * math.exp(((math.log(z) * y) - b))) / y)
	else:
		tmp = t_2
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(t - 1.0) * log(a))
	t_2 = Float64(Float64(x * exp(Float64(Float64(log(a) * Float64(t - 1.0)) - b))) / y)
	tmp = 0.0
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = Float64(x * Float64(Float64(Float64(1.0 / a) * exp(Float64(Float64(log(z) * y) - b))) / y));
	else
		tmp = t_2;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (t - 1.0) * log(a);
	t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	tmp = 0.0;
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = x * (((1.0 / a) * exp(((log(z) * y) - b))) / y);
	else
		tmp = t_2;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]}, Block[{t$95$2 = N[(N[(x * N[Exp[N[(N[(N[Log[a], $MachinePrecision] * N[(t - 1.0), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[t$95$1, -2e+27], t$95$2, If[LessEqual[t$95$1, 4e+65], N[(x * N[(N[(N[(1.0 / a), $MachinePrecision] * N[Exp[N[(N[(N[Log[z], $MachinePrecision] * y), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], t$95$2]]]]

Derivation

1. Split input into 2 regimes

2. if (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < -2e27 or 4e65 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a))

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{\left(t - 1\right)} - b}}{y} \]

      2. lift-log.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(\color{blue}{t} - 1\right) - b}}{y} \]

      3. lift--.f64 79.7

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(t - \color{blue}{1}\right) - b}}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]

   if -2e27 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < 4e65

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 3: 93.1% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \left(t - 1\right) \cdot \log a\\ t_2 := \frac{x \cdot e^{\log a \cdot \left(t - 1\right) - b}}{y}\\ \mathbf{if}\;t\_1 \leq -2 \cdot 10^{+27}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_1 \leq 4 \cdot 10^{+65}:\\ \;\;\;\;\frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* (- t 1.0) (log a)))
        (t_2 (/ (* x (exp (- (* (log a) (- t 1.0)) b))) y)))
   (if (<= t_1 -2e+27)
     t_2
     (if (<= t_1 4e+65) (/ (* x (/ (exp (- (* (log z) y) b)) a)) y) t_2))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * log(a);
	double t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = (x * (exp(((log(z) * y) - b)) / a)) / y;
	} else {
		tmp = t_2;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: t_2
    real(8) :: tmp
    t_1 = (t - 1.0d0) * log(a)
    t_2 = (x * exp(((log(a) * (t - 1.0d0)) - b))) / y
    if (t_1 <= (-2d+27)) then
        tmp = t_2
    else if (t_1 <= 4d+65) then
        tmp = (x * (exp(((log(z) * y) - b)) / a)) / y
    else
        tmp = t_2
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * Math.log(a);
	double t_2 = (x * Math.exp(((Math.log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = (x * (Math.exp(((Math.log(z) * y) - b)) / a)) / y;
	} else {
		tmp = t_2;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (t - 1.0) * math.log(a)
	t_2 = (x * math.exp(((math.log(a) * (t - 1.0)) - b))) / y
	tmp = 0
	if t_1 <= -2e+27:
		tmp = t_2
	elif t_1 <= 4e+65:
		tmp = (x * (math.exp(((math.log(z) * y) - b)) / a)) / y
	else:
		tmp = t_2
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(t - 1.0) * log(a))
	t_2 = Float64(Float64(x * exp(Float64(Float64(log(a) * Float64(t - 1.0)) - b))) / y)
	tmp = 0.0
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = Float64(Float64(x * Float64(exp(Float64(Float64(log(z) * y) - b)) / a)) / y);
	else
		tmp = t_2;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (t - 1.0) * log(a);
	t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	tmp = 0.0;
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = (x * (exp(((log(z) * y) - b)) / a)) / y;
	else
		tmp = t_2;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]}, Block[{t$95$2 = N[(N[(x * N[Exp[N[(N[(N[Log[a], $MachinePrecision] * N[(t - 1.0), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[t$95$1, -2e+27], t$95$2, If[LessEqual[t$95$1, 4e+65], N[(N[(x * N[(N[Exp[N[(N[(N[Log[z], $MachinePrecision] * y), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], t$95$2]]]]

Derivation

1. Split input into 2 regimes

2. if (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < -2e27 or 4e65 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a))

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{\left(t - 1\right)} - b}}{y} \]

      2. lift-log.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(\color{blue}{t} - 1\right) - b}}{y} \]

      3. lift--.f64 79.7

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(t - \color{blue}{1}\right) - b}}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]

   if -2e27 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < 4e65

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in y around inf

      \[\leadsto \frac{x \cdot \frac{e^{y \cdot \log z - b}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. exp-diff N/A

         \[\leadsto \frac{x \cdot \frac{\frac{e^{y \cdot \log z}}{e^{b}}}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto \frac{x \cdot \frac{\frac{e^{\log z \cdot y}}{e^{b}}}{a}}{y} \]

      3. exp-diff N/A

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

      4. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

      5. lift-log.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

      6. lift-*.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

      7. lift--.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

      8. lift-exp.f64 79.7

         \[\leadsto \frac{x \cdot \frac{e^{\log z \cdot y - b}}{a}}{y} \]

   7. Applied rewrites 79.7%

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

Alternative 4: 89.1% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \left(t - 1\right) \cdot \log a\\ t_2 := \frac{x \cdot e^{\log a \cdot \left(t - 1\right) - b}}{y}\\ \mathbf{if}\;t\_1 \leq -2 \cdot 10^{+27}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_1 \leq 4 \cdot 10^{+65}:\\ \;\;\;\;\frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y}\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* (- t 1.0) (log a)))
        (t_2 (/ (* x (exp (- (* (log a) (- t 1.0)) b))) y)))
   (if (<= t_1 -2e+27)
     t_2
     (if (<= t_1 4e+65) (/ (* (exp (- (* (log z) y) b)) x) (* a y)) t_2))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * log(a);
	double t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = (exp(((log(z) * y) - b)) * x) / (a * y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: t_2
    real(8) :: tmp
    t_1 = (t - 1.0d0) * log(a)
    t_2 = (x * exp(((log(a) * (t - 1.0d0)) - b))) / y
    if (t_1 <= (-2d+27)) then
        tmp = t_2
    else if (t_1 <= 4d+65) then
        tmp = (exp(((log(z) * y) - b)) * x) / (a * y)
    else
        tmp = t_2
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * Math.log(a);
	double t_2 = (x * Math.exp(((Math.log(a) * (t - 1.0)) - b))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= 4e+65) {
		tmp = (Math.exp(((Math.log(z) * y) - b)) * x) / (a * y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (t - 1.0) * math.log(a)
	t_2 = (x * math.exp(((math.log(a) * (t - 1.0)) - b))) / y
	tmp = 0
	if t_1 <= -2e+27:
		tmp = t_2
	elif t_1 <= 4e+65:
		tmp = (math.exp(((math.log(z) * y) - b)) * x) / (a * y)
	else:
		tmp = t_2
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(t - 1.0) * log(a))
	t_2 = Float64(Float64(x * exp(Float64(Float64(log(a) * Float64(t - 1.0)) - b))) / y)
	tmp = 0.0
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = Float64(Float64(exp(Float64(Float64(log(z) * y) - b)) * x) / Float64(a * y));
	else
		tmp = t_2;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (t - 1.0) * log(a);
	t_2 = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	tmp = 0.0;
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= 4e+65)
		tmp = (exp(((log(z) * y) - b)) * x) / (a * y);
	else
		tmp = t_2;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]}, Block[{t$95$2 = N[(N[(x * N[Exp[N[(N[(N[Log[a], $MachinePrecision] * N[(t - 1.0), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[t$95$1, -2e+27], t$95$2, If[LessEqual[t$95$1, 4e+65], N[(N[(N[Exp[N[(N[(N[Log[z], $MachinePrecision] * y), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision] * x), $MachinePrecision] / N[(a * y), $MachinePrecision]), $MachinePrecision], t$95$2]]]]

