Asymptote C

Percentage Accurate: 54.9% → 99.8%

Time: 3.0s

Alternatives: 8

Speedup: 0.7×

Specification

Math

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

FPCore

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

C

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

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

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

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

Java

public static double code(double x) {
	return (x / (x + 1.0)) - ((x + 1.0) / (x - 1.0));
}

Python

def code(x):
	return (x / (x + 1.0)) - ((x + 1.0) / (x - 1.0))

Julia

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

MATLAB

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

Wolfram

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

Initial Program: 54.9% accurate, 1.0× speedup

Math

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

FPCore

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

C

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

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

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

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

Java

public static double code(double x) {
	return (x / (x + 1.0)) - ((x + 1.0) / (x - 1.0));
}

Python

def code(x):
	return (x / (x + 1.0)) - ((x + 1.0) / (x - 1.0))

Julia

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

MATLAB

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

Wolfram

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

Alternative 1: 99.8% accurate, 0.2× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{\frac{\mathsf{fma}\left(\frac{{x}^{-1} - -3}{x}, -1, -1\right)}{x} + -3}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.0)
   (/ (+ (/ (fma (/ (- (pow x -1.0) -3.0) x) -1.0 -1.0) x) -3.0) x)
   (/ (fma 3.0 x 1.0) (- 1.0 (* x x)))))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.0) {
		tmp = ((fma(((pow(x, -1.0) - -3.0) / x), -1.0, -1.0) / x) + -3.0) / x;
	} else {
		tmp = fma(3.0, x, 1.0) / (1.0 - (x * x));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.0)
		tmp = Float64(Float64(Float64(fma(Float64(Float64((x ^ -1.0) - -3.0) / x), -1.0, -1.0) / x) + -3.0) / x);
	else
		tmp = Float64(fma(3.0, x, 1.0) / Float64(1.0 - Float64(x * x)));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.0], N[(N[(N[(N[(N[(N[(N[Power[x, -1.0], $MachinePrecision] - -3.0), $MachinePrecision] / x), $MachinePrecision] * -1.0 + -1.0), $MachinePrecision] / x), $MachinePrecision] + -3.0), $MachinePrecision] / x), $MachinePrecision], N[(N[(3.0 * x + 1.0), $MachinePrecision] / N[(1.0 - N[(x * x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0

   1. Initial program 8.6%

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

   3. Taylor expanded in x around -inf

      \[\leadsto \color{blue}{-1 \cdot \frac{3 + -1 \cdot \frac{-1 \cdot \frac{3 + \frac{1}{x}}{x} - 1}{x}}{x}} \]
   4. Step-by-step derivation

      1. mul-1-neg N/A

         \[\leadsto \mathsf{neg}\left(\frac{3 + -1 \cdot \frac{-1 \cdot \frac{3 + \frac{1}{x}}{x} - 1}{x}}{x}\right) \]

      2. distribute-neg-frac N/A

         \[\leadsto \frac{\mathsf{neg}\left(\left(3 + -1 \cdot \frac{-1 \cdot \frac{3 + \frac{1}{x}}{x} - 1}{x}\right)\right)}{\color{blue}{x}} \]

      3. lower-/.f64 N/A

         \[\leadsto \frac{\mathsf{neg}\left(\left(3 + -1 \cdot \frac{-1 \cdot \frac{3 + \frac{1}{x}}{x} - 1}{x}\right)\right)}{\color{blue}{x}} \]

   5. Applied rewrites 99.3%

      \[\leadsto \color{blue}{\frac{\frac{\mathsf{fma}\left(\frac{{x}^{-1} - -3}{x}, -1, -1\right)}{x} + -3}{x}} \]

   if 0.0 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 98.9%

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

   4. Applied rewrites 98.8%

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

   5. Taylor expanded in x around 0

      \[\leadsto \frac{\color{blue}{1 + 3 \cdot x}}{1 - x \cdot x} \]
   6. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{3 \cdot x + \color{blue}{1}}{1 - x \cdot x} \]

      2. lower-fma.f64 100.0

         \[\leadsto \frac{\mathsf{fma}\left(3, \color{blue}{x}, 1\right)}{1 - x \cdot x} \]

