Numeric.AD.Rank1.Halley:findZero from ad-4.2.4

Percentage Accurate: 82.2% → 99.9%

Time: 4.6s

Alternatives: 5

Speedup: 1.5×

Specification

Math

\[\begin{array}{l} \\ x - \frac{\left(y \cdot 2\right) \cdot z}{\left(z \cdot 2\right) \cdot z - y \cdot t} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (- x (/ (* (* y 2.0) z) (- (* (* z 2.0) z) (* y t)))))

C

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

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x - (((y * 2.0d0) * z) / (((z * 2.0d0) * z) - (y * t)))
end function

Java

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

Python

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

Julia

function code(x, y, z, t)
	return Float64(x - Float64(Float64(Float64(y * 2.0) * z) / Float64(Float64(Float64(z * 2.0) * z) - Float64(y * t))))
end

MATLAB

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

Wolfram

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

Initial Program: 82.2% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ x - \frac{\left(y \cdot 2\right) \cdot z}{\left(z \cdot 2\right) \cdot z - y \cdot t} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (- x (/ (* (* y 2.0) z) (- (* (* z 2.0) z) (* y t)))))

C

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

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x - (((y * 2.0d0) * z) / (((z * 2.0d0) * z) - (y * t)))
end function

Java

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

Python

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

Julia

function code(x, y, z, t)
	return Float64(x - Float64(Float64(Float64(y * 2.0) * z) / Float64(Float64(Float64(z * 2.0) * z) - Float64(y * t))))
end

MATLAB

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

Wolfram

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

Alternative 1: 99.9% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ x + \frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (+ x (/ -2.0 (- (/ (/ z 0.5) y) (/ t z)))))

C

double code(double x, double y, double z, double t) {
	return x + (-2.0 / (((z / 0.5) / y) - (t / z)));
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x + ((-2.0d0) / (((z / 0.5d0) / y) - (t / z)))
end function

Java

public static double code(double x, double y, double z, double t) {
	return x + (-2.0 / (((z / 0.5) / y) - (t / z)));
}

Python

def code(x, y, z, t):
	return x + (-2.0 / (((z / 0.5) / y) - (t / z)))

Julia

function code(x, y, z, t)
	return Float64(x + Float64(-2.0 / Float64(Float64(Float64(z / 0.5) / y) - Float64(t / z))))
end

MATLAB

function tmp = code(x, y, z, t)
	tmp = x + (-2.0 / (((z / 0.5) / y) - (t / z)));
end

Wolfram

code[x_, y_, z_, t_] := N[(x + N[(-2.0 / N[(N[(N[(z / 0.5), $MachinePrecision] / y), $MachinePrecision] - N[(t / z), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 84.9%

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

   1. sub-neg 84.9%

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

   2. associate-/l* 93.1%

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

   3. *-commutative 93.1%

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

   4. associate-/l* 93.0%

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

   5. distribute-neg-frac 93.0%

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

   6. metadata-eval 93.0%

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

   7. associate-/l/ 84.9%

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

   8. div-sub 76.3%

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

   9. times-frac 90.6%

      \[\leadsto x + \frac{-2}{\color{blue}{\frac{z \cdot 2}{y} \cdot \frac{z}{z}} - \frac{y \cdot t}{y \cdot z}} \]

   10. *-inverses 90.6%

       \[\leadsto x + \frac{-2}{\frac{z \cdot 2}{y} \cdot \color{blue}{1} - \frac{y \cdot t}{y \cdot z}} \]

   11. *-rgt-identity 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{\frac{z \cdot 2}{y}} - \frac{y \cdot t}{y \cdot z}} \]

   12. *-commutative 90.6%

       \[\leadsto x + \frac{-2}{\frac{\color{blue}{2 \cdot z}}{y} - \frac{y \cdot t}{y \cdot z}} \]

   13. associate-*l/ 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{\frac{2}{y} \cdot z} - \frac{y \cdot t}{y \cdot z}} \]

   14. *-commutative 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{z \cdot \frac{2}{y}} - \frac{y \cdot t}{y \cdot z}} \]

   15. times-frac 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{\frac{y}{y} \cdot \frac{t}{z}}} \]

   16. *-inverses 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{1} \cdot \frac{t}{z}} \]

   17. *-lft-identity 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{\frac{t}{z}}} \]

3. Simplified 99.9%

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

   1. expm1-log1p-u 91.5%

      \[\leadsto x + \color{blue}{\mathsf{expm1}\left(\mathsf{log1p}\left(\frac{-2}{z \cdot \frac{2}{y} - \frac{t}{z}}\right)\right)} \]

