Data.Approximate.Numerics:blog from approximate-0.2.2.1

Percentage Accurate: 99.7% → 99.9%

Time: 7.5s

Alternatives: 8

Speedup: 1.0×

Specification

Math

\[\begin{array}{l} \\ \frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (/ (* 6.0 (- x 1.0)) (+ (+ x 1.0) (* 4.0 (sqrt x)))))

C

double code(double x) {
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * sqrt(x)));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = (6.0d0 * (x - 1.0d0)) / ((x + 1.0d0) + (4.0d0 * sqrt(x)))
end function

Java

public static double code(double x) {
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * Math.sqrt(x)));
}

Python

def code(x):
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * math.sqrt(x)))

Julia

function code(x)
	return Float64(Float64(6.0 * Float64(x - 1.0)) / Float64(Float64(x + 1.0) + Float64(4.0 * sqrt(x))))
end

MATLAB

function tmp = code(x)
	tmp = (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * sqrt(x)));
end

Wolfram

code[x_] := N[(N[(6.0 * N[(x - 1.0), $MachinePrecision]), $MachinePrecision] / N[(N[(x + 1.0), $MachinePrecision] + N[(4.0 * N[Sqrt[x], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Initial Program: 99.7% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (/ (* 6.0 (- x 1.0)) (+ (+ x 1.0) (* 4.0 (sqrt x)))))

C

double code(double x) {
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * sqrt(x)));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = (6.0d0 * (x - 1.0d0)) / ((x + 1.0d0) + (4.0d0 * sqrt(x)))
end function

Java

public static double code(double x) {
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * Math.sqrt(x)));
}

Python

def code(x):
	return (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * math.sqrt(x)))

Julia

function code(x)
	return Float64(Float64(6.0 * Float64(x - 1.0)) / Float64(Float64(x + 1.0) + Float64(4.0 * sqrt(x))))
end

MATLAB

function tmp = code(x)
	tmp = (6.0 * (x - 1.0)) / ((x + 1.0) + (4.0 * sqrt(x)));
end

Wolfram

code[x_] := N[(N[(6.0 * N[(x - 1.0), $MachinePrecision]), $MachinePrecision] / N[(N[(x + 1.0), $MachinePrecision] + N[(4.0 * N[Sqrt[x], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Alternative 1: 99.9% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ 6 \cdot \frac{x + -1}{1 + \left(x + 4 \cdot \sqrt{x}\right)} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (* 6.0 (/ (+ x -1.0) (+ 1.0 (+ x (* 4.0 (sqrt x)))))))

C

double code(double x) {
	return 6.0 * ((x + -1.0) / (1.0 + (x + (4.0 * sqrt(x)))));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 6.0d0 * ((x + (-1.0d0)) / (1.0d0 + (x + (4.0d0 * sqrt(x)))))
end function

Java

public static double code(double x) {
	return 6.0 * ((x + -1.0) / (1.0 + (x + (4.0 * Math.sqrt(x)))));
}

Python

def code(x):
	return 6.0 * ((x + -1.0) / (1.0 + (x + (4.0 * math.sqrt(x)))))

Julia

function code(x)
	return Float64(6.0 * Float64(Float64(x + -1.0) / Float64(1.0 + Float64(x + Float64(4.0 * sqrt(x))))))
end

MATLAB

function tmp = code(x)
	tmp = 6.0 * ((x + -1.0) / (1.0 + (x + (4.0 * sqrt(x)))));
end

Wolfram

code[x_] := N[(6.0 * N[(N[(x + -1.0), $MachinePrecision] / N[(1.0 + N[(x + N[(4.0 * N[Sqrt[x], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
2. Step-by-step derivation

   1. *-un-lft-identity 99.8%

      \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1 \cdot \left(\left(x + 1\right) + 4 \cdot \sqrt{x}\right)}} \]

   2. times-frac 99.9%

      \[\leadsto \color{blue}{\frac{6}{1} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}}} \]

   3. metadata-eval 99.9%

      \[\leadsto \color{blue}{6} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   4. sub-neg 99.9%

      \[\leadsto 6 \cdot \frac{\color{blue}{x + \left(-1\right)}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   5. metadata-eval 99.9%

      \[\leadsto 6 \cdot \frac{x + \color{blue}{-1}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   6. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \]