   7. Applied rewrites 100.0%

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

4. Final simplification 99.7%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{\frac{\mathsf{fma}\left(\frac{{x}^{-1} - -3}{x}, -1, -1\right)}{x} + -3}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \]
5. Add Preprocessing

Alternative 2: 99.7% accurate, 0.4× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{\frac{\frac{-3}{x} - 1}{x} - 3}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.0)
   (/ (- (/ (- (/ -3.0 x) 1.0) x) 3.0) x)
   (/ (fma 3.0 x 1.0) (- 1.0 (* x x)))))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.0) {
		tmp = ((((-3.0 / x) - 1.0) / x) - 3.0) / x;
	} else {
		tmp = fma(3.0, x, 1.0) / (1.0 - (x * x));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.0)
		tmp = Float64(Float64(Float64(Float64(Float64(-3.0 / x) - 1.0) / x) - 3.0) / x);
	else
		tmp = Float64(fma(3.0, x, 1.0) / Float64(1.0 - Float64(x * x)));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.0], N[(N[(N[(N[(N[(-3.0 / x), $MachinePrecision] - 1.0), $MachinePrecision] / x), $MachinePrecision] - 3.0), $MachinePrecision] / x), $MachinePrecision], N[(N[(3.0 * x + 1.0), $MachinePrecision] / N[(1.0 - N[(x * x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0

   1. Initial program 8.6%

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

   3. Taylor expanded in x around inf

      \[\leadsto \color{blue}{\frac{-1 \cdot \frac{1 + 3 \cdot \frac{1}{x}}{x} - 3}{x}} \]
   4. Step-by-step derivation

      1. lower-/.f64 N/A

         \[\leadsto \frac{-1 \cdot \frac{1 + 3 \cdot \frac{1}{x}}{x} - 3}{\color{blue}{x}} \]

      2. lower--.f64 N/A

         \[\leadsto \frac{-1 \cdot \frac{1 + 3 \cdot \frac{1}{x}}{x} - 3}{x} \]

      3. associate-*r/ N/A

         \[\leadsto \frac{\frac{-1 \cdot \left(1 + 3 \cdot \frac{1}{x}\right)}{x} - 3}{x} \]

      4. lower-/.f64 N/A

         \[\leadsto \frac{\frac{-1 \cdot \left(1 + 3 \cdot \frac{1}{x}\right)}{x} - 3}{x} \]

      5. mul-1-neg N/A

         \[\leadsto \frac{\frac{\mathsf{neg}\left(\left(1 + 3 \cdot \frac{1}{x}\right)\right)}{x} - 3}{x} \]

      6. +-commutative N/A

         \[\leadsto \frac{\frac{\mathsf{neg}\left(\left(3 \cdot \frac{1}{x} + 1\right)\right)}{x} - 3}{x} \]

      7. distribute-neg-in N/A

         \[\leadsto \frac{\frac{\left(\mathsf{neg}\left(3 \cdot \frac{1}{x}\right)\right) + \left(\mathsf{neg}\left(1\right)\right)}{x} - 3}{x} \]

      8. distribute-lft-neg-out N/A

         \[\leadsto \frac{\frac{\left(\mathsf{neg}\left(3\right)\right) \cdot \frac{1}{x} + \left(\mathsf{neg}\left(1\right)\right)}{x} - 3}{x} \]

      9. metadata-eval N/A

         \[\leadsto \frac{\frac{\left(\mathsf{neg}\left(3\right)\right) \cdot \frac{1}{x} + -1}{x} - 3}{x} \]

      10. lower-+.f64 N/A

          \[\leadsto \frac{\frac{\left(\mathsf{neg}\left(3\right)\right) \cdot \frac{1}{x} + -1}{x} - 3}{x} \]

      11. metadata-eval N/A

          \[\leadsto \frac{\frac{-3 \cdot \frac{1}{x} + -1}{x} - 3}{x} \]

      12. associate-*r/ N/A

          \[\leadsto \frac{\frac{\frac{-3 \cdot 1}{x} + -1}{x} - 3}{x} \]

      13. metadata-eval N/A

          \[\leadsto \frac{\frac{\frac{-3}{x} + -1}{x} - 3}{x} \]