   2. expm1-udef 77.9%

      \[\leadsto x + \color{blue}{\left(e^{\mathsf{log1p}\left(\frac{-2}{z \cdot \frac{2}{y} - \frac{t}{z}}\right)} - 1\right)} \]

   3. associate-*r/ 77.9%

      \[\leadsto x + \left(e^{\mathsf{log1p}\left(\frac{-2}{\color{blue}{\frac{z \cdot 2}{y}} - \frac{t}{z}}\right)} - 1\right) \]

   4. metadata-eval 77.9%

      \[\leadsto x + \left(e^{\mathsf{log1p}\left(\frac{-2}{\frac{z \cdot \color{blue}{\frac{1}{0.5}}}{y} - \frac{t}{z}}\right)} - 1\right) \]

   5. div-inv 77.9%

      \[\leadsto x + \left(e^{\mathsf{log1p}\left(\frac{-2}{\frac{\color{blue}{\frac{z}{0.5}}}{y} - \frac{t}{z}}\right)} - 1\right) \]

5. Applied egg-rr 77.9%

   \[\leadsto x + \color{blue}{\left(e^{\mathsf{log1p}\left(\frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}}\right)} - 1\right)} \]
6. Step-by-step derivation

   1. expm1-def 91.4%

      \[\leadsto x + \color{blue}{\mathsf{expm1}\left(\mathsf{log1p}\left(\frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}}\right)\right)} \]

   2. expm1-log1p 99.9%

      \[\leadsto x + \color{blue}{\frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}}} \]

7. Simplified 99.9%

   \[\leadsto x + \color{blue}{\frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}}} \]

8. Final simplification 99.9%

   \[\leadsto x + \frac{-2}{\frac{\frac{z}{0.5}}{y} - \frac{t}{z}} \]

Alternative 2: 99.8% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ x + \frac{-2}{z \cdot \frac{2}{y} - \frac{t}{z}} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (+ x (/ -2.0 (- (* z (/ 2.0 y)) (/ t z)))))

C

double code(double x, double y, double z, double t) {
	return x + (-2.0 / ((z * (2.0 / y)) - (t / z)));
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x + ((-2.0d0) / ((z * (2.0d0 / y)) - (t / z)))
end function

Java

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

Python

def code(x, y, z, t):
	return x + (-2.0 / ((z * (2.0 / y)) - (t / z)))

Julia

function code(x, y, z, t)
	return Float64(x + Float64(-2.0 / Float64(Float64(z * Float64(2.0 / y)) - Float64(t / z))))
end

MATLAB

function tmp = code(x, y, z, t)
	tmp = x + (-2.0 / ((z * (2.0 / y)) - (t / z)));
end

Wolfram

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

Derivation

1. Initial program 84.9%

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

   1. sub-neg 84.9%

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

   2. associate-/l* 93.1%

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

   3. *-commutative 93.1%

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

   4. associate-/l* 93.0%

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

   5. distribute-neg-frac 93.0%

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

   6. metadata-eval 93.0%

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

   7. associate-/l/ 84.9%

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

   8. div-sub 76.3%

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

   9. times-frac 90.6%

      \[\leadsto x + \frac{-2}{\color{blue}{\frac{z \cdot 2}{y} \cdot \frac{z}{z}} - \frac{y \cdot t}{y \cdot z}} \]

   10. *-inverses 90.6%

       \[\leadsto x + \frac{-2}{\frac{z \cdot 2}{y} \cdot \color{blue}{1} - \frac{y \cdot t}{y \cdot z}} \]

   11. *-rgt-identity 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{\frac{z \cdot 2}{y}} - \frac{y \cdot t}{y \cdot z}} \]

   12. *-commutative 90.6%

       \[\leadsto x + \frac{-2}{\frac{\color{blue}{2 \cdot z}}{y} - \frac{y \cdot t}{y \cdot z}} \]

   13. associate-*l/ 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{\frac{2}{y} \cdot z} - \frac{y \cdot t}{y \cdot z}} \]

   14. *-commutative 90.6%

       \[\leadsto x + \frac{-2}{\color{blue}{z \cdot \frac{2}{y}} - \frac{y \cdot t}{y \cdot z}} \]

   15. times-frac 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{\frac{y}{y} \cdot \frac{t}{z}}} \]

   16. *-inverses 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{1} \cdot \frac{t}{z}} \]

   17. *-lft-identity 99.9%

       \[\leadsto x + \frac{-2}{z \cdot \frac{2}{y} - \color{blue}{\frac{t}{z}}} \]

