Example 2 from Robby

Percentage Accurate: 99.8% → 99.8%

Time: 17.3s

Alternatives: 8

Speedup: N/A×

Specification

Math

\[\begin{array}{l} t_1 := \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\\ \left|\left(ew \cdot \cos t\right) \cdot \cos t\_1 - \left(eh \cdot \sin t\right) \cdot \sin t\_1\right| \end{array} \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (let* ((t_1 (atan (/ (* (- eh) (tan t)) ew))))
   (fabs (- (* (* ew (cos t)) (cos t_1)) (* (* eh (sin t)) (sin t_1))))))

C

double code(double eh, double ew, double t) {
	double t_1 = atan(((-eh * tan(t)) / ew));
	return fabs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))));
}

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(eh, ew, t)
use fmin_fmax_functions
    real(8), intent (in) :: eh
    real(8), intent (in) :: ew
    real(8), intent (in) :: t
    real(8) :: t_1
    t_1 = atan(((-eh * tan(t)) / ew))
    code = abs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))))
end function

Java

public static double code(double eh, double ew, double t) {
	double t_1 = Math.atan(((-eh * Math.tan(t)) / ew));
	return Math.abs((((ew * Math.cos(t)) * Math.cos(t_1)) - ((eh * Math.sin(t)) * Math.sin(t_1))));
}

Python

def code(eh, ew, t):
	t_1 = math.atan(((-eh * math.tan(t)) / ew))
	return math.fabs((((ew * math.cos(t)) * math.cos(t_1)) - ((eh * math.sin(t)) * math.sin(t_1))))

Julia

function code(eh, ew, t)
	t_1 = atan(Float64(Float64(Float64(-eh) * tan(t)) / ew))
	return abs(Float64(Float64(Float64(ew * cos(t)) * cos(t_1)) - Float64(Float64(eh * sin(t)) * sin(t_1))))
end

MATLAB

function tmp = code(eh, ew, t)
	t_1 = atan(((-eh * tan(t)) / ew));
	tmp = abs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))));
end

Wolfram

code[eh_, ew_, t_] := Block[{t$95$1 = N[ArcTan[N[(N[((-eh) * N[Tan[t], $MachinePrecision]), $MachinePrecision] / ew), $MachinePrecision]], $MachinePrecision]}, N[Abs[N[(N[(N[(ew * N[Cos[t], $MachinePrecision]), $MachinePrecision] * N[Cos[t$95$1], $MachinePrecision]), $MachinePrecision] - N[(N[(eh * N[Sin[t], $MachinePrecision]), $MachinePrecision] * N[Sin[t$95$1], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]]

Initial Program: 99.8% accurate, 1.0× speedup

Math

\[\begin{array}{l} t_1 := \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\\ \left|\left(ew \cdot \cos t\right) \cdot \cos t\_1 - \left(eh \cdot \sin t\right) \cdot \sin t\_1\right| \end{array} \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (let* ((t_1 (atan (/ (* (- eh) (tan t)) ew))))
   (fabs (- (* (* ew (cos t)) (cos t_1)) (* (* eh (sin t)) (sin t_1))))))

C

double code(double eh, double ew, double t) {
	double t_1 = atan(((-eh * tan(t)) / ew));
	return fabs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))));
}

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(eh, ew, t)
use fmin_fmax_functions
    real(8), intent (in) :: eh
    real(8), intent (in) :: ew
    real(8), intent (in) :: t
    real(8) :: t_1
    t_1 = atan(((-eh * tan(t)) / ew))
    code = abs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))))
end function

Java

public static double code(double eh, double ew, double t) {
	double t_1 = Math.atan(((-eh * Math.tan(t)) / ew));
	return Math.abs((((ew * Math.cos(t)) * Math.cos(t_1)) - ((eh * Math.sin(t)) * Math.sin(t_1))));
}

Python

def code(eh, ew, t):
	t_1 = math.atan(((-eh * math.tan(t)) / ew))
	return math.fabs((((ew * math.cos(t)) * math.cos(t_1)) - ((eh * math.sin(t)) * math.sin(t_1))))