      14. lower-/.f64 99.2

          \[\leadsto \frac{\frac{\frac{-3}{x} + -1}{x} - 3}{x} \]

   5. Applied rewrites 99.2%

      \[\leadsto \color{blue}{\frac{\frac{\frac{-3}{x} + -1}{x} - 3}{x}} \]

   if 0.0 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 98.9%

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

   4. Applied rewrites 98.8%

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

   5. Taylor expanded in x around 0

      \[\leadsto \frac{\color{blue}{1 + 3 \cdot x}}{1 - x \cdot x} \]
   6. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{3 \cdot x + \color{blue}{1}}{1 - x \cdot x} \]

      2. lower-fma.f64 100.0

         \[\leadsto \frac{\mathsf{fma}\left(3, \color{blue}{x}, 1\right)}{1 - x \cdot x} \]

   7. Applied rewrites 100.0%

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

4. Final simplification 99.6%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{\frac{\frac{-3}{x} - 1}{x} - 3}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \]
5. Add Preprocessing

Alternative 3: 99.6% accurate, 0.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{-3 + \frac{-1}{x}}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.0)
   (/ (+ -3.0 (/ -1.0 x)) x)
   (/ (fma 3.0 x 1.0) (- 1.0 (* x x)))))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.0) {
		tmp = (-3.0 + (-1.0 / x)) / x;
	} else {
		tmp = fma(3.0, x, 1.0) / (1.0 - (x * x));
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.0)
		tmp = Float64(Float64(-3.0 + Float64(-1.0 / x)) / x);
	else
		tmp = Float64(fma(3.0, x, 1.0) / Float64(1.0 - Float64(x * x)));
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.0], N[(N[(-3.0 + N[(-1.0 / x), $MachinePrecision]), $MachinePrecision] / x), $MachinePrecision], N[(N[(3.0 * x + 1.0), $MachinePrecision] / N[(1.0 - N[(x * x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0

   1. Initial program 8.6%

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

   4. Applied rewrites 4.3%

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

   5. Taylor expanded in x around 0

      \[\leadsto \frac{\color{blue}{1 + 3 \cdot x}}{1 - x \cdot x} \]
   6. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{3 \cdot x + \color{blue}{1}}{1 - x \cdot x} \]

      2. lower-fma.f64 43.3

         \[\leadsto \frac{\mathsf{fma}\left(3, \color{blue}{x}, 1\right)}{1 - x \cdot x} \]

   7. Applied rewrites 43.3%

      \[\leadsto \frac{\color{blue}{\mathsf{fma}\left(3, x, 1\right)}}{1 - x \cdot x} \]
   8. Step-by-step derivation

      1. lift--.f64 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{1 - x \cdot x}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 - \color{blue}{x \cdot x}} \]

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

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{1 + \left(\mathsf{neg}\left(x\right)\right) \cdot x}} \]

      4. distribute-lft-neg-in N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \color{blue}{\left(\mathsf{neg}\left(x \cdot x\right)\right)}} \]

      5. pow2 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \left(\mathsf{neg}\left(\color{blue}{{x}^{2}}\right)\right)} \]

      6. mul-1-neg N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \color{blue}{-1 \cdot {x}^{2}}} \]

      7. +-commutative N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{-1 \cdot {x}^{2} + 1}} \]

      8. mul-1-neg N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\left(\mathsf{neg}\left({x}^{2}\right)\right)} + 1} \]

      9. pow2 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\left(\mathsf{neg}\left(\color{blue}{x \cdot x}\right)\right) + 1} \]

      10. distribute-lft-neg-in N/A

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\left(\mathsf{neg}\left(x\right)\right) \cdot x} + 1} \]

      11. lower-fma.f64 N/A

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\mathsf{fma}\left(\mathsf{neg}\left(x\right), x, 1\right)}} \]

      12. lower-neg.f64 43.3

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\mathsf{fma}\left(\color{blue}{-x}, x, 1\right)} \]

   9. Applied rewrites 43.3%

      \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\mathsf{fma}\left(-x, x, 1\right)}} \]