Derivation

1. Split input into 2 regimes

2. if (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < -2e27 or 4e65 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a))

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{\left(t - 1\right)} - b}}{y} \]

      2. lift-log.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(\color{blue}{t} - 1\right) - b}}{y} \]

      3. lift--.f64 79.7

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(t - \color{blue}{1}\right) - b}}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]

   if -2e27 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < 4e65

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in x around 0

      \[\leadsto \frac{x \cdot e^{y \cdot \log z - b}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot e^{y \cdot \log z - b}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{e^{y \cdot \log z - b} \cdot x}{a \cdot y} \]

      3. exp-diff N/A

         \[\leadsto \frac{\frac{e^{y \cdot \log z}}{e^{b}} \cdot x}{a \cdot y} \]

      4. *-commutative N/A

         \[\leadsto \frac{\frac{e^{\log z \cdot y}}{e^{b}} \cdot x}{a \cdot y} \]

      5. exp-diff N/A

         \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      7. lift-log.f64 N/A

         \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      8. lift-*.f64 N/A

         \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      9. lift--.f64 N/A

         \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      10. lift-exp.f64 N/A

          \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

      11. lower-*.f64 72.0

          \[\leadsto \frac{e^{\log z \cdot y - b} \cdot x}{a \cdot y} \]

   7. Applied rewrites 72.0%

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

Alternative 5: 88.0% accurate, 1.1× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \frac{x \cdot \frac{{z}^{y}}{a}}{y}\\ \mathbf{if}\;y \leq -1.1 \cdot 10^{+111}:\\ \;\;\;\;t\_1\\ \mathbf{elif}\;y \leq 2.8 \cdot 10^{+123}:\\ \;\;\;\;\frac{x \cdot e^{\log a \cdot \left(t - 1\right) - b}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_1\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (/ (* x (/ (pow z y) a)) y)))
   (if (<= y -1.1e+111)
     t_1
     (if (<= y 2.8e+123) (/ (* x (exp (- (* (log a) (- t 1.0)) b))) y) t_1))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (x * (pow(z, y) / a)) / y;
	double tmp;
	if (y <= -1.1e+111) {
		tmp = t_1;
	} else if (y <= 2.8e+123) {
		tmp = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = (x * ((z ** y) / a)) / y
    if (y <= (-1.1d+111)) then
        tmp = t_1
    else if (y <= 2.8d+123) then
        tmp = (x * exp(((log(a) * (t - 1.0d0)) - b))) / y
    else
        tmp = t_1
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (x * (Math.pow(z, y) / a)) / y;
	double tmp;
	if (y <= -1.1e+111) {
		tmp = t_1;
	} else if (y <= 2.8e+123) {
		tmp = (x * Math.exp(((Math.log(a) * (t - 1.0)) - b))) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (x * (math.pow(z, y) / a)) / y
	tmp = 0
	if y <= -1.1e+111:
		tmp = t_1
	elif y <= 2.8e+123:
		tmp = (x * math.exp(((math.log(a) * (t - 1.0)) - b))) / y
	else:
		tmp = t_1
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(x * Float64((z ^ y) / a)) / y)
	tmp = 0.0
	if (y <= -1.1e+111)
		tmp = t_1;
	elseif (y <= 2.8e+123)
		tmp = Float64(Float64(x * exp(Float64(Float64(log(a) * Float64(t - 1.0)) - b))) / y);
	else
		tmp = t_1;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (x * ((z ^ y) / a)) / y;
	tmp = 0.0;
	if (y <= -1.1e+111)
		tmp = t_1;
	elseif (y <= 2.8e+123)
		tmp = (x * exp(((log(a) * (t - 1.0)) - b))) / y;
	else
		tmp = t_1;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(x * N[(N[Power[z, y], $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[y, -1.1e+111], t$95$1, If[LessEqual[y, 2.8e+123], N[(N[(x * N[Exp[N[(N[(N[Log[a], $MachinePrecision] * N[(t - 1.0), $MachinePrecision]), $MachinePrecision] - b), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], t$95$1]]]

Derivation

1. Split input into 2 regimes

2. if y < -1.09999999999999999e111 or 2.80000000000000011e123 < y

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

      2. lower-pow.f64 58.9

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

   7. Applied rewrites 58.9%

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]

   if -1.09999999999999999e111 < y < 2.80000000000000011e123

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{\left(t - 1\right)} - b}}{y} \]

      2. lift-log.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(\color{blue}{t} - 1\right) - b}}{y} \]

      3. lift--.f64 79.7

         \[\leadsto \frac{x \cdot e^{\log a \cdot \left(t - \color{blue}{1}\right) - b}}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot \left(t - 1\right)} - b}}{y} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 6: 75.1% accurate, 0.6× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \left(t - 1\right) \cdot \log a\\ t_2 := \frac{x \cdot e^{\log a \cdot t}}{y}\\ \mathbf{if}\;t\_1 \leq -2 \cdot 10^{+27}:\\ \;\;\;\;t\_2\\ \mathbf{elif}\;t\_1 \leq -250:\\ \;\;\;\;x \cdot \frac{\frac{e^{-b}}{a}}{y}\\ \mathbf{elif}\;t\_1 \leq 4 \cdot 10^{+116}:\\ \;\;\;\;\frac{{z}^{y} \cdot x}{a \cdot y}\\ \mathbf{else}:\\ \;\;\;\;t\_2\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* (- t 1.0) (log a))) (t_2 (/ (* x (exp (* (log a) t))) y)))
   (if (<= t_1 -2e+27)
     t_2
     (if (<= t_1 -250.0)
       (* x (/ (/ (exp (- b)) a) y))
       (if (<= t_1 4e+116) (/ (* (pow z y) x) (* a y)) t_2)))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * log(a);
	double t_2 = (x * exp((log(a) * t))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= -250.0) {
		tmp = x * ((exp(-b) / a) / y);
	} else if (t_1 <= 4e+116) {
		tmp = (pow(z, y) * x) / (a * y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: t_2
    real(8) :: tmp
    t_1 = (t - 1.0d0) * log(a)
    t_2 = (x * exp((log(a) * t))) / y
    if (t_1 <= (-2d+27)) then
        tmp = t_2
    else if (t_1 <= (-250.0d0)) then
        tmp = x * ((exp(-b) / a) / y)
    else if (t_1 <= 4d+116) then
        tmp = ((z ** y) * x) / (a * y)
    else
        tmp = t_2
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (t - 1.0) * Math.log(a);
	double t_2 = (x * Math.exp((Math.log(a) * t))) / y;
	double tmp;
	if (t_1 <= -2e+27) {
		tmp = t_2;
	} else if (t_1 <= -250.0) {
		tmp = x * ((Math.exp(-b) / a) / y);
	} else if (t_1 <= 4e+116) {
		tmp = (Math.pow(z, y) * x) / (a * y);
	} else {
		tmp = t_2;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (t - 1.0) * math.log(a)
	t_2 = (x * math.exp((math.log(a) * t))) / y
	tmp = 0
	if t_1 <= -2e+27:
		tmp = t_2
	elif t_1 <= -250.0:
		tmp = x * ((math.exp(-b) / a) / y)
	elif t_1 <= 4e+116:
		tmp = (math.pow(z, y) * x) / (a * y)
	else:
		tmp = t_2
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(t - 1.0) * log(a))
	t_2 = Float64(Float64(x * exp(Float64(log(a) * t))) / y)
	tmp = 0.0
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= -250.0)
		tmp = Float64(x * Float64(Float64(exp(Float64(-b)) / a) / y));
	elseif (t_1 <= 4e+116)
		tmp = Float64(Float64((z ^ y) * x) / Float64(a * y));
	else
		tmp = t_2;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (t - 1.0) * log(a);
	t_2 = (x * exp((log(a) * t))) / y;
	tmp = 0.0;
	if (t_1 <= -2e+27)
		tmp = t_2;
	elseif (t_1 <= -250.0)
		tmp = x * ((exp(-b) / a) / y);
	elseif (t_1 <= 4e+116)
		tmp = ((z ^ y) * x) / (a * y);
	else
		tmp = t_2;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(t - 1.0), $MachinePrecision] * N[Log[a], $MachinePrecision]), $MachinePrecision]}, Block[{t$95$2 = N[(N[(x * N[Exp[N[(N[Log[a], $MachinePrecision] * t), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[t$95$1, -2e+27], t$95$2, If[LessEqual[t$95$1, -250.0], N[(x * N[(N[(N[Exp[(-b)], $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[t$95$1, 4e+116], N[(N[(N[Power[z, y], $MachinePrecision] * x), $MachinePrecision] / N[(a * y), $MachinePrecision]), $MachinePrecision], t$95$2]]]]]