Julia

function code(eh, ew, t)
	t_1 = atan(Float64(Float64(Float64(-eh) * tan(t)) / ew))
	return abs(Float64(Float64(Float64(ew * cos(t)) * cos(t_1)) - Float64(Float64(eh * sin(t)) * sin(t_1))))
end

MATLAB

function tmp = code(eh, ew, t)
	t_1 = atan(((-eh * tan(t)) / ew));
	tmp = abs((((ew * cos(t)) * cos(t_1)) - ((eh * sin(t)) * sin(t_1))));
end

Wolfram

code[eh_, ew_, t_] := Block[{t$95$1 = N[ArcTan[N[(N[((-eh) * N[Tan[t], $MachinePrecision]), $MachinePrecision] / ew), $MachinePrecision]], $MachinePrecision]}, N[Abs[N[(N[(N[(ew * N[Cos[t], $MachinePrecision]), $MachinePrecision] * N[Cos[t$95$1], $MachinePrecision]), $MachinePrecision] - N[(N[(eh * N[Sin[t], $MachinePrecision]), $MachinePrecision] * N[Sin[t$95$1], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]]

Alternative 1: 99.8% accurate, 1.2× speedup

Math

\[\begin{array}{l} t_1 := \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right)\\ \left|\mathsf{fma}\left(\frac{\cos t}{\cosh t\_1}, ew, \tanh t\_1 \cdot \left(\sin t \cdot eh\right)\right)\right| \end{array} \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (let* ((t_1 (asinh (* (/ eh ew) (tan t)))))
   (fabs (fma (/ (cos t) (cosh t_1)) ew (* (tanh t_1) (* (sin t) eh))))))

C

double code(double eh, double ew, double t) {
	double t_1 = asinh(((eh / ew) * tan(t)));
	return fabs(fma((cos(t) / cosh(t_1)), ew, (tanh(t_1) * (sin(t) * eh))));
}

Julia

function code(eh, ew, t)
	t_1 = asinh(Float64(Float64(eh / ew) * tan(t)))
	return abs(fma(Float64(cos(t) / cosh(t_1)), ew, Float64(tanh(t_1) * Float64(sin(t) * eh))))
end

Wolfram

code[eh_, ew_, t_] := Block[{t$95$1 = N[ArcSinh[N[(N[(eh / ew), $MachinePrecision] * N[Tan[t], $MachinePrecision]), $MachinePrecision]], $MachinePrecision]}, N[Abs[N[(N[(N[Cos[t], $MachinePrecision] / N[Cosh[t$95$1], $MachinePrecision]), $MachinePrecision] * ew + N[(N[Tanh[t$95$1], $MachinePrecision] * N[(N[Sin[t], $MachinePrecision] * eh), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]]

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]
2. Step-by-step derivation

3. Applied rewrites 99.8%

   \[\leadsto \left|\color{blue}{\frac{\cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)}\right| \]

5. Applied rewrites 99.8%

   \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right)}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)}\right| \]
6. Add Preprocessing

Alternative 2: 98.5% accurate, 1.8× speedup

Math

\[\left|\frac{\cos t \cdot ew}{1} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)\right| \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (fabs
  (-
   (/ (* (cos t) ew) 1.0)
   (* (tanh (asinh (* (/ (tan t) ew) (- eh)))) (* (sin t) eh)))))

C

double code(double eh, double ew, double t) {
	return fabs((((cos(t) * ew) / 1.0) - (tanh(asinh(((tan(t) / ew) * -eh))) * (sin(t) * eh))));
}

Python

def code(eh, ew, t):
	return math.fabs((((math.cos(t) * ew) / 1.0) - (math.tanh(math.asinh(((math.tan(t) / ew) * -eh))) * (math.sin(t) * eh))))

Julia

function code(eh, ew, t)
	return abs(Float64(Float64(Float64(cos(t) * ew) / 1.0) - Float64(tanh(asinh(Float64(Float64(tan(t) / ew) * Float64(-eh)))) * Float64(sin(t) * eh))))
end

MATLAB

function tmp = code(eh, ew, t)
	tmp = abs((((cos(t) * ew) / 1.0) - (tanh(asinh(((tan(t) / ew) * -eh))) * (sin(t) * eh))));
end