   7. associate-+r+ 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \]

   8. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \]

   9. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]

3. Applied egg-rr 99.9%

   \[\leadsto \color{blue}{6 \cdot \frac{x + -1}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]
4. Step-by-step derivation

   1. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

5. Applied egg-rr 99.9%

   \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

6. Final simplification 99.9%

   \[\leadsto 6 \cdot \frac{x + -1}{1 + \left(x + 4 \cdot \sqrt{x}\right)} \]

Alternative 2: 95.6% accurate, 12.6× speedup

Math

\[\begin{array}{l} \\ 6 \cdot \frac{x + -1}{x + 1} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (* 6.0 (/ (+ x -1.0) (+ x 1.0))))

C

double code(double x) {
	return 6.0 * ((x + -1.0) / (x + 1.0));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 6.0d0 * ((x + (-1.0d0)) / (x + 1.0d0))
end function

Java

public static double code(double x) {
	return 6.0 * ((x + -1.0) / (x + 1.0));
}

Python

def code(x):
	return 6.0 * ((x + -1.0) / (x + 1.0))

Julia

function code(x)
	return Float64(6.0 * Float64(Float64(x + -1.0) / Float64(x + 1.0)))
end

MATLAB

function tmp = code(x)
	tmp = 6.0 * ((x + -1.0) / (x + 1.0));
end

Wolfram

code[x_] := N[(6.0 * N[(N[(x + -1.0), $MachinePrecision] / N[(x + 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
2. Step-by-step derivation

   1. *-un-lft-identity 99.8%

      \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1 \cdot \left(\left(x + 1\right) + 4 \cdot \sqrt{x}\right)}} \]

   2. times-frac 99.9%

      \[\leadsto \color{blue}{\frac{6}{1} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}}} \]

   3. metadata-eval 99.9%

      \[\leadsto \color{blue}{6} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   4. sub-neg 99.9%

      \[\leadsto 6 \cdot \frac{\color{blue}{x + \left(-1\right)}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   5. metadata-eval 99.9%

      \[\leadsto 6 \cdot \frac{x + \color{blue}{-1}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   6. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \]

   7. associate-+r+ 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \]

   8. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \]

   9. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]

3. Applied egg-rr 99.9%

   \[\leadsto \color{blue}{6 \cdot \frac{x + -1}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]
4. Step-by-step derivation

   1. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

5. Applied egg-rr 99.9%

   \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

6. Taylor expanded in x around inf 94.5%

   \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{x}} \]

7. Final simplification 94.5%

   \[\leadsto 6 \cdot \frac{x + -1}{x + 1} \]

Alternative 3: 95.6% accurate, 12.6× speedup

Math

\[\begin{array}{l} \\ \frac{6}{\frac{-1 - x}{1 - x}} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (/ 6.0 (/ (- -1.0 x) (- 1.0 x))))

C

double code(double x) {
	return 6.0 / ((-1.0 - x) / (1.0 - x));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 6.0d0 / (((-1.0d0) - x) / (1.0d0 - x))
end function

Java

public static double code(double x) {
	return 6.0 / ((-1.0 - x) / (1.0 - x));
}

Python

def code(x):
	return 6.0 / ((-1.0 - x) / (1.0 - x))

Julia

function code(x)
	return Float64(6.0 / Float64(Float64(-1.0 - x) / Float64(1.0 - x)))
end

MATLAB

function tmp = code(x)
	tmp = 6.0 / ((-1.0 - x) / (1.0 - x));
end

Wolfram

code[x_] := N[(6.0 / N[(N[(-1.0 - x), $MachinePrecision] / N[(1.0 - x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 99.8%

   \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
2. Step-by-step derivation

   1. *-un-lft-identity 99.8%

      \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1 \cdot \left(\left(x + 1\right) + 4 \cdot \sqrt{x}\right)}} \]

   2. times-frac 99.9%

      \[\leadsto \color{blue}{\frac{6}{1} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}}} \]