3. Simplified 99.9%

   \[\leadsto \color{blue}{x + \frac{-2}{z \cdot \frac{2}{y} - \frac{t}{z}}} \]

4. Final simplification 99.9%

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

Alternative 3: 88.6% accurate, 1.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;z \leq -8.2 \cdot 10^{+14} \lor \neg \left(z \leq 3.8 \cdot 10^{+75}\right):\\ \;\;\;\;x - \frac{y}{z}\\ \mathbf{else}:\\ \;\;\;\;x - \frac{z}{t \cdot -0.5}\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (if (or (<= z -8.2e+14) (not (<= z 3.8e+75)))
   (- x (/ y z))
   (- x (/ z (* t -0.5)))))

C

double code(double x, double y, double z, double t) {
	double tmp;
	if ((z <= -8.2e+14) || !(z <= 3.8e+75)) {
		tmp = x - (y / z);
	} else {
		tmp = x - (z / (t * -0.5));
	}
	return tmp;
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8) :: tmp
    if ((z <= (-8.2d+14)) .or. (.not. (z <= 3.8d+75))) then
        tmp = x - (y / z)
    else
        tmp = x - (z / (t * (-0.5d0)))
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t) {
	double tmp;
	if ((z <= -8.2e+14) || !(z <= 3.8e+75)) {
		tmp = x - (y / z);
	} else {
		tmp = x - (z / (t * -0.5));
	}
	return tmp;
}

Python

def code(x, y, z, t):
	tmp = 0
	if (z <= -8.2e+14) or not (z <= 3.8e+75):
		tmp = x - (y / z)
	else:
		tmp = x - (z / (t * -0.5))
	return tmp

Julia

function code(x, y, z, t)
	tmp = 0.0
	if ((z <= -8.2e+14) || !(z <= 3.8e+75))
		tmp = Float64(x - Float64(y / z));
	else
		tmp = Float64(x - Float64(z / Float64(t * -0.5)));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t)
	tmp = 0.0;
	if ((z <= -8.2e+14) || ~((z <= 3.8e+75)))
		tmp = x - (y / z);
	else
		tmp = x - (z / (t * -0.5));
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_] := If[Or[LessEqual[z, -8.2e+14], N[Not[LessEqual[z, 3.8e+75]], $MachinePrecision]], N[(x - N[(y / z), $MachinePrecision]), $MachinePrecision], N[(x - N[(z / N[(t * -0.5), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if z < -8.2e14 or 3.8000000000000002e75 < z

   1. Initial program 73.5%

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

      1. associate-/l* 89.2%

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

      2. *-commutative 89.2%

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

      3. associate-*r/ 89.2%

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

      4. div-sub 89.2%

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

      5. *-commutative 89.2%

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

      6. associate-/l* 96.4%

         \[\leadsto x - 2 \cdot \frac{y}{\color{blue}{\frac{z}{\frac{z}{z \cdot 2}}} - \frac{y \cdot t}{z}} \]

      7. associate-/r* 96.4%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\color{blue}{\frac{\frac{z}{z}}{2}}} - \frac{y \cdot t}{z}} \]

      8. *-inverses 96.4%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\frac{\color{blue}{1}}{2}} - \frac{y \cdot t}{z}} \]

      9. metadata-eval 96.4%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\color{blue}{0.5}} - \frac{y \cdot t}{z}} \]

      10. *-commutative 96.4%

          \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{0.5} - \frac{\color{blue}{t \cdot y}}{z}} \]

      11. associate-*l/ 98.2%

          \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{0.5} - \color{blue}{\frac{t}{z} \cdot y}} \]

   3. Simplified 98.2%

      \[\leadsto \color{blue}{x - 2 \cdot \frac{y}{\frac{z}{0.5} - \frac{t}{z} \cdot y}} \]

   4. Taylor expanded in y around 0 90.2%

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

   if -8.2e14 < z < 3.8000000000000002e75

   1. Initial program 93.3%

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

      1. *-commutative 93.3%

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

      2. associate-/l* 96.0%

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

      3. div-sub 96.0%

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

      4. sub-neg 96.0%

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

      5. *-commutative 96.0%

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

      6. associate-*l* 96.0%

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

      7. *-commutative 96.0%

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

      8. times-frac 96.0%

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

      9. metadata-eval 96.0%

         \[\leadsto x - \frac{z}{\color{blue}{1} \cdot \frac{z \cdot z}{y} + \left(-\frac{y \cdot t}{y \cdot 2}\right)} \]