Wolfram

code[eh_, ew_, t_] := N[Abs[N[(N[(N[(N[Cos[t], $MachinePrecision] * ew), $MachinePrecision] / 1.0), $MachinePrecision] - N[(N[Tanh[N[ArcSinh[N[(N[(N[Tan[t], $MachinePrecision] / ew), $MachinePrecision] * (-eh)), $MachinePrecision]], $MachinePrecision]], $MachinePrecision] * N[(N[Sin[t], $MachinePrecision] * eh), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]
2. Step-by-step derivation

3. Applied rewrites 99.8%

   \[\leadsto \left|\color{blue}{\frac{\cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)}\right| \]

4. Taylor expanded in eh around 0

   \[\leadsto \left|\frac{\cos t \cdot ew}{\color{blue}{1}} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)\right| \]
5. Step-by-step derivation

6. Applied rewrites 98.5%

   \[\leadsto \left|\frac{\cos t \cdot ew}{\color{blue}{1}} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)\right| \]
7. Add Preprocessing

Alternative 3: 98.5% accurate, 1.8× speedup

Math

\[\left|\mathsf{fma}\left(\cos t, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)\right| \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (fabs
  (fma
   (cos t)
   ew
   (* (* (sin t) (- eh)) (tanh (asinh (* (/ (tan t) ew) (- eh))))))))

C

double code(double eh, double ew, double t) {
	return fabs(fma(cos(t), ew, ((sin(t) * -eh) * tanh(asinh(((tan(t) / ew) * -eh))))));
}

Julia

function code(eh, ew, t)
	return abs(fma(cos(t), ew, Float64(Float64(sin(t) * Float64(-eh)) * tanh(asinh(Float64(Float64(tan(t) / ew) * Float64(-eh)))))))
end

Wolfram

code[eh_, ew_, t_] := N[Abs[N[(N[Cos[t], $MachinePrecision] * ew + N[(N[(N[Sin[t], $MachinePrecision] * (-eh)), $MachinePrecision] * N[Tanh[N[ArcSinh[N[(N[(N[Tan[t], $MachinePrecision] / ew), $MachinePrecision] * (-eh)), $MachinePrecision]], $MachinePrecision]], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

3. Applied rewrites 99.8%

   \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)}\right| \]

4. Taylor expanded in eh around 0

   \[\leadsto \left|\mathsf{fma}\left(\color{blue}{\cos t}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)\right| \]
5. Step-by-step derivation

   1. lower-cos.f64 98.5%

      \[\leadsto \left|\mathsf{fma}\left(\cos t, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)\right| \]

6. Applied rewrites 98.5%

   \[\leadsto \left|\mathsf{fma}\left(\color{blue}{\cos t}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)\right| \]
7. Add Preprocessing

Alternative 4: 90.6% accurate, 1.8× speedup

Math

\[\begin{array}{l} t_1 := \frac{\left(-\left|eh\right|\right) \cdot t}{ew}\\ t_2 := ew \cdot \cos t\\ \mathbf{if}\;\left|eh\right| \leq 1.75 \cdot 10^{-114}:\\ \;\;\;\;\left|t\_2\right|\\ \mathbf{else}:\\ \;\;\;\;\left|\frac{t\_2}{\sqrt{\mathsf{fma}\left(t\_1, t\_1, 1\right)}} - \tanh \sinh^{-1} t\_1 \cdot \left(\sin t \cdot \left|eh\right|\right)\right|\\ \end{array} \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (let* ((t_1 (/ (* (- (fabs eh)) t) ew)) (t_2 (* ew (cos t))))
   (if (<= (fabs eh) 1.75e-114)
     (fabs t_2)
     (fabs
      (-
       (/ t_2 (sqrt (fma t_1 t_1 1.0)))
       (* (tanh (asinh t_1)) (* (sin t) (fabs eh))))))))

C

double code(double eh, double ew, double t) {
	double t_1 = (-fabs(eh) * t) / ew;
	double t_2 = ew * cos(t);
	double tmp;
	if (fabs(eh) <= 1.75e-114) {
		tmp = fabs(t_2);
	} else {
		tmp = fabs(((t_2 / sqrt(fma(t_1, t_1, 1.0))) - (tanh(asinh(t_1)) * (sin(t) * fabs(eh)))));
	}
	return tmp;
}

