expq2 (section 3.11)

Percentage Accurate: 37.4% → 100.0%

Time: 5.2s

Alternatives: 7

Speedup: 2.1×

Specification

\[710 > x\]

Math

\[\begin{array}{l} \\ \frac{e^{x}}{e^{x} - 1} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (/ (exp x) (- (exp x) 1.0)))

C

double code(double x) {
	return exp(x) / (exp(x) - 1.0);
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = exp(x) / (exp(x) - 1.0d0)
end function

Java

public static double code(double x) {
	return Math.exp(x) / (Math.exp(x) - 1.0);
}

Python

def code(x):
	return math.exp(x) / (math.exp(x) - 1.0)

Julia

function code(x)
	return Float64(exp(x) / Float64(exp(x) - 1.0))
end

MATLAB

function tmp = code(x)
	tmp = exp(x) / (exp(x) - 1.0);
end

Wolfram

code[x_] := N[(N[Exp[x], $MachinePrecision] / N[(N[Exp[x], $MachinePrecision] - 1.0), $MachinePrecision]), $MachinePrecision]

Initial Program: 37.4% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{e^{x}}{e^{x} - 1} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (/ (exp x) (- (exp x) 1.0)))

C

double code(double x) {
	return exp(x) / (exp(x) - 1.0);
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = exp(x) / (exp(x) - 1.0d0)
end function

Java

public static double code(double x) {
	return Math.exp(x) / (Math.exp(x) - 1.0);
}

Python

def code(x):
	return math.exp(x) / (math.exp(x) - 1.0)

Julia

function code(x)
	return Float64(exp(x) / Float64(exp(x) - 1.0))
end

MATLAB

function tmp = code(x)
	tmp = exp(x) / (exp(x) - 1.0);
end

Wolfram

code[x_] := N[(N[Exp[x], $MachinePrecision] / N[(N[Exp[x], $MachinePrecision] - 1.0), $MachinePrecision]), $MachinePrecision]

Alternative 1: 100.0% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{e^{x}}{\mathsf{expm1}\left(x\right)} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (/ (exp x) (expm1 x)))

C

double code(double x) {
	return exp(x) / expm1(x);
}

Java

public static double code(double x) {
	return Math.exp(x) / Math.expm1(x);
}

Python

def code(x):
	return math.exp(x) / math.expm1(x)

Julia

function code(x)
	return Float64(exp(x) / expm1(x))
end

Wolfram

code[x_] := N[(N[Exp[x], $MachinePrecision] / N[(Exp[x] - 1), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 36.9%

   \[\frac{e^{x}}{e^{x} - 1} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift--.f64 N/A

      \[\leadsto \frac{e^{x}}{\color{blue}{e^{x} - 1}} \]

   2. lift-exp.f64 N/A

      \[\leadsto \frac{e^{x}}{\color{blue}{e^{x}} - 1} \]

   3. lower-expm1.f64 100.0

      \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]

4. Applied rewrites 100.0%

   \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]
5. Add Preprocessing

Alternative 2: 99.5% accurate, 1.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq -3.8:\\ \;\;\;\;\frac{e^{x}}{\left(1 + x\right) - 1}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, {x}^{-1} - -0.5\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= x -3.8)
   (/ (exp x) (- (+ 1.0 x) 1.0))
   (fma
    (fma (* x x) -0.001388888888888889 0.08333333333333333)
    x
    (- (pow x -1.0) -0.5))))

C

double code(double x) {
	double tmp;
	if (x <= -3.8) {
		tmp = exp(x) / ((1.0 + x) - 1.0);
	} else {
		tmp = fma(fma((x * x), -0.001388888888888889, 0.08333333333333333), x, (pow(x, -1.0) - -0.5));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (x <= -3.8)
		tmp = Float64(exp(x) / Float64(Float64(1.0 + x) - 1.0));
	else
		tmp = fma(fma(Float64(x * x), -0.001388888888888889, 0.08333333333333333), x, Float64((x ^ -1.0) - -0.5));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[x, -3.8], N[(N[Exp[x], $MachinePrecision] / N[(N[(1.0 + x), $MachinePrecision] - 1.0), $MachinePrecision]), $MachinePrecision], N[(N[(N[(x * x), $MachinePrecision] * -0.001388888888888889 + 0.08333333333333333), $MachinePrecision] * x + N[(N[Power[x, -1.0], $MachinePrecision] - -0.5), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < -3.7999999999999998

   1. Initial program 100.0%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \frac{e^{x}}{\color{blue}{\left(1 + x\right)} - 1} \]
   4. Step-by-step derivation