      10. *-lft-identity 96.0%

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

      11. associate-*r/ 96.9%

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

      12. fma-def 96.9%

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

      13. associate-/r* 96.9%

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

      14. distribute-neg-frac 96.9%

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

      15. *-commutative 96.9%

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

      16. associate-/l* 99.9%

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

      17. *-inverses 99.9%

          \[\leadsto x - \frac{z}{\mathsf{fma}\left(z, \frac{z}{y}, \frac{-\frac{t}{\color{blue}{1}}}{2}\right)} \]

      18. /-rgt-identity 99.9%

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

   3. Simplified 99.9%

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

   4. Taylor expanded in z around 0 90.1%

      \[\leadsto x - \frac{z}{\color{blue}{-0.5 \cdot t}} \]
3. Recombined 2 regimes into one program.

4. Final simplification 90.1%

   \[\leadsto \begin{array}{l} \mathbf{if}\;z \leq -8.2 \cdot 10^{+14} \lor \neg \left(z \leq 3.8 \cdot 10^{+75}\right):\\ \;\;\;\;x - \frac{y}{z}\\ \mathbf{else}:\\ \;\;\;\;x - \frac{z}{t \cdot -0.5}\\ \end{array} \]

Alternative 4: 81.1% accurate, 1.9× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;z \leq -9 \cdot 10^{+67} \lor \neg \left(z \leq 1.95 \cdot 10^{+76}\right):\\ \;\;\;\;x - \frac{y}{z}\\ \mathbf{else}:\\ \;\;\;\;x\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (if (or (<= z -9e+67) (not (<= z 1.95e+76))) (- x (/ y z)) x))

C

double code(double x, double y, double z, double t) {
	double tmp;
	if ((z <= -9e+67) || !(z <= 1.95e+76)) {
		tmp = x - (y / z);
	} else {
		tmp = x;
	}
	return tmp;
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8) :: tmp
    if ((z <= (-9d+67)) .or. (.not. (z <= 1.95d+76))) then
        tmp = x - (y / z)
    else
        tmp = x
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t) {
	double tmp;
	if ((z <= -9e+67) || !(z <= 1.95e+76)) {
		tmp = x - (y / z);
	} else {
		tmp = x;
	}
	return tmp;
}

Python

def code(x, y, z, t):
	tmp = 0
	if (z <= -9e+67) or not (z <= 1.95e+76):
		tmp = x - (y / z)
	else:
		tmp = x
	return tmp

Julia

function code(x, y, z, t)
	tmp = 0.0
	if ((z <= -9e+67) || !(z <= 1.95e+76))
		tmp = Float64(x - Float64(y / z));
	else
		tmp = x;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t)
	tmp = 0.0;
	if ((z <= -9e+67) || ~((z <= 1.95e+76)))
		tmp = x - (y / z);
	else
		tmp = x;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_] := If[Or[LessEqual[z, -9e+67], N[Not[LessEqual[z, 1.95e+76]], $MachinePrecision]], N[(x - N[(y / z), $MachinePrecision]), $MachinePrecision], x]

Derivation

1. Split input into 2 regimes

2. if z < -8.9999999999999997e67 or 1.94999999999999995e76 < z

   1. Initial program 72.5%

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

      1. associate-/l* 88.9%

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

      2. *-commutative 88.9%

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

      3. associate-*r/ 88.9%

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

      4. div-sub 88.9%

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

      5. *-commutative 88.9%

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

      6. associate-/l* 97.0%

         \[\leadsto x - 2 \cdot \frac{y}{\color{blue}{\frac{z}{\frac{z}{z \cdot 2}}} - \frac{y \cdot t}{z}} \]

      7. associate-/r* 97.0%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\color{blue}{\frac{\frac{z}{z}}{2}}} - \frac{y \cdot t}{z}} \]

      8. *-inverses 97.0%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\frac{\color{blue}{1}}{2}} - \frac{y \cdot t}{z}} \]

      9. metadata-eval 97.0%

         \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{\color{blue}{0.5}} - \frac{y \cdot t}{z}} \]

      10. *-commutative 97.0%

          \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{0.5} - \frac{\color{blue}{t \cdot y}}{z}} \]

      11. associate-*l/ 99.0%

          \[\leadsto x - 2 \cdot \frac{y}{\frac{z}{0.5} - \color{blue}{\frac{t}{z} \cdot y}} \]

   3. Simplified 99.0%

      \[\leadsto \color{blue}{x - 2 \cdot \frac{y}{\frac{z}{0.5} - \frac{t}{z} \cdot y}} \]