Julia

function code(eh, ew, t)
	t_1 = Float64(Float64(Float64(-abs(eh)) * t) / ew)
	t_2 = Float64(ew * cos(t))
	tmp = 0.0
	if (abs(eh) <= 1.75e-114)
		tmp = abs(t_2);
	else
		tmp = abs(Float64(Float64(t_2 / sqrt(fma(t_1, t_1, 1.0))) - Float64(tanh(asinh(t_1)) * Float64(sin(t) * abs(eh)))));
	end
	return tmp
end

Wolfram

code[eh_, ew_, t_] := Block[{t$95$1 = N[(N[((-N[Abs[eh], $MachinePrecision]) * t), $MachinePrecision] / ew), $MachinePrecision]}, Block[{t$95$2 = N[(ew * N[Cos[t], $MachinePrecision]), $MachinePrecision]}, If[LessEqual[N[Abs[eh], $MachinePrecision], 1.75e-114], N[Abs[t$95$2], $MachinePrecision], N[Abs[N[(N[(t$95$2 / N[Sqrt[N[(t$95$1 * t$95$1 + 1.0), $MachinePrecision]], $MachinePrecision]), $MachinePrecision] - N[(N[Tanh[N[ArcSinh[t$95$1], $MachinePrecision]], $MachinePrecision] * N[(N[Sin[t], $MachinePrecision] * N[Abs[eh], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]]]]

Derivation

1. Split input into 2 regimes

2. if eh < 1.75e-114

   1. Initial program 99.8%

      \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

   3. Applied rewrites 99.8%

      \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)}\right| \]

   4. Taylor expanded in eh around 0

      \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]
   5. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \left|ew \cdot \color{blue}{\cos t}\right| \]

      2. lower-cos.f64 61.6%

         \[\leadsto \left|ew \cdot \cos t\right| \]

   6. Applied rewrites 61.6%

      \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]

   if 1.75e-114 < eh

   1. Initial program 99.8%

      \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

   2. Taylor expanded in t around 0

      \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\color{blue}{-1 \cdot \left(eh \cdot t\right)}}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]
   3. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \color{blue}{\left(eh \cdot t\right)}}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

      2. lower-*.f64 89.4%

         \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot \color{blue}{t}\right)}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

   4. Applied rewrites 89.4%

      \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\color{blue}{-1 \cdot \left(eh \cdot t\right)}}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

   5. Taylor expanded in t around 0

      \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot t\right)}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\color{blue}{-1 \cdot \left(eh \cdot t\right)}}{ew}\right)\right| \]
   6. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot t\right)}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{-1 \cdot \color{blue}{\left(eh \cdot t\right)}}{ew}\right)\right| \]

      2. lower-*.f64 89.4%

         \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot t\right)}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot \color{blue}{t}\right)}{ew}\right)\right| \]

   7. Applied rewrites 89.4%

      \[\leadsto \left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{-1 \cdot \left(eh \cdot t\right)}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\color{blue}{-1 \cdot \left(eh \cdot t\right)}}{ew}\right)\right| \]
   8. Step-by-step derivation

   9. Applied rewrites 89.1%

      \[\leadsto \left|\color{blue}{\frac{ew \cdot \cos t}{\sqrt{\mathsf{fma}\left(\frac{\left(-eh\right) \cdot t}{ew}, \frac{\left(-eh\right) \cdot t}{ew}, 1\right)}} - \tanh \sinh^{-1} \left(\frac{\left(-eh\right) \cdot t}{ew}\right) \cdot \left(\sin t \cdot eh\right)}\right| \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 5: 90.3% accurate, 2.1× speedup

Math

\[\begin{array}{l} t_1 := \frac{\left|eh\right| \cdot t}{ew}\\ \mathbf{if}\;\left|eh\right| \leq 1.75 \cdot 10^{-114}:\\ \;\;\;\;\left|ew \cdot \cos t\right|\\ \mathbf{else}:\\ \;\;\;\;\left|\mathsf{fma}\left(\frac{\cos t}{\cosh t\_1}, ew, \tanh t\_1 \cdot \left(\sin t \cdot \left|eh\right|\right)\right)\right|\\ \end{array} \]