      1. lower-+.f64 100.0

         \[\leadsto \frac{e^{x}}{\color{blue}{\left(1 + x\right)} - 1} \]

   5. Applied rewrites 100.0%

      \[\leadsto \frac{e^{x}}{\color{blue}{\left(1 + x\right)} - 1} \]

   if -3.7999999999999998 < x

   1. Initial program 6.1%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing
   3. Step-by-step derivation

      1. lift--.f64 N/A

         \[\leadsto \frac{e^{x}}{\color{blue}{e^{x} - 1}} \]

      2. lift-exp.f64 N/A

         \[\leadsto \frac{e^{x}}{\color{blue}{e^{x}} - 1} \]

      3. lower-expm1.f64 100.0

         \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]

   4. Applied rewrites 100.0%

      \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]

   5. Taylor expanded in x around 0

      \[\leadsto \color{blue}{\frac{1 + x \cdot \left(\frac{1}{2} + x \cdot \left(\frac{1}{12} + \frac{-1}{720} \cdot {x}^{2}\right)\right)}{x}} \]

   7. Applied rewrites 99.4%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, \frac{1}{x} - -0.5\right)} \]
3. Recombined 2 regimes into one program.

4. Final simplification 99.6%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq -3.8:\\ \;\;\;\;\frac{e^{x}}{\left(1 + x\right) - 1}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, {x}^{-1} - -0.5\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 3: 83.6% accurate, 1.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq -6.2:\\ \;\;\;\;\sqrt{{\left(x \cdot x\right)}^{-1}}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, {x}^{-1} - -0.5\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= x -6.2)
   (sqrt (pow (* x x) -1.0))
   (fma
    (fma (* x x) -0.001388888888888889 0.08333333333333333)
    x
    (- (pow x -1.0) -0.5))))

C

double code(double x) {
	double tmp;
	if (x <= -6.2) {
		tmp = sqrt(pow((x * x), -1.0));
	} else {
		tmp = fma(fma((x * x), -0.001388888888888889, 0.08333333333333333), x, (pow(x, -1.0) - -0.5));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (x <= -6.2)
		tmp = sqrt((Float64(x * x) ^ -1.0));
	else
		tmp = fma(fma(Float64(x * x), -0.001388888888888889, 0.08333333333333333), x, Float64((x ^ -1.0) - -0.5));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[x, -6.2], N[Sqrt[N[Power[N[(x * x), $MachinePrecision], -1.0], $MachinePrecision]], $MachinePrecision], N[(N[(N[(x * x), $MachinePrecision] * -0.001388888888888889 + 0.08333333333333333), $MachinePrecision] * x + N[(N[Power[x, -1.0], $MachinePrecision] - -0.5), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < -6.20000000000000018

   1. Initial program 100.0%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{\frac{1}{x}} \]
   4. Step-by-step derivation

      1. lower-/.f64 5.4

         \[\leadsto \color{blue}{\frac{1}{x}} \]

   5. Applied rewrites 5.4%

      \[\leadsto \color{blue}{\frac{1}{x}} \]
   6. Step-by-step derivation

   7. Applied rewrites 53.1%

      \[\leadsto \sqrt{{x}^{-2}} \]
   8. Step-by-step derivation

   9. Applied rewrites 53.1%

      \[\leadsto \sqrt{\frac{1}{x \cdot x}} \]

   if -6.20000000000000018 < x

   1. Initial program 6.1%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing
   3. Step-by-step derivation

      1. lift--.f64 N/A

         \[\leadsto \frac{e^{x}}{\color{blue}{e^{x} - 1}} \]

      2. lift-exp.f64 N/A

         \[\leadsto \frac{e^{x}}{\color{blue}{e^{x}} - 1} \]

      3. lower-expm1.f64 100.0

         \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]

   4. Applied rewrites 100.0%

      \[\leadsto \frac{e^{x}}{\color{blue}{\mathsf{expm1}\left(x\right)}} \]

   5. Taylor expanded in x around 0

      \[\leadsto \color{blue}{\frac{1 + x \cdot \left(\frac{1}{2} + x \cdot \left(\frac{1}{12} + \frac{-1}{720} \cdot {x}^{2}\right)\right)}{x}} \]

   7. Applied rewrites 99.4%

      \[\leadsto \color{blue}{\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, \frac{1}{x} - -0.5\right)} \]
3. Recombined 2 regimes into one program.