Derivation

1. Split input into 3 regimes

2. if (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < -2e27 or 4.00000000000000006e116 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a))

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around inf

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

      1. *-commutative N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{t}}}{y} \]

      2. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot e^{\log a \cdot \color{blue}{t}}}{y} \]

      3. lift-log.f64 48.4

         \[\leadsto \frac{x \cdot e^{\log a \cdot t}}{y} \]

   4. Applied rewrites 48.4%

      \[\leadsto \frac{x \cdot e^{\color{blue}{\log a \cdot t}}}{y} \]

   if -2e27 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < -250

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   if -250 < (*.f64 (-.f64 t #s(literal 1 binary64)) (log.f64 a)) < 4.00000000000000006e116

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

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

Alternative 7: 71.7% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := e^{-b}\\ \mathbf{if}\;b \leq -4.6 \cdot 10^{+21}:\\ \;\;\;\;x \cdot \frac{t\_1}{y}\\ \mathbf{elif}\;b \leq 2.9 \cdot 10^{-10}:\\ \;\;\;\;\frac{{z}^{y} \cdot x}{a \cdot y}\\ \mathbf{else}:\\ \;\;\;\;x \cdot \frac{\frac{t\_1}{a}}{y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (exp (- b))))
   (if (<= b -4.6e+21)
     (* x (/ t_1 y))
     (if (<= b 2.9e-10) (/ (* (pow z y) x) (* a y)) (* x (/ (/ t_1 a) y))))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = exp(-b);
	double tmp;
	if (b <= -4.6e+21) {
		tmp = x * (t_1 / y);
	} else if (b <= 2.9e-10) {
		tmp = (pow(z, y) * x) / (a * y);
	} else {
		tmp = x * ((t_1 / a) / y);
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = exp(-b)
    if (b <= (-4.6d+21)) then
        tmp = x * (t_1 / y)
    else if (b <= 2.9d-10) then
        tmp = ((z ** y) * x) / (a * y)
    else
        tmp = x * ((t_1 / a) / y)
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = Math.exp(-b);
	double tmp;
	if (b <= -4.6e+21) {
		tmp = x * (t_1 / y);
	} else if (b <= 2.9e-10) {
		tmp = (Math.pow(z, y) * x) / (a * y);
	} else {
		tmp = x * ((t_1 / a) / y);
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = math.exp(-b)
	tmp = 0
	if b <= -4.6e+21:
		tmp = x * (t_1 / y)
	elif b <= 2.9e-10:
		tmp = (math.pow(z, y) * x) / (a * y)
	else:
		tmp = x * ((t_1 / a) / y)
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = exp(Float64(-b))
	tmp = 0.0
	if (b <= -4.6e+21)
		tmp = Float64(x * Float64(t_1 / y));
	elseif (b <= 2.9e-10)
		tmp = Float64(Float64((z ^ y) * x) / Float64(a * y));
	else
		tmp = Float64(x * Float64(Float64(t_1 / a) / y));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = exp(-b);
	tmp = 0.0;
	if (b <= -4.6e+21)
		tmp = x * (t_1 / y);
	elseif (b <= 2.9e-10)
		tmp = ((z ^ y) * x) / (a * y);
	else
		tmp = x * ((t_1 / a) / y);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[Exp[(-b)], $MachinePrecision]}, If[LessEqual[b, -4.6e+21], N[(x * N[(t$95$1 / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 2.9e-10], N[(N[(N[Power[z, y], $MachinePrecision] * x), $MachinePrecision] / N[(a * y), $MachinePrecision]), $MachinePrecision], N[(x * N[(N[(t$95$1 / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision]]]]

Derivation

1. Split input into 3 regimes

2. if b < -4.6e21

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in b around inf

      \[\leadsto \frac{x \cdot e^{\color{blue}{-1 \cdot b}}}{y} \]
   3. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto \frac{x \cdot e^{\mathsf{neg}\left(b\right)}}{y} \]

      2. lower-neg.f64 47.1

         \[\leadsto \frac{x \cdot e^{-b}}{y} \]

   4. Applied rewrites 47.1%

      \[\leadsto \frac{x \cdot e^{\color{blue}{-b}}}{y} \]
   5. Step-by-step derivation

      1. lift-/.f64 N/A

         \[\leadsto \color{blue}{\frac{x \cdot e^{-b}}{y}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\color{blue}{x \cdot e^{-b}}}{y} \]

      3. associate-/l* N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      4. lower-*.f64 N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      5. lower-/.f64 47.1

         \[\leadsto x \cdot \color{blue}{\frac{e^{-b}}{y}} \]

   6. Applied rewrites 47.1%

      \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

   if -4.6e21 < b < 2.89999999999999981e-10

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

      \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

   if 2.89999999999999981e-10 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Alternative 8: 59.0% accurate, 1.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := \frac{x \cdot \frac{e^{-b}}{a}}{y}\\ \mathbf{if}\;b \leq -1.02 \cdot 10^{-234}:\\ \;\;\;\;t\_1\\ \mathbf{elif}\;b \leq 1.6 \cdot 10^{-211}:\\ \;\;\;\;x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_1\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (/ (* x (/ (exp (- b)) a)) y)))
   (if (<= b -1.02e-234)
     t_1
     (if (<= b 1.6e-211) (* x (/ (/ (* (* b b) 0.5) a) y)) t_1))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (x * (exp(-b) / a)) / y;
	double tmp;
	if (b <= -1.02e-234) {
		tmp = t_1;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else {
		tmp = t_1;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = (x * (exp(-b) / a)) / y
    if (b <= (-1.02d-234)) then
        tmp = t_1
    else if (b <= 1.6d-211) then
        tmp = x * ((((b * b) * 0.5d0) / a) / y)
    else
        tmp = t_1
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = (x * (Math.exp(-b) / a)) / y;
	double tmp;
	if (b <= -1.02e-234) {
		tmp = t_1;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else {
		tmp = t_1;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = (x * (math.exp(-b) / a)) / y
	tmp = 0
	if b <= -1.02e-234:
		tmp = t_1
	elif b <= 1.6e-211:
		tmp = x * ((((b * b) * 0.5) / a) / y)
	else:
		tmp = t_1
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(Float64(x * Float64(exp(Float64(-b)) / a)) / y)
	tmp = 0.0
	if (b <= -1.02e-234)
		tmp = t_1;
	elseif (b <= 1.6e-211)
		tmp = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y));
	else
		tmp = t_1;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = (x * (exp(-b) / a)) / y;
	tmp = 0.0;
	if (b <= -1.02e-234)
		tmp = t_1;
	elseif (b <= 1.6e-211)
		tmp = x * ((((b * b) * 0.5) / a) / y);
	else
		tmp = t_1;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(N[(x * N[(N[Exp[(-b)], $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision]}, If[LessEqual[b, -1.02e-234], t$95$1, If[LessEqual[b, 1.6e-211], N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], t$95$1]]]