FPCore

(FPCore (eh ew t)
 :precision binary64
 (let* ((t_1 (/ (* (fabs eh) t) ew)))
   (if (<= (fabs eh) 1.75e-114)
     (fabs (* ew (cos t)))
     (fabs
      (fma (/ (cos t) (cosh t_1)) ew (* (tanh t_1) (* (sin t) (fabs eh))))))))

C

double code(double eh, double ew, double t) {
	double t_1 = (fabs(eh) * t) / ew;
	double tmp;
	if (fabs(eh) <= 1.75e-114) {
		tmp = fabs((ew * cos(t)));
	} else {
		tmp = fabs(fma((cos(t) / cosh(t_1)), ew, (tanh(t_1) * (sin(t) * fabs(eh)))));
	}
	return tmp;
}

Julia

function code(eh, ew, t)
	t_1 = Float64(Float64(abs(eh) * t) / ew)
	tmp = 0.0
	if (abs(eh) <= 1.75e-114)
		tmp = abs(Float64(ew * cos(t)));
	else
		tmp = abs(fma(Float64(cos(t) / cosh(t_1)), ew, Float64(tanh(t_1) * Float64(sin(t) * abs(eh)))));
	end
	return tmp
end

Wolfram

code[eh_, ew_, t_] := Block[{t$95$1 = N[(N[(N[Abs[eh], $MachinePrecision] * t), $MachinePrecision] / ew), $MachinePrecision]}, If[LessEqual[N[Abs[eh], $MachinePrecision], 1.75e-114], N[Abs[N[(ew * N[Cos[t], $MachinePrecision]), $MachinePrecision]], $MachinePrecision], N[Abs[N[(N[(N[Cos[t], $MachinePrecision] / N[Cosh[t$95$1], $MachinePrecision]), $MachinePrecision] * ew + N[(N[Tanh[t$95$1], $MachinePrecision] * N[(N[Sin[t], $MachinePrecision] * N[Abs[eh], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]]]

Derivation

1. Split input into 2 regimes

2. if eh < 1.75e-114

   1. Initial program 99.8%

      \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

   3. Applied rewrites 99.8%

      \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)}\right| \]

   4. Taylor expanded in eh around 0

      \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]
   5. Step-by-step derivation

      1. lower-*.f64 N/A

         \[\leadsto \left|ew \cdot \color{blue}{\cos t}\right| \]

      2. lower-cos.f64 61.6%

         \[\leadsto \left|ew \cdot \cos t\right| \]

   6. Applied rewrites 61.6%

      \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]

   if 1.75e-114 < eh

   1. Initial program 99.8%

      \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]
   2. Step-by-step derivation

   3. Applied rewrites 99.8%

      \[\leadsto \left|\color{blue}{\frac{\cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)} - \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right) \cdot \left(\sin t \cdot eh\right)}\right| \]

   5. Applied rewrites 99.8%

      \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right)}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)}\right| \]

   6. Taylor expanded in t around 0

      \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \color{blue}{\left(\frac{eh \cdot t}{ew}\right)}}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]
   7. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{\color{blue}{ew}}\right)}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]

      2. lower-*.f64 88.6%

         \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{ew}\right)}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]

   8. Applied rewrites 88.6%

      \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \color{blue}{\left(\frac{eh \cdot t}{ew}\right)}}, ew, \tanh \sinh^{-1} \left(\frac{eh}{ew} \cdot \tan t\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]

   9. Taylor expanded in t around 0

      \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{ew}\right)}, ew, \tanh \color{blue}{\left(\frac{eh \cdot t}{ew}\right)} \cdot \left(\sin t \cdot eh\right)\right)\right| \]
   10. Step-by-step derivation

       1. lower-/.f64 N/A

          \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{ew}\right)}, ew, \tanh \left(\frac{eh \cdot t}{\color{blue}{ew}}\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]

       2. lower-*.f64 88.8%

          \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{ew}\right)}, ew, \tanh \left(\frac{eh \cdot t}{ew}\right) \cdot \left(\sin t \cdot eh\right)\right)\right| \]