   3. metadata-eval 99.9%

      \[\leadsto \color{blue}{6} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   4. sub-neg 99.9%

      \[\leadsto 6 \cdot \frac{\color{blue}{x + \left(-1\right)}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   5. metadata-eval 99.9%

      \[\leadsto 6 \cdot \frac{x + \color{blue}{-1}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   6. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \]

   7. associate-+r+ 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \]

   8. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \]

   9. +-commutative 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]

3. Applied egg-rr 99.9%

   \[\leadsto \color{blue}{6 \cdot \frac{x + -1}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]
4. Step-by-step derivation

   1. fma-udef 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

5. Applied egg-rr 99.9%

   \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

6. Taylor expanded in x around inf 94.5%

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

   1. frac-2neg 94.5%

      \[\leadsto 6 \cdot \color{blue}{\frac{-\left(x + -1\right)}{-\left(1 + x\right)}} \]

   2. associate-*r/ 94.4%

      \[\leadsto \color{blue}{\frac{6 \cdot \left(-\left(x + -1\right)\right)}{-\left(1 + x\right)}} \]

   3. +-commutative 94.4%

      \[\leadsto \frac{6 \cdot \left(-\color{blue}{\left(-1 + x\right)}\right)}{-\left(1 + x\right)} \]

   4. distribute-neg-in 94.4%

      \[\leadsto \frac{6 \cdot \color{blue}{\left(\left(--1\right) + \left(-x\right)\right)}}{-\left(1 + x\right)} \]

   5. metadata-eval 94.4%

      \[\leadsto \frac{6 \cdot \left(\color{blue}{1} + \left(-x\right)\right)}{-\left(1 + x\right)} \]

   6. sub-neg 94.4%

      \[\leadsto \frac{6 \cdot \color{blue}{\left(1 - x\right)}}{-\left(1 + x\right)} \]

   7. distribute-neg-in 94.4%

      \[\leadsto \frac{6 \cdot \left(1 - x\right)}{\color{blue}{\left(-1\right) + \left(-x\right)}} \]

   8. metadata-eval 94.4%

      \[\leadsto \frac{6 \cdot \left(1 - x\right)}{\color{blue}{-1} + \left(-x\right)} \]

8. Applied egg-rr 94.4%

   \[\leadsto \color{blue}{\frac{6 \cdot \left(1 - x\right)}{-1 + \left(-x\right)}} \]
9. Step-by-step derivation

   1. associate-/l* 94.5%

      \[\leadsto \color{blue}{\frac{6}{\frac{-1 + \left(-x\right)}{1 - x}}} \]

   2. unsub-neg 94.5%

      \[\leadsto \frac{6}{\frac{\color{blue}{-1 - x}}{1 - x}} \]

10. Simplified 94.5%

    \[\leadsto \color{blue}{\frac{6}{\frac{-1 - x}{1 - x}}} \]

11. Final simplification 94.5%

    \[\leadsto \frac{6}{\frac{-1 - x}{1 - x}} \]

Alternative 4: 95.6% accurate, 16.0× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq 0.5:\\ \;\;\;\;-6\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{6}{x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (if (<= x 0.5) -6.0 (- 6.0 (/ 6.0 x))))

C

double code(double x) {
	double tmp;
	if (x <= 0.5) {
		tmp = -6.0;
	} else {
		tmp = 6.0 - (6.0 / x);
	}
	return tmp;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    real(8) :: tmp
    if (x <= 0.5d0) then
        tmp = -6.0d0
    else
        tmp = 6.0d0 - (6.0d0 / x)
    end if
    code = tmp
end function

Java

public static double code(double x) {
	double tmp;
	if (x <= 0.5) {
		tmp = -6.0;
	} else {
		tmp = 6.0 - (6.0 / x);
	}
	return tmp;
}

Python

def code(x):
	tmp = 0
	if x <= 0.5:
		tmp = -6.0
	else:
		tmp = 6.0 - (6.0 / x)
	return tmp

Julia

function code(x)
	tmp = 0.0
	if (x <= 0.5)
		tmp = -6.0;
	else
		tmp = Float64(6.0 - Float64(6.0 / x));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x)
	tmp = 0.0;
	if (x <= 0.5)
		tmp = -6.0;
	else
		tmp = 6.0 - (6.0 / x);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_] := If[LessEqual[x, 0.5], -6.0, N[(6.0 - N[(6.0 / x), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < 0.5