4. Final simplification 84.2%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq -6.2:\\ \;\;\;\;\sqrt{{\left(x \cdot x\right)}^{-1}}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, -0.001388888888888889, 0.08333333333333333\right), x, {x}^{-1} - -0.5\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 4: 83.5% accurate, 1.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq -700:\\ \;\;\;\;\sqrt{{\left(x \cdot x\right)}^{-1}}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(0.08333333333333333, x, {x}^{-1} - -0.5\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= x -700.0)
   (sqrt (pow (* x x) -1.0))
   (fma 0.08333333333333333 x (- (pow x -1.0) -0.5))))

C

double code(double x) {
	double tmp;
	if (x <= -700.0) {
		tmp = sqrt(pow((x * x), -1.0));
	} else {
		tmp = fma(0.08333333333333333, x, (pow(x, -1.0) - -0.5));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (x <= -700.0)
		tmp = sqrt((Float64(x * x) ^ -1.0));
	else
		tmp = fma(0.08333333333333333, x, Float64((x ^ -1.0) - -0.5));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[x, -700.0], N[Sqrt[N[Power[N[(x * x), $MachinePrecision], -1.0], $MachinePrecision]], $MachinePrecision], N[(0.08333333333333333 * x + N[(N[Power[x, -1.0], $MachinePrecision] - -0.5), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < -700

   1. Initial program 100.0%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{\frac{1}{x}} \]
   4. Step-by-step derivation

      1. lower-/.f64 5.4

         \[\leadsto \color{blue}{\frac{1}{x}} \]

   5. Applied rewrites 5.4%

      \[\leadsto \color{blue}{\frac{1}{x}} \]
   6. Step-by-step derivation

   7. Applied rewrites 53.1%

      \[\leadsto \sqrt{{x}^{-2}} \]
   8. Step-by-step derivation

   9. Applied rewrites 53.1%

      \[\leadsto \sqrt{\frac{1}{x \cdot x}} \]

   if -700 < x

   1. Initial program 6.1%

      \[\frac{e^{x}}{e^{x} - 1} \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{\frac{1 + x \cdot \left(\frac{1}{2} + \frac{1}{12} \cdot x\right)}{x}} \]
   4. Step-by-step derivation

      1. distribute-lft-in N/A

         \[\leadsto \frac{1 + \color{blue}{\left(x \cdot \frac{1}{2} + x \cdot \left(\frac{1}{12} \cdot x\right)\right)}}{x} \]

      2. *-commutative N/A

         \[\leadsto \frac{1 + \left(\color{blue}{\frac{1}{2} \cdot x} + x \cdot \left(\frac{1}{12} \cdot x\right)\right)}{x} \]

      3. associate-+r+ N/A

         \[\leadsto \frac{\color{blue}{\left(1 + \frac{1}{2} \cdot x\right) + x \cdot \left(\frac{1}{12} \cdot x\right)}}{x} \]

      4. div-add N/A

         \[\leadsto \color{blue}{\frac{1 + \frac{1}{2} \cdot x}{x} + \frac{x \cdot \left(\frac{1}{12} \cdot x\right)}{x}} \]

      5. *-commutative N/A

         \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \frac{\color{blue}{\left(\frac{1}{12} \cdot x\right) \cdot x}}{x} \]

      6. associate-/l* N/A

         \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \color{blue}{\left(\frac{1}{12} \cdot x\right) \cdot \frac{x}{x}} \]

      7. *-inverses N/A

         \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \left(\frac{1}{12} \cdot x\right) \cdot \color{blue}{1} \]

      8. *-rgt-identity N/A

         \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \color{blue}{\frac{1}{12} \cdot x} \]

      9. +-commutative N/A

         \[\leadsto \color{blue}{\frac{1}{12} \cdot x + \frac{1 + \frac{1}{2} \cdot x}{x}} \]

      10. lower-fma.f64 N/A

          \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{1}{12}, x, \frac{1 + \frac{1}{2} \cdot x}{x}\right)} \]

      11. fp-cancel-sign-sub-inv N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{\color{blue}{1 - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}}{x}\right) \]

      12. div-sub N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x} - \frac{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}{x}}\right) \]

      13. associate-/l* N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \frac{x}{x}}\right) \]

      14. *-inverses N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \color{blue}{1}\right) \]

      15. metadata-eval N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\frac{-1}{2}} \cdot 1\right) \]

      16. metadata-eval N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\frac{-1}{2}}\right) \]

      17. metadata-eval N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right)}\right) \]

      18. lower--.f64 N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)}\right) \]

      19. lower-/.f64 N/A

          \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x}} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)\right) \]

      20. metadata-eval 99.3

          \[\leadsto \mathsf{fma}\left(0.08333333333333333, x, \frac{1}{x} - \color{blue}{-0.5}\right) \]

   5. Applied rewrites 99.3%

      \[\leadsto \color{blue}{\mathsf{fma}\left(0.08333333333333333, x, \frac{1}{x} - -0.5\right)} \]
3. Recombined 2 regimes into one program.