   11. Applied rewrites 88.8%

       \[\leadsto \left|\mathsf{fma}\left(\frac{\cos t}{\cosh \left(\frac{eh \cdot t}{ew}\right)}, ew, \tanh \color{blue}{\left(\frac{eh \cdot t}{ew}\right)} \cdot \left(\sin t \cdot eh\right)\right)\right| \]
3. Recombined 2 regimes into one program.
4. Add Preprocessing

Alternative 6: 61.6% accurate, 6.7× speedup

Math

\[\left|ew \cdot \cos t\right| \]

FPCore

(FPCore (eh ew t) :precision binary64 (fabs (* ew (cos t))))

C

double code(double eh, double ew, double t) {
	return fabs((ew * cos(t)));
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

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

real(8) function code(eh, ew, t)
use fmin_fmax_functions
    real(8), intent (in) :: eh
    real(8), intent (in) :: ew
    real(8), intent (in) :: t
    code = abs((ew * cos(t)))
end function

Java

public static double code(double eh, double ew, double t) {
	return Math.abs((ew * Math.cos(t)));
}

Python

def code(eh, ew, t):
	return math.fabs((ew * math.cos(t)))

Julia

function code(eh, ew, t)
	return abs(Float64(ew * cos(t)))
end

MATLAB

function tmp = code(eh, ew, t)
	tmp = abs((ew * cos(t)));
end

Wolfram

code[eh_, ew_, t_] := N[Abs[N[(ew * N[Cos[t], $MachinePrecision]), $MachinePrecision]], $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

3. Applied rewrites 99.8%

   \[\leadsto \left|\color{blue}{\mathsf{fma}\left(\frac{\cos t}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}, ew, \left(\sin t \cdot \left(-eh\right)\right) \cdot \tanh \sinh^{-1} \left(\frac{\tan t}{ew} \cdot \left(-eh\right)\right)\right)}\right| \]

4. Taylor expanded in eh around 0

   \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]
5. Step-by-step derivation

   1. lower-*.f64 N/A

      \[\leadsto \left|ew \cdot \color{blue}{\cos t}\right| \]

   2. lower-cos.f64 61.6%

      \[\leadsto \left|ew \cdot \cos t\right| \]

6. Applied rewrites 61.6%

   \[\leadsto \left|\color{blue}{ew \cdot \cos t}\right| \]
7. Add Preprocessing

Alternative 7: 42.1% accurate, 78.2× speedup

Math

\[\left|-ew\right| \]

FPCore

(FPCore (eh ew t) :precision binary64 (fabs (- ew)))

C

double code(double eh, double ew, double t) {
	return fabs(-ew);
}

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(eh, ew, t)
use fmin_fmax_functions
    real(8), intent (in) :: eh
    real(8), intent (in) :: ew
    real(8), intent (in) :: t
    code = abs(-ew)
end function

Java

public static double code(double eh, double ew, double t) {
	return Math.abs(-ew);
}

Python

def code(eh, ew, t):
	return math.fabs(-ew)

Julia

function code(eh, ew, t)
	return abs(Float64(-ew))
end

MATLAB

function tmp = code(eh, ew, t)
	tmp = abs(-ew);
end

Wolfram

code[eh_, ew_, t_] := N[Abs[(-ew)], $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

3. Applied rewrites 39.3%

   \[\leadsto \color{blue}{\sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}}} \]

4. Taylor expanded in t around 0

   \[\leadsto \sqrt{\color{blue}{-1 \cdot ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]
5. Step-by-step derivation

   1. lower-*.f64 19.2%

      \[\leadsto \sqrt{-1 \cdot \color{blue}{ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]

6. Applied rewrites 19.2%

   \[\leadsto \sqrt{\color{blue}{-1 \cdot ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]

7. Taylor expanded in t around 0

   \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{\color{blue}{-1 \cdot ew}} \]
8. Step-by-step derivation

   1. lower-*.f64 20.8%

      \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{-1 \cdot \color{blue}{ew}} \]

9. Applied rewrites 20.8%

   \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{\color{blue}{-1 \cdot ew}} \]
10. Step-by-step derivation

    1. lower-unsound-*.f64 N/A

       \[\leadsto \color{blue}{\sqrt{-1 \cdot ew} \cdot \sqrt{-1 \cdot ew}} \]