   1. Initial program 99.9%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   2. Taylor expanded in x around 0 94.2%

      \[\leadsto \color{blue}{-6} \]

   if 0.5 < x

   1. Initial program 99.7%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
   2. Step-by-step derivation

      1. associate-/l* 99.9%

         \[\leadsto \color{blue}{\frac{6}{\frac{\left(x + 1\right) + 4 \cdot \sqrt{x}}{x - 1}}} \]

      2. associate-/r/ 99.8%

         \[\leadsto \color{blue}{\frac{6}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \cdot \left(x - 1\right)} \]

      3. +-commutative 99.8%

         \[\leadsto \frac{6}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \cdot \left(x - 1\right) \]

      4. associate-+r+ 99.8%

         \[\leadsto \frac{6}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \cdot \left(x - 1\right) \]

      5. fma-udef 99.8%

         \[\leadsto \frac{6}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \cdot \left(x - 1\right) \]

      6. +-commutative 99.8%

         \[\leadsto \frac{6}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \cdot \left(x - 1\right) \]

      7. sub-neg 99.8%

         \[\leadsto \frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \color{blue}{\left(x + \left(-1\right)\right)} \]

      8. metadata-eval 99.8%

         \[\leadsto \frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \left(x + \color{blue}{-1}\right) \]

   3. Applied egg-rr 99.8%

      \[\leadsto \color{blue}{\frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \left(x + -1\right)} \]

   4. Taylor expanded in x around inf 94.7%

      \[\leadsto \color{blue}{\frac{6}{x}} \cdot \left(x + -1\right) \]

   5. Taylor expanded in x around 0 94.8%

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

      1. associate-*r/ 94.8%

         \[\leadsto 6 - \color{blue}{\frac{6 \cdot 1}{x}} \]

      2. metadata-eval 94.8%

         \[\leadsto 6 - \frac{\color{blue}{6}}{x} \]

   7. Simplified 94.8%

      \[\leadsto \color{blue}{6 - \frac{6}{x}} \]
3. Recombined 2 regimes into one program.

4. Final simplification 94.5%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq 0.5:\\ \;\;\;\;-6\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{6}{x}\\ \end{array} \]

Alternative 5: 95.6% accurate, 16.0× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq 0.49:\\ \;\;\;\;x \cdot 12 - 6\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{6}{x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= x 0.49) (- (* x 12.0) 6.0) (- 6.0 (/ 6.0 x))))

C

double code(double x) {
	double tmp;
	if (x <= 0.49) {
		tmp = (x * 12.0) - 6.0;
	} else {
		tmp = 6.0 - (6.0 / x);
	}
	return tmp;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    real(8) :: tmp
    if (x <= 0.49d0) then
        tmp = (x * 12.0d0) - 6.0d0
    else
        tmp = 6.0d0 - (6.0d0 / x)
    end if
    code = tmp
end function

Java

public static double code(double x) {
	double tmp;
	if (x <= 0.49) {
		tmp = (x * 12.0) - 6.0;
	} else {
		tmp = 6.0 - (6.0 / x);
	}
	return tmp;
}

Python

def code(x):
	tmp = 0
	if x <= 0.49:
		tmp = (x * 12.0) - 6.0
	else:
		tmp = 6.0 - (6.0 / x)
	return tmp

Julia

function code(x)
	tmp = 0.0
	if (x <= 0.49)
		tmp = Float64(Float64(x * 12.0) - 6.0);
	else
		tmp = Float64(6.0 - Float64(6.0 / x));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x)
	tmp = 0.0;
	if (x <= 0.49)
		tmp = (x * 12.0) - 6.0;
	else
		tmp = 6.0 - (6.0 / x);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_] := If[LessEqual[x, 0.49], N[(N[(x * 12.0), $MachinePrecision] - 6.0), $MachinePrecision], N[(6.0 - N[(6.0 / x), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < 0.48999999999999999

   1. Initial program 99.9%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
   2. Step-by-step derivation

      1. *-un-lft-identity 99.9%

         \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1 \cdot \left(\left(x + 1\right) + 4 \cdot \sqrt{x}\right)}} \]