   10. Taylor expanded in x around inf

       \[\leadsto \color{blue}{-1 \cdot \frac{3 + \frac{1}{x}}{x}} \]
   11. Step-by-step derivation

       1. associate-*r/ N/A

          \[\leadsto \frac{-1 \cdot \left(3 + \frac{1}{x}\right)}{\color{blue}{x}} \]

       2. mul-1-neg N/A

          \[\leadsto \frac{\mathsf{neg}\left(\left(3 + \frac{1}{x}\right)\right)}{x} \]

       3. lower-/.f64 N/A

          \[\leadsto \frac{\mathsf{neg}\left(\left(3 + \frac{1}{x}\right)\right)}{\color{blue}{x}} \]

       4. mul-1-neg N/A

          \[\leadsto \frac{-1 \cdot \left(3 + \frac{1}{x}\right)}{x} \]

       5. distribute-lft-in N/A

          \[\leadsto \frac{-1 \cdot 3 + -1 \cdot \frac{1}{x}}{x} \]

       6. metadata-eval N/A

          \[\leadsto \frac{-3 + -1 \cdot \frac{1}{x}}{x} \]

       7. metadata-eval N/A

          \[\leadsto \frac{-3 + \frac{1}{-1} \cdot \frac{1}{x}}{x} \]

       8. times-frac N/A

          \[\leadsto \frac{-3 + \frac{1 \cdot 1}{-1 \cdot x}}{x} \]

       9. metadata-eval N/A

          \[\leadsto \frac{-3 + \frac{1}{-1 \cdot x}}{x} \]

       10. metadata-eval N/A

           \[\leadsto \frac{-3 + \frac{\mathsf{neg}\left(-1\right)}{-1 \cdot x}}{x} \]

       11. mul-1-neg N/A

           \[\leadsto \frac{-3 + \frac{\mathsf{neg}\left(-1\right)}{\mathsf{neg}\left(x\right)}}{x} \]

       12. frac-2neg N/A

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

       13. lower-+.f64 N/A

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

       14. lower-/.f64 99.0

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

   12. Applied rewrites 99.0%

       \[\leadsto \color{blue}{\frac{-3 + \frac{-1}{x}}{x}} \]

   if 0.0 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 98.9%

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

   4. Applied rewrites 98.8%

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

   5. Taylor expanded in x around 0

      \[\leadsto \frac{\color{blue}{1 + 3 \cdot x}}{1 - x \cdot x} \]
   6. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{3 \cdot x + \color{blue}{1}}{1 - x \cdot x} \]

      2. lower-fma.f64 100.0

         \[\leadsto \frac{\mathsf{fma}\left(3, \color{blue}{x}, 1\right)}{1 - x \cdot x} \]

   7. Applied rewrites 100.0%

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

4. Final simplification 99.5%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;\frac{-3 + \frac{-1}{x}}{x}\\ \mathbf{else}:\\ \;\;\;\;\frac{\mathsf{fma}\left(3, x, 1\right)}{1 - x \cdot x}\\ \end{array} \]
5. Add Preprocessing

Alternative 4: 98.9% accurate, 0.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;\frac{-3 + \frac{-1}{x}}{x}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.02)
   (/ (+ -3.0 (/ -1.0 x)) x)
   (fma (- x -3.0) x 1.0)))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.02) {
		tmp = (-3.0 + (-1.0 / x)) / x;
	} else {
		tmp = fma((x - -3.0), x, 1.0);
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.02)
		tmp = Float64(Float64(-3.0 + Float64(-1.0 / x)) / x);
	else
		tmp = fma(Float64(x - -3.0), x, 1.0);
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.02], N[(N[(-3.0 + N[(-1.0 / x), $MachinePrecision]), $MachinePrecision] / x), $MachinePrecision], N[(N[(x - -3.0), $MachinePrecision] * x + 1.0), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0200000000000000004

   1. Initial program 10.3%

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

   4. Applied rewrites 6.1%

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

   5. Taylor expanded in x around 0

      \[\leadsto \frac{\color{blue}{1 + 3 \cdot x}}{1 - x \cdot x} \]
   6. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto \frac{3 \cdot x + \color{blue}{1}}{1 - x \cdot x} \]