Derivation

1. Split input into 2 regimes

2. if b < -1.01999999999999999e-234 or 1.59999999999999993e-211 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto \frac{x \cdot \frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.0

         \[\leadsto \frac{x \cdot \frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.0%

      \[\leadsto \frac{x \cdot \frac{e^{-b}}{\color{blue}{a}}}{y} \]

   if -1.01999999999999999e-234 < b < 1.59999999999999993e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 9: 58.7% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := x \cdot \frac{e^{-b}}{y}\\ \mathbf{if}\;b \leq -1.05 \cdot 10^{-58}:\\ \;\;\;\;t\_1\\ \mathbf{elif}\;b \leq -1.02 \cdot 10^{-234}:\\ \;\;\;\;\frac{\frac{x}{a}}{y}\\ \mathbf{elif}\;b \leq 1.6 \cdot 10^{-211}:\\ \;\;\;\;x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{elif}\;b \leq 200000000:\\ \;\;\;\;\frac{x \cdot \frac{1}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_1\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* x (/ (exp (- b)) y))))
   (if (<= b -1.05e-58)
     t_1
     (if (<= b -1.02e-234)
       (/ (/ x a) y)
       (if (<= b 1.6e-211)
         (* x (/ (/ (* (* b b) 0.5) a) y))
         (if (<= b 200000000.0) (/ (* x (/ 1.0 a)) y) t_1))))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * (exp(-b) / y);
	double tmp;
	if (b <= -1.05e-58) {
		tmp = t_1;
	} else if (b <= -1.02e-234) {
		tmp = (x / a) / y;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 200000000.0) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = x * (exp(-b) / y)
    if (b <= (-1.05d-58)) then
        tmp = t_1
    else if (b <= (-1.02d-234)) then
        tmp = (x / a) / y
    else if (b <= 1.6d-211) then
        tmp = x * ((((b * b) * 0.5d0) / a) / y)
    else if (b <= 200000000.0d0) then
        tmp = (x * (1.0d0 / a)) / y
    else
        tmp = t_1
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * (Math.exp(-b) / y);
	double tmp;
	if (b <= -1.05e-58) {
		tmp = t_1;
	} else if (b <= -1.02e-234) {
		tmp = (x / a) / y;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 200000000.0) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = x * (math.exp(-b) / y)
	tmp = 0
	if b <= -1.05e-58:
		tmp = t_1
	elif b <= -1.02e-234:
		tmp = (x / a) / y
	elif b <= 1.6e-211:
		tmp = x * ((((b * b) * 0.5) / a) / y)
	elif b <= 200000000.0:
		tmp = (x * (1.0 / a)) / y
	else:
		tmp = t_1
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(x * Float64(exp(Float64(-b)) / y))
	tmp = 0.0
	if (b <= -1.05e-58)
		tmp = t_1;
	elseif (b <= -1.02e-234)
		tmp = Float64(Float64(x / a) / y);
	elseif (b <= 1.6e-211)
		tmp = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y));
	elseif (b <= 200000000.0)
		tmp = Float64(Float64(x * Float64(1.0 / a)) / y);
	else
		tmp = t_1;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = x * (exp(-b) / y);
	tmp = 0.0;
	if (b <= -1.05e-58)
		tmp = t_1;
	elseif (b <= -1.02e-234)
		tmp = (x / a) / y;
	elseif (b <= 1.6e-211)
		tmp = x * ((((b * b) * 0.5) / a) / y);
	elseif (b <= 200000000.0)
		tmp = (x * (1.0 / a)) / y;
	else
		tmp = t_1;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(x * N[(N[Exp[(-b)], $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision]}, If[LessEqual[b, -1.05e-58], t$95$1, If[LessEqual[b, -1.02e-234], N[(N[(x / a), $MachinePrecision] / y), $MachinePrecision], If[LessEqual[b, 1.6e-211], N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 200000000.0], N[(N[(x * N[(1.0 / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], t$95$1]]]]]

Derivation

1. Split input into 4 regimes

2. if b < -1.04999999999999994e-58 or 2e8 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in b around inf

      \[\leadsto \frac{x \cdot e^{\color{blue}{-1 \cdot b}}}{y} \]
   3. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto \frac{x \cdot e^{\mathsf{neg}\left(b\right)}}{y} \]

      2. lower-neg.f64 47.1

         \[\leadsto \frac{x \cdot e^{-b}}{y} \]

   4. Applied rewrites 47.1%

      \[\leadsto \frac{x \cdot e^{\color{blue}{-b}}}{y} \]
   5. Step-by-step derivation

      1. lift-/.f64 N/A

         \[\leadsto \color{blue}{\frac{x \cdot e^{-b}}{y}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\color{blue}{x \cdot e^{-b}}}{y} \]

      3. associate-/l* N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      4. lower-*.f64 N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      5. lower-/.f64 47.1

         \[\leadsto x \cdot \color{blue}{\frac{e^{-b}}{y}} \]

   6. Applied rewrites 47.1%

      \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

   if -1.04999999999999994e-58 < b < -1.01999999999999999e-234

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

      \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   9. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x}{a \cdot y} \]

      2. lift-*.f64 31.2

         \[\leadsto \frac{x}{a \cdot y} \]

   10. Applied rewrites 31.2%

       \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   11. Step-by-step derivation

       1. lift-*.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       2. lift-/.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       3. associate-/r* N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       4. lower-/.f64 N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       5. lower-/.f64 30.6

          \[\leadsto \frac{\frac{x}{a}}{y} \]

   12. Applied rewrites 30.6%

       \[\leadsto \frac{\frac{x}{a}}{y} \]

   if -1.01999999999999999e-234 < b < 1.59999999999999993e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   if 1.59999999999999993e-211 < b < 2e8

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

      2. lower-pow.f64 58.9

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

   7. Applied rewrites 58.9%

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]
   9. Step-by-step derivation

   10. Applied rewrites 30.6%

       \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]
3. Recombined 4 regimes into one program.
4. Add Preprocessing

Alternative 10: 57.5% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := x \cdot \frac{\frac{e^{-b}}{a}}{y}\\ \mathbf{if}\;b \leq -9.2 \cdot 10^{-235}:\\ \;\;\;\;t\_1\\ \mathbf{elif}\;b \leq 1.6 \cdot 10^{-211}:\\ \;\;\;\;x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{elif}\;b \leq 7.7 \cdot 10^{-25}:\\ \;\;\;\;\frac{x \cdot \frac{1}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;t\_1\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* x (/ (/ (exp (- b)) a) y))))
   (if (<= b -9.2e-235)
     t_1
     (if (<= b 1.6e-211)
       (* x (/ (/ (* (* b b) 0.5) a) y))
       (if (<= b 7.7e-25) (/ (* x (/ 1.0 a)) y) t_1)))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * ((exp(-b) / a) / y);
	double tmp;
	if (b <= -9.2e-235) {
		tmp = t_1;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 7.7e-25) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = x * ((exp(-b) / a) / y)
    if (b <= (-9.2d-235)) then
        tmp = t_1
    else if (b <= 1.6d-211) then
        tmp = x * ((((b * b) * 0.5d0) / a) / y)
    else if (b <= 7.7d-25) then
        tmp = (x * (1.0d0 / a)) / y
    else
        tmp = t_1
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * ((Math.exp(-b) / a) / y);
	double tmp;
	if (b <= -9.2e-235) {
		tmp = t_1;
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 7.7e-25) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = t_1;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = x * ((math.exp(-b) / a) / y)
	tmp = 0
	if b <= -9.2e-235:
		tmp = t_1
	elif b <= 1.6e-211:
		tmp = x * ((((b * b) * 0.5) / a) / y)
	elif b <= 7.7e-25:
		tmp = (x * (1.0 / a)) / y
	else:
		tmp = t_1
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(x * Float64(Float64(exp(Float64(-b)) / a) / y))
	tmp = 0.0
	if (b <= -9.2e-235)
		tmp = t_1;
	elseif (b <= 1.6e-211)
		tmp = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y));
	elseif (b <= 7.7e-25)
		tmp = Float64(Float64(x * Float64(1.0 / a)) / y);
	else
		tmp = t_1;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = x * ((exp(-b) / a) / y);
	tmp = 0.0;
	if (b <= -9.2e-235)
		tmp = t_1;
	elseif (b <= 1.6e-211)
		tmp = x * ((((b * b) * 0.5) / a) / y);
	elseif (b <= 7.7e-25)
		tmp = (x * (1.0 / a)) / y;
	else
		tmp = t_1;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(x * N[(N[(N[Exp[(-b)], $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision]}, If[LessEqual[b, -9.2e-235], t$95$1, If[LessEqual[b, 1.6e-211], N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 7.7e-25], N[(N[(x * N[(1.0 / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], t$95$1]]]]