      2. times-frac 99.9%

         \[\leadsto \color{blue}{\frac{6}{1} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}}} \]

      3. metadata-eval 99.9%

         \[\leadsto \color{blue}{6} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      4. sub-neg 99.9%

         \[\leadsto 6 \cdot \frac{\color{blue}{x + \left(-1\right)}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      5. metadata-eval 99.9%

         \[\leadsto 6 \cdot \frac{x + \color{blue}{-1}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      6. +-commutative 99.9%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \]

      7. associate-+r+ 99.9%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \]

      8. fma-udef 99.9%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \]

      9. +-commutative 99.9%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]

   3. Applied egg-rr 99.9%

      \[\leadsto \color{blue}{6 \cdot \frac{x + -1}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]
   4. Step-by-step derivation

      1. fma-udef 99.9%

         \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

   5. Applied egg-rr 99.9%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

   6. Taylor expanded in x around inf 94.2%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{x}} \]

   7. Taylor expanded in x around 0 94.2%

      \[\leadsto \color{blue}{12 \cdot x - 6} \]

   if 0.48999999999999999 < x

   1. Initial program 99.7%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
   2. Step-by-step derivation

      1. associate-/l* 99.9%

         \[\leadsto \color{blue}{\frac{6}{\frac{\left(x + 1\right) + 4 \cdot \sqrt{x}}{x - 1}}} \]

      2. associate-/r/ 99.8%

         \[\leadsto \color{blue}{\frac{6}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \cdot \left(x - 1\right)} \]

      3. +-commutative 99.8%

         \[\leadsto \frac{6}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \cdot \left(x - 1\right) \]

      4. associate-+r+ 99.8%

         \[\leadsto \frac{6}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \cdot \left(x - 1\right) \]

      5. fma-udef 99.8%

         \[\leadsto \frac{6}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \cdot \left(x - 1\right) \]

      6. +-commutative 99.8%

         \[\leadsto \frac{6}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \cdot \left(x - 1\right) \]

      7. sub-neg 99.8%

         \[\leadsto \frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \color{blue}{\left(x + \left(-1\right)\right)} \]

      8. metadata-eval 99.8%

         \[\leadsto \frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \left(x + \color{blue}{-1}\right) \]

   3. Applied egg-rr 99.8%

      \[\leadsto \color{blue}{\frac{6}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)} \cdot \left(x + -1\right)} \]

   4. Taylor expanded in x around inf 94.7%

      \[\leadsto \color{blue}{\frac{6}{x}} \cdot \left(x + -1\right) \]

   5. Taylor expanded in x around 0 94.8%

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

      1. associate-*r/ 94.8%

         \[\leadsto 6 - \color{blue}{\frac{6 \cdot 1}{x}} \]

      2. metadata-eval 94.8%

         \[\leadsto 6 - \frac{\color{blue}{6}}{x} \]

   7. Simplified 94.8%

      \[\leadsto \color{blue}{6 - \frac{6}{x}} \]
3. Recombined 2 regimes into one program.

4. Final simplification 94.5%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq 0.49:\\ \;\;\;\;x \cdot 12 - 6\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{6}{x}\\ \end{array} \]

Alternative 6: 95.6% accurate, 16.0× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq 2:\\ \;\;\;\;6 \cdot \left(x + -1\right)\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{12}{x}\\ \end{array} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (if (<= x 2.0) (* 6.0 (+ x -1.0)) (- 6.0 (/ 12.0 x))))

C

double code(double x) {
	double tmp;
	if (x <= 2.0) {
		tmp = 6.0 * (x + -1.0);
	} else {
		tmp = 6.0 - (12.0 / x);
	}
	return tmp;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    real(8) :: tmp
    if (x <= 2.0d0) then
        tmp = 6.0d0 * (x + (-1.0d0))
    else
        tmp = 6.0d0 - (12.0d0 / x)
    end if
    code = tmp
end function

Java

public static double code(double x) {
	double tmp;
	if (x <= 2.0) {
		tmp = 6.0 * (x + -1.0);
	} else {
		tmp = 6.0 - (12.0 / x);
	}
	return tmp;
}