4. Final simplification 84.1%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq -700:\\ \;\;\;\;\sqrt{{\left(x \cdot x\right)}^{-1}}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(0.08333333333333333, x, {x}^{-1} - -0.5\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 5: 67.2% accurate, 2.1× speedup

Math

\[\begin{array}{l} \\ {x}^{-1} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (pow x -1.0))

C

double code(double x) {
	return pow(x, -1.0);
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = x ** (-1.0d0)
end function

Java

public static double code(double x) {
	return Math.pow(x, -1.0);
}

Python

def code(x):
	return math.pow(x, -1.0)

Julia

function code(x)
	return x ^ -1.0
end

MATLAB

function tmp = code(x)
	tmp = x ^ -1.0;
end

Wolfram

code[x_] := N[Power[x, -1.0], $MachinePrecision]

Derivation

1. Initial program 36.9%

   \[\frac{e^{x}}{e^{x} - 1} \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \color{blue}{\frac{1}{x}} \]
4. Step-by-step derivation

   1. lower-/.f64 67.7

      \[\leadsto \color{blue}{\frac{1}{x}} \]

5. Applied rewrites 67.7%

   \[\leadsto \color{blue}{\frac{1}{x}} \]

6. Final simplification 67.7%

   \[\leadsto {x}^{-1} \]
7. Add Preprocessing

Alternative 6: 3.3% accurate, 35.8× speedup

Math

\[\begin{array}{l} \\ 0.08333333333333333 \cdot x \end{array} \]

FPCore

(FPCore (x) :precision binary64 (* 0.08333333333333333 x))

C

double code(double x) {
	return 0.08333333333333333 * x;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 0.08333333333333333d0 * x
end function

Java

public static double code(double x) {
	return 0.08333333333333333 * x;
}

Python

def code(x):
	return 0.08333333333333333 * x

Julia

function code(x)
	return Float64(0.08333333333333333 * x)
end

MATLAB

function tmp = code(x)
	tmp = 0.08333333333333333 * x;
end

Wolfram

code[x_] := N[(0.08333333333333333 * x), $MachinePrecision]

Derivation

1. Initial program 36.9%

   \[\frac{e^{x}}{e^{x} - 1} \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \color{blue}{\frac{1 + x \cdot \left(\frac{1}{2} + \frac{1}{12} \cdot x\right)}{x}} \]
4. Step-by-step derivation

   1. distribute-lft-in N/A

      \[\leadsto \frac{1 + \color{blue}{\left(x \cdot \frac{1}{2} + x \cdot \left(\frac{1}{12} \cdot x\right)\right)}}{x} \]

   2. *-commutative N/A

      \[\leadsto \frac{1 + \left(\color{blue}{\frac{1}{2} \cdot x} + x \cdot \left(\frac{1}{12} \cdot x\right)\right)}{x} \]

   3. associate-+r+ N/A

      \[\leadsto \frac{\color{blue}{\left(1 + \frac{1}{2} \cdot x\right) + x \cdot \left(\frac{1}{12} \cdot x\right)}}{x} \]

   4. div-add N/A

      \[\leadsto \color{blue}{\frac{1 + \frac{1}{2} \cdot x}{x} + \frac{x \cdot \left(\frac{1}{12} \cdot x\right)}{x}} \]

   5. *-commutative N/A

      \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \frac{\color{blue}{\left(\frac{1}{12} \cdot x\right) \cdot x}}{x} \]

   6. associate-/l* N/A

      \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \color{blue}{\left(\frac{1}{12} \cdot x\right) \cdot \frac{x}{x}} \]

   7. *-inverses N/A

      \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \left(\frac{1}{12} \cdot x\right) \cdot \color{blue}{1} \]

   8. *-rgt-identity N/A

      \[\leadsto \frac{1 + \frac{1}{2} \cdot x}{x} + \color{blue}{\frac{1}{12} \cdot x} \]

   9. +-commutative N/A

      \[\leadsto \color{blue}{\frac{1}{12} \cdot x + \frac{1 + \frac{1}{2} \cdot x}{x}} \]

   10. lower-fma.f64 N/A

       \[\leadsto \color{blue}{\mathsf{fma}\left(\frac{1}{12}, x, \frac{1 + \frac{1}{2} \cdot x}{x}\right)} \]