Derivation

1. Split input into 3 regimes

2. if b < -9.19999999999999989e-235 or 7.7000000000000002e-25 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   if -9.19999999999999989e-235 < b < 1.59999999999999993e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   if 1.59999999999999993e-211 < b < 7.7000000000000002e-25

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

      2. lower-pow.f64 58.9

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

   7. Applied rewrites 58.9%

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]
   9. Step-by-step derivation

   10. Applied rewrites 30.6%

       \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Alternative 11: 57.0% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := e^{-b}\\ \mathbf{if}\;b \leq -9.2 \cdot 10^{-235}:\\ \;\;\;\;x \cdot \frac{t\_1}{a \cdot y}\\ \mathbf{elif}\;b \leq 1.66 \cdot 10^{-211}:\\ \;\;\;\;x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{elif}\;b \leq 200000000:\\ \;\;\;\;\frac{x \cdot \frac{\mathsf{fma}\left(\log z, y, 1\right)}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;x \cdot \frac{t\_1}{y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (exp (- b))))
   (if (<= b -9.2e-235)
     (* x (/ t_1 (* a y)))
     (if (<= b 1.66e-211)
       (* x (/ (/ (* (* b b) 0.5) a) y))
       (if (<= b 200000000.0)
         (/ (* x (/ (fma (log z) y 1.0) a)) y)
         (* x (/ t_1 y)))))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = exp(-b);
	double tmp;
	if (b <= -9.2e-235) {
		tmp = x * (t_1 / (a * y));
	} else if (b <= 1.66e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 200000000.0) {
		tmp = (x * (fma(log(z), y, 1.0) / a)) / y;
	} else {
		tmp = x * (t_1 / y);
	}
	return tmp;
}

Julia

function code(x, y, z, t, a, b)
	t_1 = exp(Float64(-b))
	tmp = 0.0
	if (b <= -9.2e-235)
		tmp = Float64(x * Float64(t_1 / Float64(a * y)));
	elseif (b <= 1.66e-211)
		tmp = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y));
	elseif (b <= 200000000.0)
		tmp = Float64(Float64(x * Float64(fma(log(z), y, 1.0) / a)) / y);
	else
		tmp = Float64(x * Float64(t_1 / y));
	end
	return tmp
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[Exp[(-b)], $MachinePrecision]}, If[LessEqual[b, -9.2e-235], N[(x * N[(t$95$1 / N[(a * y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 1.66e-211], N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 200000000.0], N[(N[(x * N[(N[(N[Log[z], $MachinePrecision] * y + 1.0), $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], N[(x * N[(t$95$1 / y), $MachinePrecision]), $MachinePrecision]]]]]

Derivation

1. Split input into 4 regimes

2. if b < -9.19999999999999989e-235

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{a \cdot \color{blue}{y}} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

      3. lower-exp.f64 N/A

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

      4. lower-*.f64 54.3

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

   7. Applied rewrites 54.3%

      \[\leadsto x \cdot \frac{e^{-b}}{\color{blue}{a \cdot y}} \]

   if -9.19999999999999989e-235 < b < 1.6599999999999999e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   if 1.6599999999999999e-211 < b < 2e8

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

      2. lower-pow.f64 58.9

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

   7. Applied rewrites 58.9%

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot \frac{1 + y \cdot \log z}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{x \cdot \frac{y \cdot \log z + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto \frac{x \cdot \frac{\log z \cdot y + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto \frac{x \cdot \frac{\mathsf{fma}\left(\log z, y, 1\right)}{a}}{y} \]

      4. lift-log.f64 31.1

         \[\leadsto \frac{x \cdot \frac{\mathsf{fma}\left(\log z, y, 1\right)}{a}}{y} \]

   10. Applied rewrites 31.1%

       \[\leadsto \frac{x \cdot \frac{\mathsf{fma}\left(\log z, y, 1\right)}{a}}{y} \]

   if 2e8 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in b around inf

      \[\leadsto \frac{x \cdot e^{\color{blue}{-1 \cdot b}}}{y} \]
   3. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto \frac{x \cdot e^{\mathsf{neg}\left(b\right)}}{y} \]

      2. lower-neg.f64 47.1

         \[\leadsto \frac{x \cdot e^{-b}}{y} \]

   4. Applied rewrites 47.1%

      \[\leadsto \frac{x \cdot e^{\color{blue}{-b}}}{y} \]
   5. Step-by-step derivation

      1. lift-/.f64 N/A

         \[\leadsto \color{blue}{\frac{x \cdot e^{-b}}{y}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\color{blue}{x \cdot e^{-b}}}{y} \]

      3. associate-/l* N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      4. lower-*.f64 N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      5. lower-/.f64 47.1

         \[\leadsto x \cdot \color{blue}{\frac{e^{-b}}{y}} \]

   6. Applied rewrites 47.1%

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

Alternative 12: 56.8% accurate, 1.4× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := e^{-b}\\ \mathbf{if}\;b \leq -9.2 \cdot 10^{-235}:\\ \;\;\;\;x \cdot \frac{t\_1}{a \cdot y}\\ \mathbf{elif}\;b \leq 1.6 \cdot 10^{-211}:\\ \;\;\;\;x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{elif}\;b \leq 200000000:\\ \;\;\;\;\frac{x \cdot \frac{1}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;x \cdot \frac{t\_1}{y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (exp (- b))))
   (if (<= b -9.2e-235)
     (* x (/ t_1 (* a y)))
     (if (<= b 1.6e-211)
       (* x (/ (/ (* (* b b) 0.5) a) y))
       (if (<= b 200000000.0) (/ (* x (/ 1.0 a)) y) (* x (/ t_1 y)))))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = exp(-b);
	double tmp;
	if (b <= -9.2e-235) {
		tmp = x * (t_1 / (a * y));
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 200000000.0) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = x * (t_1 / y);
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = exp(-b)
    if (b <= (-9.2d-235)) then
        tmp = x * (t_1 / (a * y))
    else if (b <= 1.6d-211) then
        tmp = x * ((((b * b) * 0.5d0) / a) / y)
    else if (b <= 200000000.0d0) then
        tmp = (x * (1.0d0 / a)) / y
    else
        tmp = x * (t_1 / y)
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = Math.exp(-b);
	double tmp;
	if (b <= -9.2e-235) {
		tmp = x * (t_1 / (a * y));
	} else if (b <= 1.6e-211) {
		tmp = x * ((((b * b) * 0.5) / a) / y);
	} else if (b <= 200000000.0) {
		tmp = (x * (1.0 / a)) / y;
	} else {
		tmp = x * (t_1 / y);
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = math.exp(-b)
	tmp = 0
	if b <= -9.2e-235:
		tmp = x * (t_1 / (a * y))
	elif b <= 1.6e-211:
		tmp = x * ((((b * b) * 0.5) / a) / y)
	elif b <= 200000000.0:
		tmp = (x * (1.0 / a)) / y
	else:
		tmp = x * (t_1 / y)
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = exp(Float64(-b))
	tmp = 0.0
	if (b <= -9.2e-235)
		tmp = Float64(x * Float64(t_1 / Float64(a * y)));
	elseif (b <= 1.6e-211)
		tmp = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y));
	elseif (b <= 200000000.0)
		tmp = Float64(Float64(x * Float64(1.0 / a)) / y);
	else
		tmp = Float64(x * Float64(t_1 / y));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = exp(-b);
	tmp = 0.0;
	if (b <= -9.2e-235)
		tmp = x * (t_1 / (a * y));
	elseif (b <= 1.6e-211)
		tmp = x * ((((b * b) * 0.5) / a) / y);
	elseif (b <= 200000000.0)
		tmp = (x * (1.0 / a)) / y;
	else
		tmp = x * (t_1 / y);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[Exp[(-b)], $MachinePrecision]}, If[LessEqual[b, -9.2e-235], N[(x * N[(t$95$1 / N[(a * y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 1.6e-211], N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], If[LessEqual[b, 200000000.0], N[(N[(x * N[(1.0 / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], N[(x * N[(t$95$1 / y), $MachinePrecision]), $MachinePrecision]]]]]