    2. lower-unsound-sqrt.f64 N/A

       \[\leadsto \color{blue}{\sqrt{-1 \cdot ew}} \cdot \sqrt{-1 \cdot ew} \]

    3. lower-unsound-sqrt.f64 N/A

       \[\leadsto \sqrt{-1 \cdot ew} \cdot \color{blue}{\sqrt{-1 \cdot ew}} \]

    4. sqrt-prod N/A

       \[\leadsto \color{blue}{\sqrt{\left(-1 \cdot ew\right) \cdot \left(-1 \cdot ew\right)}} \]

11. Applied rewrites 42.1%

    \[\leadsto \color{blue}{\left|-ew\right|} \]
12. Add Preprocessing

Alternative 8: 21.8% accurate, 125.6× speedup

Math

\[-ew \]

FPCore

(FPCore (eh ew t) :precision binary64 (- ew))

C

double code(double eh, double ew, double t) {
	return -ew;
}

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(eh, ew, t)
use fmin_fmax_functions
    real(8), intent (in) :: eh
    real(8), intent (in) :: ew
    real(8), intent (in) :: t
    code = -ew
end function

Java

public static double code(double eh, double ew, double t) {
	return -ew;
}

Python

def code(eh, ew, t):
	return -ew

Julia

function code(eh, ew, t)
	return Float64(-ew)
end

MATLAB

function tmp = code(eh, ew, t)
	tmp = -ew;
end

Wolfram

code[eh_, ew_, t_] := (-ew)

Derivation

1. Initial program 99.8%

   \[\left|\left(ew \cdot \cos t\right) \cdot \cos \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right) - \left(eh \cdot \sin t\right) \cdot \sin \tan^{-1} \left(\frac{\left(-eh\right) \cdot \tan t}{ew}\right)\right| \]

3. Applied rewrites 39.3%

   \[\leadsto \color{blue}{\sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}}} \]

4. Taylor expanded in t around 0

   \[\leadsto \sqrt{\color{blue}{-1 \cdot ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]
5. Step-by-step derivation

   1. lower-*.f64 19.2%

      \[\leadsto \sqrt{-1 \cdot \color{blue}{ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]

6. Applied rewrites 19.2%

   \[\leadsto \sqrt{\color{blue}{-1 \cdot ew}} \cdot \sqrt{\frac{\left(-\left(\sin t \cdot eh\right) \cdot \left(eh \cdot \frac{\tan t}{ew}\right)\right) - \cos t \cdot ew}{\cosh \sinh^{-1} \left(\frac{\tan t \cdot eh}{ew}\right)}} \]

7. Taylor expanded in t around 0

   \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{\color{blue}{-1 \cdot ew}} \]
8. Step-by-step derivation

   1. lower-*.f64 20.8%

      \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{-1 \cdot \color{blue}{ew}} \]

9. Applied rewrites 20.8%

   \[\leadsto \sqrt{-1 \cdot ew} \cdot \sqrt{\color{blue}{-1 \cdot ew}} \]
10. Step-by-step derivation

    1. lift-*.f64 N/A

       \[\leadsto \color{blue}{\sqrt{-1 \cdot ew} \cdot \sqrt{-1 \cdot ew}} \]

    2. lift-sqrt.f64 N/A

       \[\leadsto \color{blue}{\sqrt{-1 \cdot ew}} \cdot \sqrt{-1 \cdot ew} \]

    3. lift-sqrt.f64 N/A

       \[\leadsto \sqrt{-1 \cdot ew} \cdot \color{blue}{\sqrt{-1 \cdot ew}} \]

    4. rem-square-sqrt 21.8%

       \[\leadsto \color{blue}{-1 \cdot ew} \]

11. Applied rewrites 21.8%

    \[\leadsto \color{blue}{-ew} \]
12. Add Preprocessing

Reproduce

herbie shell --seed 2025183 
(FPCore (eh ew t)
  :name "Example 2 from Robby"
  :precision binary64
  (fabs (- (* (* ew (cos t)) (cos (atan (/ (* (- eh) (tan t)) ew)))) (* (* eh (sin t)) (sin (atan (/ (* (- eh) (tan t)) ew)))))))