      2. lower-fma.f64 45.0

         \[\leadsto \frac{\mathsf{fma}\left(3, \color{blue}{x}, 1\right)}{1 - x \cdot x} \]

   7. Applied rewrites 45.0%

      \[\leadsto \frac{\color{blue}{\mathsf{fma}\left(3, x, 1\right)}}{1 - x \cdot x} \]
   8. Step-by-step derivation

      1. lift--.f64 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{1 - x \cdot x}} \]

      2. lift-*.f64 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 - \color{blue}{x \cdot x}} \]

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

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{1 + \left(\mathsf{neg}\left(x\right)\right) \cdot x}} \]

      4. distribute-lft-neg-in N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \color{blue}{\left(\mathsf{neg}\left(x \cdot x\right)\right)}} \]

      5. pow2 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \left(\mathsf{neg}\left(\color{blue}{{x}^{2}}\right)\right)} \]

      6. mul-1-neg N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{1 + \color{blue}{-1 \cdot {x}^{2}}} \]

      7. +-commutative N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{-1 \cdot {x}^{2} + 1}} \]

      8. mul-1-neg N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\left(\mathsf{neg}\left({x}^{2}\right)\right)} + 1} \]

      9. pow2 N/A

         \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\left(\mathsf{neg}\left(\color{blue}{x \cdot x}\right)\right) + 1} \]

      10. distribute-lft-neg-in N/A

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\left(\mathsf{neg}\left(x\right)\right) \cdot x} + 1} \]

      11. lower-fma.f64 N/A

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\mathsf{fma}\left(\mathsf{neg}\left(x\right), x, 1\right)}} \]

      12. lower-neg.f64 45.0

          \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\mathsf{fma}\left(\color{blue}{-x}, x, 1\right)} \]

   9. Applied rewrites 45.0%

      \[\leadsto \frac{\mathsf{fma}\left(3, x, 1\right)}{\color{blue}{\mathsf{fma}\left(-x, x, 1\right)}} \]

   10. Taylor expanded in x around inf

       \[\leadsto \color{blue}{-1 \cdot \frac{3 + \frac{1}{x}}{x}} \]
   11. Step-by-step derivation

       1. associate-*r/ N/A

          \[\leadsto \frac{-1 \cdot \left(3 + \frac{1}{x}\right)}{\color{blue}{x}} \]

       2. mul-1-neg N/A

          \[\leadsto \frac{\mathsf{neg}\left(\left(3 + \frac{1}{x}\right)\right)}{x} \]

       3. lower-/.f64 N/A

          \[\leadsto \frac{\mathsf{neg}\left(\left(3 + \frac{1}{x}\right)\right)}{\color{blue}{x}} \]

       4. mul-1-neg N/A

          \[\leadsto \frac{-1 \cdot \left(3 + \frac{1}{x}\right)}{x} \]

       5. distribute-lft-in N/A

          \[\leadsto \frac{-1 \cdot 3 + -1 \cdot \frac{1}{x}}{x} \]

       6. metadata-eval N/A

          \[\leadsto \frac{-3 + -1 \cdot \frac{1}{x}}{x} \]

       7. metadata-eval N/A

          \[\leadsto \frac{-3 + \frac{1}{-1} \cdot \frac{1}{x}}{x} \]

       8. times-frac N/A

          \[\leadsto \frac{-3 + \frac{1 \cdot 1}{-1 \cdot x}}{x} \]

       9. metadata-eval N/A

          \[\leadsto \frac{-3 + \frac{1}{-1 \cdot x}}{x} \]

       10. metadata-eval N/A

           \[\leadsto \frac{-3 + \frac{\mathsf{neg}\left(-1\right)}{-1 \cdot x}}{x} \]

       11. mul-1-neg N/A

           \[\leadsto \frac{-3 + \frac{\mathsf{neg}\left(-1\right)}{\mathsf{neg}\left(x\right)}}{x} \]

       12. frac-2neg N/A

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

       13. lower-+.f64 N/A

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

       14. lower-/.f64 98.2

           \[\leadsto \frac{-3 + \frac{-1}{x}}{x} \]