Derivation

1. Split input into 4 regimes

2. if b < -9.19999999999999989e-235

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\mathsf{neg}\left(b\right)}}{a \cdot \color{blue}{y}} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

      3. lower-exp.f64 N/A

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

      4. lower-*.f64 54.3

         \[\leadsto x \cdot \frac{e^{-b}}{a \cdot y} \]

   7. Applied rewrites 54.3%

      \[\leadsto x \cdot \frac{e^{-b}}{\color{blue}{a \cdot y}} \]

   if -9.19999999999999989e-235 < b < 1.59999999999999993e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   if 1.59999999999999993e-211 < b < 2e8

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \frac{x \cdot \color{blue}{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}}{y} \]
   3. Step-by-step derivation

      1. associate--l+ N/A

         \[\leadsto \frac{x \cdot e^{-1 \cdot \log a + \left(y \cdot \log z - b\right)}}{y} \]

      2. exp-sum N/A

         \[\leadsto \frac{x \cdot \left(e^{-1 \cdot \log a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      3. *-commutative N/A

         \[\leadsto \frac{x \cdot \left(e^{\log a \cdot -1} \cdot e^{\color{blue}{y \cdot \log z} - b}\right)}{y} \]

      4. pow-to-exp N/A

         \[\leadsto \frac{x \cdot \left({a}^{-1} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      5. inv-pow N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      6. lower-*.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot \color{blue}{e^{y \cdot \log z - b}}\right)}{y} \]

      7. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\color{blue}{y \cdot \log z - b}}\right)}{y} \]

      8. lower-exp.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      9. lower--.f64 N/A

         \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{y \cdot \log z - b}\right)}{y} \]

      10. *-commutative N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      11. lower-*.f64 N/A

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

      12. lift-log.f64 79.7

          \[\leadsto \frac{x \cdot \left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}{y} \]

   4. Applied rewrites 79.7%

      \[\leadsto \frac{x \cdot \color{blue}{\left(\frac{1}{a} \cdot e^{\log z \cdot y - b}\right)}}{y} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

      2. lower-pow.f64 58.9

         \[\leadsto \frac{x \cdot \frac{{z}^{y}}{a}}{y} \]

   7. Applied rewrites 58.9%

      \[\leadsto \frac{x \cdot \frac{{z}^{y}}{\color{blue}{a}}}{y} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]
   9. Step-by-step derivation

   10. Applied rewrites 30.6%

       \[\leadsto \frac{x \cdot \frac{1}{a}}{y} \]

   if 2e8 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in b around inf

      \[\leadsto \frac{x \cdot e^{\color{blue}{-1 \cdot b}}}{y} \]
   3. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto \frac{x \cdot e^{\mathsf{neg}\left(b\right)}}{y} \]

      2. lower-neg.f64 47.1

         \[\leadsto \frac{x \cdot e^{-b}}{y} \]

   4. Applied rewrites 47.1%

      \[\leadsto \frac{x \cdot e^{\color{blue}{-b}}}{y} \]
   5. Step-by-step derivation

      1. lift-/.f64 N/A

         \[\leadsto \color{blue}{\frac{x \cdot e^{-b}}{y}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\color{blue}{x \cdot e^{-b}}}{y} \]

      3. associate-/l* N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      4. lower-*.f64 N/A

         \[\leadsto \color{blue}{x \cdot \frac{e^{-b}}{y}} \]

      5. lower-/.f64 47.1

         \[\leadsto x \cdot \color{blue}{\frac{e^{-b}}{y}} \]

   6. Applied rewrites 47.1%

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

Alternative 13: 39.6% accurate, 1.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_1 := x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y}\\ \mathbf{if}\;b \leq -1.05 \cdot 10^{+170}:\\ \;\;\;\;t\_1\\ \mathbf{elif}\;b \leq -1.02 \cdot 10^{-234}:\\ \;\;\;\;\frac{x \cdot \frac{\left(-b\right) + 1}{a}}{y}\\ \mathbf{elif}\;b \leq 1.6 \cdot 10^{-211}:\\ \;\;\;\;t\_1\\ \mathbf{else}:\\ \;\;\;\;\frac{\frac{x}{a}}{y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (let* ((t_1 (* x (/ (/ (* (* b b) 0.5) a) y))))
   (if (<= b -1.05e+170)
     t_1
     (if (<= b -1.02e-234)
       (/ (* x (/ (+ (- b) 1.0) a)) y)
       (if (<= b 1.6e-211) t_1 (/ (/ x a) y))))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * ((((b * b) * 0.5) / a) / y);
	double tmp;
	if (b <= -1.05e+170) {
		tmp = t_1;
	} else if (b <= -1.02e-234) {
		tmp = (x * ((-b + 1.0) / a)) / y;
	} else if (b <= 1.6e-211) {
		tmp = t_1;
	} else {
		tmp = (x / a) / y;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: t_1
    real(8) :: tmp
    t_1 = x * ((((b * b) * 0.5d0) / a) / y)
    if (b <= (-1.05d+170)) then
        tmp = t_1
    else if (b <= (-1.02d-234)) then
        tmp = (x * ((-b + 1.0d0) / a)) / y
    else if (b <= 1.6d-211) then
        tmp = t_1
    else
        tmp = (x / a) / y
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double t_1 = x * ((((b * b) * 0.5) / a) / y);
	double tmp;
	if (b <= -1.05e+170) {
		tmp = t_1;
	} else if (b <= -1.02e-234) {
		tmp = (x * ((-b + 1.0) / a)) / y;
	} else if (b <= 1.6e-211) {
		tmp = t_1;
	} else {
		tmp = (x / a) / y;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	t_1 = x * ((((b * b) * 0.5) / a) / y)
	tmp = 0
	if b <= -1.05e+170:
		tmp = t_1
	elif b <= -1.02e-234:
		tmp = (x * ((-b + 1.0) / a)) / y
	elif b <= 1.6e-211:
		tmp = t_1
	else:
		tmp = (x / a) / y
	return tmp

Julia

function code(x, y, z, t, a, b)
	t_1 = Float64(x * Float64(Float64(Float64(Float64(b * b) * 0.5) / a) / y))
	tmp = 0.0
	if (b <= -1.05e+170)
		tmp = t_1;
	elseif (b <= -1.02e-234)
		tmp = Float64(Float64(x * Float64(Float64(Float64(-b) + 1.0) / a)) / y);
	elseif (b <= 1.6e-211)
		tmp = t_1;
	else
		tmp = Float64(Float64(x / a) / y);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	t_1 = x * ((((b * b) * 0.5) / a) / y);
	tmp = 0.0;
	if (b <= -1.05e+170)
		tmp = t_1;
	elseif (b <= -1.02e-234)
		tmp = (x * ((-b + 1.0) / a)) / y;
	elseif (b <= 1.6e-211)
		tmp = t_1;
	else
		tmp = (x / a) / y;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := Block[{t$95$1 = N[(x * N[(N[(N[(N[(b * b), $MachinePrecision] * 0.5), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision]}, If[LessEqual[b, -1.05e+170], t$95$1, If[LessEqual[b, -1.02e-234], N[(N[(x * N[(N[((-b) + 1.0), $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], If[LessEqual[b, 1.6e-211], t$95$1, N[(N[(x / a), $MachinePrecision] / y), $MachinePrecision]]]]]