Python

def code(x):
	tmp = 0
	if x <= 2.0:
		tmp = 6.0 * (x + -1.0)
	else:
		tmp = 6.0 - (12.0 / x)
	return tmp

Julia

function code(x)
	tmp = 0.0
	if (x <= 2.0)
		tmp = Float64(6.0 * Float64(x + -1.0));
	else
		tmp = Float64(6.0 - Float64(12.0 / x));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x)
	tmp = 0.0;
	if (x <= 2.0)
		tmp = 6.0 * (x + -1.0);
	else
		tmp = 6.0 - (12.0 / x);
	end
	tmp_2 = tmp;
end

Wolfram

code[x_] := If[LessEqual[x, 2.0], N[(6.0 * N[(x + -1.0), $MachinePrecision]), $MachinePrecision], N[(6.0 - N[(12.0 / x), $MachinePrecision]), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if x < 2

   1. Initial program 99.9%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   2. Taylor expanded in x around 0 93.0%

      \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1}} \]

   if 2 < x

   1. Initial program 99.7%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]
   2. Step-by-step derivation

      1. *-un-lft-identity 99.7%

         \[\leadsto \frac{6 \cdot \left(x - 1\right)}{\color{blue}{1 \cdot \left(\left(x + 1\right) + 4 \cdot \sqrt{x}\right)}} \]

      2. times-frac 100.0%

         \[\leadsto \color{blue}{\frac{6}{1} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}}} \]

      3. metadata-eval 100.0%

         \[\leadsto \color{blue}{6} \cdot \frac{x - 1}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      4. sub-neg 100.0%

         \[\leadsto 6 \cdot \frac{\color{blue}{x + \left(-1\right)}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      5. metadata-eval 100.0%

         \[\leadsto 6 \cdot \frac{x + \color{blue}{-1}}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

      6. +-commutative 100.0%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{4 \cdot \sqrt{x} + \left(x + 1\right)}} \]

      7. associate-+r+ 100.0%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\left(4 \cdot \sqrt{x} + x\right) + 1}} \]

      8. fma-udef 100.0%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{\mathsf{fma}\left(4, \sqrt{x}, x\right)} + 1} \]

      9. +-commutative 100.0%

         \[\leadsto 6 \cdot \frac{x + -1}{\color{blue}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]

   3. Applied egg-rr 100.0%

      \[\leadsto \color{blue}{6 \cdot \frac{x + -1}{1 + \mathsf{fma}\left(4, \sqrt{x}, x\right)}} \]
   4. Step-by-step derivation

      1. fma-udef 100.0%

         \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

   5. Applied egg-rr 100.0%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{\left(4 \cdot \sqrt{x} + x\right)}} \]

   6. Taylor expanded in x around inf 96.0%

      \[\leadsto 6 \cdot \frac{x + -1}{1 + \color{blue}{x}} \]

   7. Taylor expanded in x around inf 96.0%

      \[\leadsto \color{blue}{6 - 12 \cdot \frac{1}{x}} \]
   8. Step-by-step derivation

      1. associate-*r/ 96.0%

         \[\leadsto 6 - \color{blue}{\frac{12 \cdot 1}{x}} \]

      2. metadata-eval 96.0%

         \[\leadsto 6 - \frac{\color{blue}{12}}{x} \]

   9. Simplified 96.0%

      \[\leadsto \color{blue}{6 - \frac{12}{x}} \]
3. Recombined 2 regimes into one program.

4. Final simplification 94.5%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq 2:\\ \;\;\;\;6 \cdot \left(x + -1\right)\\ \mathbf{else}:\\ \;\;\;\;6 - \frac{12}{x}\\ \end{array} \]

Alternative 7: 95.6% accurate, 36.9× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;x \leq 1:\\ \;\;\;\;-6\\ \mathbf{else}:\\ \;\;\;\;6\\ \end{array} \end{array} \]

FPCore

(FPCore (x) :precision binary64 (if (<= x 1.0) -6.0 6.0))

C

double code(double x) {
	double tmp;
	if (x <= 1.0) {
		tmp = -6.0;
	} else {
		tmp = 6.0;
	}
	return tmp;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    real(8) :: tmp
    if (x <= 1.0d0) then
        tmp = -6.0d0
    else
        tmp = 6.0d0
    end if
    code = tmp
end function