   4. Taylor expanded in y around 0 91.1%

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

   if -8.9999999999999997e67 < z < 1.94999999999999995e76

   1. Initial program 92.5%

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

      1. sub-neg 92.5%

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

      2. associate-/l* 95.6%

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

      3. distribute-neg-frac 95.6%

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

      4. associate-/r/ 95.6%

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

      5. distribute-rgt-neg-in 95.6%

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

      6. metadata-eval 95.6%

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

      7. associate-*l* 95.6%

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

   3. Simplified 95.6%

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

   4. Taylor expanded in z around 0 81.5%

      \[\leadsto x + \frac{y \cdot -2}{\color{blue}{-1 \cdot \left(y \cdot t\right)}} \cdot z \]
   5. Step-by-step derivation

      1. neg-mul-1 81.5%

         \[\leadsto x + \frac{y \cdot -2}{\color{blue}{-y \cdot t}} \cdot z \]

      2. distribute-rgt-neg-in 81.5%

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

   6. Simplified 81.5%

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

   7. Taylor expanded in x around inf 74.9%

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

4. Final simplification 81.0%

   \[\leadsto \begin{array}{l} \mathbf{if}\;z \leq -9 \cdot 10^{+67} \lor \neg \left(z \leq 1.95 \cdot 10^{+76}\right):\\ \;\;\;\;x - \frac{y}{z}\\ \mathbf{else}:\\ \;\;\;\;x\\ \end{array} \]

Alternative 5: 75.1% accurate, 17.0× speedup

Math

\[\begin{array}{l} \\ x \end{array} \]

FPCore

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

C

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

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x
end function

Java

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

Python

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

Julia

function code(x, y, z, t)
	return x
end

MATLAB

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

Wolfram

code[x_, y_, z_, t_] := x

Derivation

1. Initial program 84.9%

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

   1. sub-neg 84.9%

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

   2. associate-/l* 93.1%

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

   3. distribute-neg-frac 93.1%

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

   4. associate-/r/ 93.0%

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

   5. distribute-rgt-neg-in 93.0%

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

   6. metadata-eval 93.0%

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

   7. associate-*l* 93.0%

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

3. Simplified 93.0%

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

4. Taylor expanded in z around 0 61.6%

   \[\leadsto x + \frac{y \cdot -2}{\color{blue}{-1 \cdot \left(y \cdot t\right)}} \cdot z \]
5. Step-by-step derivation

   1. neg-mul-1 61.6%

      \[\leadsto x + \frac{y \cdot -2}{\color{blue}{-y \cdot t}} \cdot z \]

   2. distribute-rgt-neg-in 61.6%

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

6. Simplified 61.6%

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

7. Taylor expanded in x around inf 75.6%

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

8. Final simplification 75.6%

   \[\leadsto x \]

Developer target: 99.9% accurate, 1.3× speedup

Math

\[\begin{array}{l} \\ x - \frac{1}{\frac{z}{y} - \frac{\frac{t}{2}}{z}} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (- x (/ 1.0 (- (/ z y) (/ (/ t 2.0) z)))))

C

double code(double x, double y, double z, double t) {
	return x - (1.0 / ((z / y) - ((t / 2.0) / z)));
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    code = x - (1.0d0 / ((z / y) - ((t / 2.0d0) / z)))
end function

Java

public static double code(double x, double y, double z, double t) {
	return x - (1.0 / ((z / y) - ((t / 2.0) / z)));
}

Python

def code(x, y, z, t):
	return x - (1.0 / ((z / y) - ((t / 2.0) / z)))

Julia

function code(x, y, z, t)
	return Float64(x - Float64(1.0 / Float64(Float64(z / y) - Float64(Float64(t / 2.0) / z))))
end

MATLAB

function tmp = code(x, y, z, t)
	tmp = x - (1.0 / ((z / y) - ((t / 2.0) / z)));
end

Wolfram

code[x_, y_, z_, t_] := N[(x - N[(1.0 / N[(N[(z / y), $MachinePrecision] - N[(N[(t / 2.0), $MachinePrecision] / z), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Reproduce

herbie shell --seed 2023252 
(FPCore (x y z t)
  :name "Numeric.AD.Rank1.Halley:findZero from ad-4.2.4"
  :precision binary64

  :herbie-target
  (- x (/ 1.0 (- (/ z y) (/ (/ t 2.0) z))))

  (- x (/ (* (* y 2.0) z) (- (* (* z 2.0) z) (* y t)))))