Derivation

1. Split input into 3 regimes

2. if b < -1.04999999999999999e170 or -1.01999999999999999e-234 < b < 1.59999999999999993e-211

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + b \cdot \left(\frac{1}{2} \cdot b - 1\right)}{a}}{y} \]
   9. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{b \cdot \left(\frac{1}{2} \cdot b - 1\right) + 1}{a}}{y} \]

      2. *-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(\frac{1}{2} \cdot b - 1\right) \cdot b + 1}{a}}{y} \]

      3. lower-fma.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      4. lower--.f64 N/A

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(\frac{1}{2} \cdot b - 1, b, 1\right)}{a}}{y} \]

      5. lower-*.f64 39.5

         \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   10. Applied rewrites 39.5%

       \[\leadsto x \cdot \frac{\frac{\mathsf{fma}\left(0.5 \cdot b - 1, b, 1\right)}{a}}{y} \]

   11. Taylor expanded in b around inf

       \[\leadsto x \cdot \frac{\frac{\frac{1}{2} \cdot {b}^{2}}{a}}{y} \]
   12. Step-by-step derivation

       1. *-commutative N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       2. lower-*.f64 N/A

          \[\leadsto x \cdot \frac{\frac{{b}^{2} \cdot \frac{1}{2}}{a}}{y} \]

       3. unpow2 N/A

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot \frac{1}{2}}{a}}{y} \]

       4. lower-*.f64 37.2

          \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   13. Applied rewrites 37.2%

       \[\leadsto x \cdot \frac{\frac{\left(b \cdot b\right) \cdot 0.5}{a}}{y} \]

   if -1.04999999999999999e170 < b < -1.01999999999999999e-234

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + -1 \cdot b}{a}}{y} \]
   9. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(\mathsf{neg}\left(b\right)\right)}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(-b\right)}{a}}{y} \]

      3. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

      4. lower-+.f64 32.5

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

   10. Applied rewrites 32.5%

       \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]
   11. Step-by-step derivation

       1. lift-*.f64 N/A

          \[\leadsto x \cdot \color{blue}{\frac{\frac{\left(-b\right) + 1}{a}}{y}} \]

       2. lift-/.f64 N/A

          \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       3. associate-*r/ N/A

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       4. lower-/.f64 N/A

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       5. lower-*.f64 31.7

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{y} \]

   12. Applied rewrites 31.7%

       \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

   if 1.59999999999999993e-211 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

      \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   9. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x}{a \cdot y} \]

      2. lift-*.f64 31.2

         \[\leadsto \frac{x}{a \cdot y} \]

   10. Applied rewrites 31.2%

       \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   11. Step-by-step derivation

       1. lift-*.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       2. lift-/.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       3. associate-/r* N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       4. lower-/.f64 N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       5. lower-/.f64 30.6

          \[\leadsto \frac{\frac{x}{a}}{y} \]

   12. Applied rewrites 30.6%

       \[\leadsto \frac{\frac{x}{a}}{y} \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Alternative 14: 35.2% accurate, 2.4× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;b \leq 0.0105:\\ \;\;\;\;\frac{x \cdot \frac{\left(-b\right) + 1}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;\frac{x}{a \cdot y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (if (<= b 0.0105) (/ (* x (/ (+ (- b) 1.0) a)) y) (/ x (* a y))))

C

double code(double x, double y, double z, double t, double a, double b) {
	double tmp;
	if (b <= 0.0105) {
		tmp = (x * ((-b + 1.0) / a)) / y;
	} else {
		tmp = x / (a * y);
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: tmp
    if (b <= 0.0105d0) then
        tmp = (x * ((-b + 1.0d0) / a)) / y
    else
        tmp = x / (a * y)
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double tmp;
	if (b <= 0.0105) {
		tmp = (x * ((-b + 1.0) / a)) / y;
	} else {
		tmp = x / (a * y);
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	tmp = 0
	if b <= 0.0105:
		tmp = (x * ((-b + 1.0) / a)) / y
	else:
		tmp = x / (a * y)
	return tmp

Julia

function code(x, y, z, t, a, b)
	tmp = 0.0
	if (b <= 0.0105)
		tmp = Float64(Float64(x * Float64(Float64(Float64(-b) + 1.0) / a)) / y);
	else
		tmp = Float64(x / Float64(a * y));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	tmp = 0.0;
	if (b <= 0.0105)
		tmp = (x * ((-b + 1.0) / a)) / y;
	else
		tmp = x / (a * y);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := If[LessEqual[b, 0.0105], N[(N[(x * N[(N[((-b) + 1.0), $MachinePrecision] / a), $MachinePrecision]), $MachinePrecision] / y), $MachinePrecision], N[(x / N[(a * y), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if b < 0.0105000000000000007

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + -1 \cdot b}{a}}{y} \]
   9. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(\mathsf{neg}\left(b\right)\right)}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(-b\right)}{a}}{y} \]

      3. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

      4. lower-+.f64 32.5

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

   10. Applied rewrites 32.5%

       \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]
   11. Step-by-step derivation

       1. lift-*.f64 N/A

          \[\leadsto x \cdot \color{blue}{\frac{\frac{\left(-b\right) + 1}{a}}{y}} \]

       2. lift-/.f64 N/A

          \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       3. associate-*r/ N/A

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       4. lower-/.f64 N/A

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

       5. lower-*.f64 31.7

          \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{y} \]

   12. Applied rewrites 31.7%

       \[\leadsto \frac{x \cdot \frac{\left(-b\right) + 1}{a}}{\color{blue}{y}} \]

   if 0.0105000000000000007 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

      \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   9. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x}{a \cdot y} \]

      2. lift-*.f64 31.2

         \[\leadsto \frac{x}{a \cdot y} \]

   10. Applied rewrites 31.2%

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

Alternative 15: 34.7% accurate, 2.4× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;b \leq -6 \cdot 10^{+87}:\\ \;\;\;\;x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y}\\ \mathbf{else}:\\ \;\;\;\;\frac{\frac{x}{a}}{y}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t a b)
 :precision binary64
 (if (<= b -6e+87) (* x (/ (/ (+ (- b) 1.0) a) y)) (/ (/ x a) y)))

C

double code(double x, double y, double z, double t, double a, double b) {
	double tmp;
	if (b <= -6e+87) {
		tmp = x * (((-b + 1.0) / a) / y);
	} else {
		tmp = (x / a) / y;
	}
	return tmp;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

    interface fmax
        module procedure fmax88
        module procedure fmax44
        module procedure fmax84
        module procedure fmax48
    end interface
    interface fmin
        module procedure fmin88
        module procedure fmin44
        module procedure fmin84
        module procedure fmin48
    end interface
contains
    real(8) function fmax88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(4) function fmax44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, max(x, y), y /= y), x /= x)
    end function
    real(8) function fmax84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmax48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
    end function
    real(8) function fmin88(x, y) result (res)
        real(8), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(4) function fmin44(x, y) result (res)
        real(4), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(y, merge(x, min(x, y), y /= y), x /= x)
    end function
    real(8) function fmin84(x, y) result(res)
        real(8), intent (in) :: x
        real(4), intent (in) :: y
        res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
    end function
    real(8) function fmin48(x, y) result(res)
        real(4), intent (in) :: x
        real(8), intent (in) :: y
        res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
    end function
end module

real(8) function code(x, y, z, t, a, b)
use fmin_fmax_functions
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8), intent (in) :: a
    real(8), intent (in) :: b
    real(8) :: tmp
    if (b <= (-6d+87)) then
        tmp = x * (((-b + 1.0d0) / a) / y)
    else
        tmp = (x / a) / y
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t, double a, double b) {
	double tmp;
	if (b <= -6e+87) {
		tmp = x * (((-b + 1.0) / a) / y);
	} else {
		tmp = (x / a) / y;
	}
	return tmp;
}

Python

def code(x, y, z, t, a, b):
	tmp = 0
	if b <= -6e+87:
		tmp = x * (((-b + 1.0) / a) / y)
	else:
		tmp = (x / a) / y
	return tmp