Java

public static double code(double x) {
	double tmp;
	if (x <= 1.0) {
		tmp = -6.0;
	} else {
		tmp = 6.0;
	}
	return tmp;
}

Python

def code(x):
	tmp = 0
	if x <= 1.0:
		tmp = -6.0
	else:
		tmp = 6.0
	return tmp

Julia

function code(x)
	tmp = 0.0
	if (x <= 1.0)
		tmp = -6.0;
	else
		tmp = 6.0;
	end
	return tmp
end

MATLAB

function tmp_2 = code(x)
	tmp = 0.0;
	if (x <= 1.0)
		tmp = -6.0;
	else
		tmp = 6.0;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_] := If[LessEqual[x, 1.0], -6.0, 6.0]

Derivation

1. Split input into 2 regimes

2. if x < 1

   1. Initial program 99.9%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   2. Taylor expanded in x around 0 92.9%

      \[\leadsto \color{blue}{-6} \]

   if 1 < x

   1. Initial program 99.7%

      \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

   2. Taylor expanded in x around inf 96.0%

      \[\leadsto \color{blue}{6} \]
3. Recombined 2 regimes into one program.

4. Final simplification 94.5%

   \[\leadsto \begin{array}{l} \mathbf{if}\;x \leq 1:\\ \;\;\;\;-6\\ \mathbf{else}:\\ \;\;\;\;6\\ \end{array} \]

Alternative 8: 48.0% accurate, 113.0× speedup

Math

\[\begin{array}{l} \\ -6 \end{array} \]

FPCore

(FPCore (x) :precision binary64 -6.0)

C

double code(double x) {
	return -6.0;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = -6.0d0
end function

Java

public static double code(double x) {
	return -6.0;
}

Python

def code(x):
	return -6.0

Julia

function code(x)
	return -6.0
end

MATLAB

function tmp = code(x)
	tmp = -6.0;
end

Wolfram

code[x_] := -6.0

Derivation

1. Initial program 99.8%

   \[\frac{6 \cdot \left(x - 1\right)}{\left(x + 1\right) + 4 \cdot \sqrt{x}} \]

2. Taylor expanded in x around 0 45.1%

   \[\leadsto \color{blue}{-6} \]

3. Final simplification 45.1%

   \[\leadsto -6 \]

Developer target: 99.9% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{6}{\frac{\left(x + 1\right) + 4 \cdot \sqrt{x}}{x - 1}} \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (/ 6.0 (/ (+ (+ x 1.0) (* 4.0 (sqrt x))) (- x 1.0))))

C

double code(double x) {
	return 6.0 / (((x + 1.0) + (4.0 * sqrt(x))) / (x - 1.0));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = 6.0d0 / (((x + 1.0d0) + (4.0d0 * sqrt(x))) / (x - 1.0d0))
end function

Java

public static double code(double x) {
	return 6.0 / (((x + 1.0) + (4.0 * Math.sqrt(x))) / (x - 1.0));
}

Python

def code(x):
	return 6.0 / (((x + 1.0) + (4.0 * math.sqrt(x))) / (x - 1.0))

Julia

function code(x)
	return Float64(6.0 / Float64(Float64(Float64(x + 1.0) + Float64(4.0 * sqrt(x))) / Float64(x - 1.0)))
end

MATLAB

function tmp = code(x)
	tmp = 6.0 / (((x + 1.0) + (4.0 * sqrt(x))) / (x - 1.0));
end

Wolfram

code[x_] := N[(6.0 / N[(N[(N[(x + 1.0), $MachinePrecision] + N[(4.0 * N[Sqrt[x], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] / N[(x - 1.0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Reproduce

herbie shell --seed 2023334 
(FPCore (x)
  :name "Data.Approximate.Numerics:blog from approximate-0.2.2.1"
  :precision binary64

  :herbie-target
  (/ 6.0 (/ (+ (+ x 1.0) (* 4.0 (sqrt x))) (- x 1.0)))

  (/ (* 6.0 (- x 1.0)) (+ (+ x 1.0) (* 4.0 (sqrt x)))))