   12. Applied rewrites 98.2%

       \[\leadsto \color{blue}{\frac{-3 + \frac{-1}{x}}{x}} \]

   if 0.0200000000000000004 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 100.0%

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

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{1 + x \cdot \left(3 + x\right)} \]
   4. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \left(3 + x\right) + \color{blue}{1} \]

      2. *-commutative N/A

         \[\leadsto \left(3 + x\right) \cdot x + 1 \]

      3. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3 + x, \color{blue}{x}, 1\right) \]

      4. +-commutative N/A

         \[\leadsto \mathsf{fma}\left(x + 3, x, 1\right) \]

      5. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x + 3 \cdot 1, x, 1\right) \]

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

         \[\leadsto \mathsf{fma}\left(x - \left(\mathsf{neg}\left(3\right)\right) \cdot 1, x, 1\right) \]

      7. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3 \cdot 1, x, 1\right) \]

      8. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

      9. lower--.f64 99.2

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

   5. Applied rewrites 99.2%

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

4. Final simplification 98.7%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;\frac{-3 + \frac{-1}{x}}{x}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 5: 98.5% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;\frac{-3}{x}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.02)
   (/ -3.0 x)
   (fma (- x -3.0) x 1.0)))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.02) {
		tmp = -3.0 / x;
	} else {
		tmp = fma((x - -3.0), x, 1.0);
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.02)
		tmp = Float64(-3.0 / x);
	else
		tmp = fma(Float64(x - -3.0), x, 1.0);
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.02], N[(-3.0 / x), $MachinePrecision], N[(N[(x - -3.0), $MachinePrecision] * x + 1.0), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0200000000000000004

   1. Initial program 10.3%

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

   3. Taylor expanded in x around inf

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

      1. lower-/.f64 97.1

         \[\leadsto \frac{-3}{\color{blue}{x}} \]

   5. Applied rewrites 97.1%

      \[\leadsto \color{blue}{\frac{-3}{x}} \]

   if 0.0200000000000000004 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 100.0%

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

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{1 + x \cdot \left(3 + x\right)} \]
   4. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \left(3 + x\right) + \color{blue}{1} \]

      2. *-commutative N/A

         \[\leadsto \left(3 + x\right) \cdot x + 1 \]

      3. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3 + x, \color{blue}{x}, 1\right) \]

      4. +-commutative N/A

         \[\leadsto \mathsf{fma}\left(x + 3, x, 1\right) \]

      5. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x + 3 \cdot 1, x, 1\right) \]

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

         \[\leadsto \mathsf{fma}\left(x - \left(\mathsf{neg}\left(3\right)\right) \cdot 1, x, 1\right) \]

      7. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3 \cdot 1, x, 1\right) \]

      8. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

      9. lower--.f64 99.2

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

   5. Applied rewrites 99.2%

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

4. Final simplification 98.2%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;\frac{-3}{x}\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 6: 52.9% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;1 - 1\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.0)
   (- 1.0 1.0)
   (fma (- x -3.0) x 1.0)))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.0) {
		tmp = 1.0 - 1.0;
	} else {
		tmp = fma((x - -3.0), x, 1.0);
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.0)
		tmp = Float64(1.0 - 1.0);
	else
		tmp = fma(Float64(x - -3.0), x, 1.0);
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.0], N[(1.0 - 1.0), $MachinePrecision], N[(N[(x - -3.0), $MachinePrecision] * x + 1.0), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0

   1. Initial program 8.6%

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

   3. Taylor expanded in x around inf

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

   5. Applied rewrites 6.7%

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

   6. Taylor expanded in x around inf

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

   8. Applied rewrites 5.9%

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

   if 0.0 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 98.9%

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

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{1 + x \cdot \left(3 + x\right)} \]
   4. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto x \cdot \left(3 + x\right) + \color{blue}{1} \]

      2. *-commutative N/A

         \[\leadsto \left(3 + x\right) \cdot x + 1 \]

      3. lower-fma.f64 N/A

         \[\leadsto \mathsf{fma}\left(3 + x, \color{blue}{x}, 1\right) \]

      4. +-commutative N/A

         \[\leadsto \mathsf{fma}\left(x + 3, x, 1\right) \]

      5. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x + 3 \cdot 1, x, 1\right) \]