Julia

function code(x, y, z, t, a, b)
	tmp = 0.0
	if (b <= -6e+87)
		tmp = Float64(x * Float64(Float64(Float64(Float64(-b) + 1.0) / a) / y));
	else
		tmp = Float64(Float64(x / a) / y);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t, a, b)
	tmp = 0.0;
	if (b <= -6e+87)
		tmp = x * (((-b + 1.0) / a) / y);
	else
		tmp = (x / a) / y;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := If[LessEqual[b, -6e+87], N[(x * N[(N[(N[((-b) + 1.0), $MachinePrecision] / a), $MachinePrecision] / y), $MachinePrecision]), $MachinePrecision], N[(N[(x / a), $MachinePrecision] / y), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if b < -5.9999999999999998e87

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in y around 0

      \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{\mathsf{neg}\left(b\right)}}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

      3. lower-exp.f64 58.6

         \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   7. Applied rewrites 58.6%

      \[\leadsto x \cdot \frac{\frac{e^{-b}}{a}}{y} \]

   8. Taylor expanded in b around 0

      \[\leadsto x \cdot \frac{\frac{1 + -1 \cdot b}{a}}{y} \]
   9. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(\mathsf{neg}\left(b\right)\right)}{a}}{y} \]

      2. lift-neg.f64 N/A

         \[\leadsto x \cdot \frac{\frac{1 + \left(-b\right)}{a}}{y} \]

      3. +-commutative N/A

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

      4. lower-+.f64 32.5

         \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

   10. Applied rewrites 32.5%

       \[\leadsto x \cdot \frac{\frac{\left(-b\right) + 1}{a}}{y} \]

   if -5.9999999999999998e87 < b

   1. Initial program 98.2%

      \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

   2. Taylor expanded in t around 0

      \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
   3. Step-by-step derivation

      1. associate-/l* N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      2. lower-*.f64 N/A

         \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

      3. lower-/.f64 N/A

         \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

   4. Applied rewrites 80.3%

      \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

   5. Taylor expanded in b around 0

      \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
   6. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

      2. *-commutative N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      3. lower-*.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      4. lower-pow.f64 N/A

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

      5. lower-*.f64 54.0

         \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   7. Applied rewrites 54.0%

      \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

   8. Taylor expanded in y around 0

      \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   9. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{x}{a \cdot y} \]

      2. lift-*.f64 31.2

         \[\leadsto \frac{x}{a \cdot y} \]

   10. Applied rewrites 31.2%

       \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
   11. Step-by-step derivation

       1. lift-*.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       2. lift-/.f64 N/A

          \[\leadsto \frac{x}{a \cdot y} \]

       3. associate-/r* N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       4. lower-/.f64 N/A

          \[\leadsto \frac{\frac{x}{a}}{y} \]

       5. lower-/.f64 30.6

          \[\leadsto \frac{\frac{x}{a}}{y} \]

   12. Applied rewrites 30.6%

       \[\leadsto \frac{\frac{x}{a}}{y} \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 16: 31.2% accurate, 5.5× speedup

Math

\[\begin{array}{l} \\ \frac{\frac{x}{a}}{y} \end{array} \]

FPCore

(FPCore (x y z t a b) :precision binary64 (/ (/ x a) y))

C

double code(double x, double y, double z, double t, double a, double b) {
	return (x / a) / y;
}

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

Java

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

Python

def code(x, y, z, t, a, b):
	return (x / a) / y

Julia

function code(x, y, z, t, a, b)
	return Float64(Float64(x / a) / y)
end

MATLAB

function tmp = code(x, y, z, t, a, b)
	tmp = (x / a) / y;
end

Wolfram

code[x_, y_, z_, t_, a_, b_] := N[(N[(x / a), $MachinePrecision] / y), $MachinePrecision]

Derivation

1. Initial program 98.2%

   \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

2. Taylor expanded in t around 0

   \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
3. Step-by-step derivation

   1. associate-/l* N/A

      \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

   2. lower-*.f64 N/A

      \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

   3. lower-/.f64 N/A

      \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

4. Applied rewrites 80.3%

   \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

5. Taylor expanded in b around 0

   \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
6. Step-by-step derivation

   1. lower-/.f64 N/A

      \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

   2. *-commutative N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   3. lower-*.f64 N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   4. lower-pow.f64 N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   5. lower-*.f64 54.0

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

7. Applied rewrites 54.0%

   \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

8. Taylor expanded in y around 0

   \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
9. Step-by-step derivation

   1. lower-/.f64 N/A

      \[\leadsto \frac{x}{a \cdot y} \]

   2. lift-*.f64 31.2

      \[\leadsto \frac{x}{a \cdot y} \]

10. Applied rewrites 31.2%

    \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
11. Step-by-step derivation

    1. lift-*.f64 N/A

       \[\leadsto \frac{x}{a \cdot y} \]

    2. lift-/.f64 N/A

       \[\leadsto \frac{x}{a \cdot y} \]

    3. associate-/r* N/A

       \[\leadsto \frac{\frac{x}{a}}{y} \]

    4. lower-/.f64 N/A

       \[\leadsto \frac{\frac{x}{a}}{y} \]

    5. lower-/.f64 30.6

       \[\leadsto \frac{\frac{x}{a}}{y} \]

12. Applied rewrites 30.6%

    \[\leadsto \frac{\frac{x}{a}}{y} \]
13. Add Preprocessing

Alternative 17: 30.6% accurate, 5.8× speedup

Math

\[\begin{array}{l} \\ \frac{x}{a \cdot y} \end{array} \]

FPCore

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

C

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

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

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

code[x_, y_, z_, t_, a_, b_] := N[(x / N[(a * y), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 98.2%

   \[\frac{x \cdot e^{\left(y \cdot \log z + \left(t - 1\right) \cdot \log a\right) - b}}{y} \]

2. Taylor expanded in t around 0

   \[\leadsto \color{blue}{\frac{x \cdot e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]
3. Step-by-step derivation

   1. associate-/l* N/A

      \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

   2. lower-*.f64 N/A

      \[\leadsto x \cdot \color{blue}{\frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{y}} \]

   3. lower-/.f64 N/A

      \[\leadsto x \cdot \frac{e^{\left(-1 \cdot \log a + y \cdot \log z\right) - b}}{\color{blue}{y}} \]

4. Applied rewrites 80.3%

   \[\leadsto \color{blue}{x \cdot \frac{\frac{1}{a} \cdot e^{\log z \cdot y - b}}{y}} \]

5. Taylor expanded in b around 0

   \[\leadsto \frac{x \cdot {z}^{y}}{\color{blue}{a \cdot y}} \]
6. Step-by-step derivation

   1. lower-/.f64 N/A

      \[\leadsto \frac{x \cdot {z}^{y}}{a \cdot \color{blue}{y}} \]

   2. *-commutative N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   3. lower-*.f64 N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   4. lower-pow.f64 N/A

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

   5. lower-*.f64 54.0

      \[\leadsto \frac{{z}^{y} \cdot x}{a \cdot y} \]

7. Applied rewrites 54.0%

   \[\leadsto \frac{{z}^{y} \cdot x}{\color{blue}{a \cdot y}} \]

8. Taylor expanded in y around 0

   \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
9. Step-by-step derivation

   1. lower-/.f64 N/A

      \[\leadsto \frac{x}{a \cdot y} \]

   2. lift-*.f64 31.2

      \[\leadsto \frac{x}{a \cdot y} \]

10. Applied rewrites 31.2%

    \[\leadsto \frac{x}{a \cdot \color{blue}{y}} \]
11. Add Preprocessing

Reproduce

herbie shell --seed 2025142 
(FPCore (x y z t a b)
  :name "Numeric.SpecFunctions:incompleteBetaWorker from math-functions-0.1.5.2, A"
  :precision binary64
  (/ (* x (exp (- (+ (* y (log z)) (* (- t 1.0) (log a))) b))) y))