   11. fp-cancel-sign-sub-inv N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{\color{blue}{1 - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}}{x}\right) \]

   12. div-sub N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x} - \frac{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}{x}}\right) \]

   13. associate-/l* N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \frac{x}{x}}\right) \]

   14. *-inverses N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \color{blue}{1}\right) \]

   15. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\frac{-1}{2}} \cdot 1\right) \]

   16. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\frac{-1}{2}}\right) \]

   17. metadata-eval N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right)}\right) \]

   18. lower--.f64 N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)}\right) \]

   19. lower-/.f64 N/A

       \[\leadsto \mathsf{fma}\left(\frac{1}{12}, x, \color{blue}{\frac{1}{x}} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)\right) \]

   20. metadata-eval 67.4

       \[\leadsto \mathsf{fma}\left(0.08333333333333333, x, \frac{1}{x} - \color{blue}{-0.5}\right) \]

5. Applied rewrites 67.4%

   \[\leadsto \color{blue}{\mathsf{fma}\left(0.08333333333333333, x, \frac{1}{x} - -0.5\right)} \]

6. Taylor expanded in x around inf

   \[\leadsto \frac{1}{12} \cdot \color{blue}{x} \]
7. Step-by-step derivation

8. Applied rewrites 3.3%

   \[\leadsto 0.08333333333333333 \cdot \color{blue}{x} \]
9. Add Preprocessing

Alternative 7: 3.2% accurate, 215.0× speedup

Math

\[\begin{array}{l} \\ 0.5 \end{array} \]

FPCore

(FPCore (x) :precision binary64 0.5)

C

double code(double x) {
	return 0.5;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 0.5d0
end function

Java

public static double code(double x) {
	return 0.5;
}

Python

def code(x):
	return 0.5

Julia

function code(x)
	return 0.5
end

MATLAB

function tmp = code(x)
	tmp = 0.5;
end

Wolfram

code[x_] := 0.5

Derivation

1. Initial program 36.9%

   \[\frac{e^{x}}{e^{x} - 1} \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \color{blue}{\frac{1 + \frac{1}{2} \cdot x}{x}} \]
4. Step-by-step derivation

   1. fp-cancel-sign-sub-inv N/A

      \[\leadsto \frac{\color{blue}{1 - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}}{x} \]

   2. div-sub N/A

      \[\leadsto \color{blue}{\frac{1}{x} - \frac{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot x}{x}} \]

   3. associate-/l* N/A

      \[\leadsto \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \frac{x}{x}} \]

   4. *-inverses N/A

      \[\leadsto \frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \cdot \color{blue}{1} \]

   5. metadata-eval N/A

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

   6. metadata-eval N/A

      \[\leadsto \frac{1}{x} - \color{blue}{\frac{-1}{2}} \]

   7. metadata-eval N/A

      \[\leadsto \frac{1}{x} - \color{blue}{\left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} \]

   8. lower--.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{x} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right)} \]

   9. lower-/.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{x}} - \left(\mathsf{neg}\left(\frac{1}{2}\right)\right) \]

   10. metadata-eval 67.5

       \[\leadsto \frac{1}{x} - \color{blue}{-0.5} \]

5. Applied rewrites 67.5%

   \[\leadsto \color{blue}{\frac{1}{x} - -0.5} \]

6. Taylor expanded in x around inf

   \[\leadsto \frac{1}{2} \]
7. Step-by-step derivation

8. Applied rewrites 3.3%

   \[\leadsto 0.5 \]
9. Add Preprocessing

Developer Target 1: 100.0% accurate, 1.9× speedup

Math

\[\begin{array}{l} \\ \frac{-1}{\mathsf{expm1}\left(-x\right)} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (/ (- 1.0) (expm1 (- x))))

C

double code(double x) {
	return -1.0 / expm1(-x);
}

Java

public static double code(double x) {
	return -1.0 / Math.expm1(-x);
}

Python

def code(x):
	return -1.0 / math.expm1(-x)

Julia

function code(x)
	return Float64(Float64(-1.0) / expm1(Float64(-x)))
end

Wolfram

code[x_] := N[((-1.0) / N[(Exp[(-x)] - 1), $MachinePrecision]), $MachinePrecision]

Reproduce

herbie shell --seed 2024340 
(FPCore (x)
  :name "expq2 (section 3.11)"
  :precision binary64
  :pre (> 710.0 x)

  :alt
  (! :herbie-platform default (/ (- 1) (expm1 (- x))))

  (/ (exp x) (- (exp x) 1.0)))