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

         \[\leadsto \mathsf{fma}\left(x - \left(\mathsf{neg}\left(3\right)\right) \cdot 1, x, 1\right) \]

      7. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3 \cdot 1, x, 1\right) \]

      8. metadata-eval N/A

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

      9. lower--.f64 96.5

         \[\leadsto \mathsf{fma}\left(x - -3, x, 1\right) \]

   5. Applied rewrites 96.5%

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

4. Final simplification 52.2%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0:\\ \;\;\;\;1 - 1\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(x - -3, x, 1\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 7: 52.7% accurate, 0.7× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;1 - 1\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(3, x, 1\right)\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= (- (/ x (- x -1.0)) (/ (- x -1.0) (- x 1.0))) 0.02)
   (- 1.0 1.0)
   (fma 3.0 x 1.0)))

C

double code(double x) {
	double tmp;
	if (((x / (x - -1.0)) - ((x - -1.0) / (x - 1.0))) <= 0.02) {
		tmp = 1.0 - 1.0;
	} else {
		tmp = fma(3.0, x, 1.0);
	}
	return tmp;
}

Julia

function code(x)
	tmp = 0.0
	if (Float64(Float64(x / Float64(x - -1.0)) - Float64(Float64(x - -1.0) / Float64(x - 1.0))) <= 0.02)
		tmp = Float64(1.0 - 1.0);
	else
		tmp = fma(3.0, x, 1.0);
	end
	return tmp
end

Wolfram

code[x_] := If[LessEqual[N[(N[(x / N[(x - -1.0), $MachinePrecision]), $MachinePrecision] - N[(N[(x - -1.0), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision], 0.02], N[(1.0 - 1.0), $MachinePrecision], N[(3.0 * x + 1.0), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64)))) < 0.0200000000000000004

   1. Initial program 10.3%

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

   3. Taylor expanded in x around inf

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

   5. Applied rewrites 7.1%

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

   6. Taylor expanded in x around inf

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

   8. Applied rewrites 5.8%

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

   if 0.0200000000000000004 < (-.f64 (/.f64 x (+.f64 x #s(literal 1 binary64))) (/.f64 (+.f64 x #s(literal 1 binary64)) (-.f64 x #s(literal 1 binary64))))

   1. Initial program 100.0%

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

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{1 + 3 \cdot x} \]
   4. Step-by-step derivation

      1. +-commutative N/A

         \[\leadsto 3 \cdot x + \color{blue}{1} \]

      2. lower-fma.f64 99.0

         \[\leadsto \mathsf{fma}\left(3, \color{blue}{x}, 1\right) \]

   5. Applied rewrites 99.0%

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

4. Final simplification 52.0%

   \[\leadsto \begin{array}{l} \mathbf{if}\;\frac{x}{x - -1} - \frac{x - -1}{x - 1} \leq 0.02:\\ \;\;\;\;1 - 1\\ \mathbf{else}:\\ \;\;\;\;\mathsf{fma}\left(3, x, 1\right)\\ \end{array} \]
5. Add Preprocessing

Alternative 8: 51.2% accurate, 35.0× speedup

Math

\[\begin{array}{l} \\ 1 \end{array} \]

FPCore

(FPCore (x) :precision binary64 1.0)

C

double code(double x) {
	return 1.0;
}

Fortran

module fmin_fmax_functions
    implicit none
    private
    public fmax
    public fmin

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

real(8) function code(x)
use fmin_fmax_functions
    real(8), intent (in) :: x
    code = 1.0d0
end function

Java

public static double code(double x) {
	return 1.0;
}

Python

def code(x):
	return 1.0

Julia

function code(x)
	return 1.0
end

MATLAB

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

Wolfram

code[x_] := 1.0

Derivation

1. Initial program 54.8%

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

3. Taylor expanded in x around 0

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

5. Applied rewrites 50.2%

   \[\leadsto \color{blue}{1} \]
6. Add Preprocessing

Reproduce

herbie shell --seed 2025085 
(FPCore (x)
  :name "Asymptote C"
  :precision binary64
  (- (/ x (+ x 1.0)) (/ (+ x 1.0) (- x 1.0))))