Numeric.Log:$clog1p from log-domain-0.10.2.1, A

Percentage Accurate: 100.0% → 100.0%

Time: 5.8s

Alternatives: 8

Speedup: 1.2×

• 44.0% of points produce a very large (infinite) output. You may want to add a precondition.

Specification

Math

\[\begin{array}{l} \\ \left(x \cdot 2 + x \cdot x\right) + y \cdot y \end{array} \]

FPCore

(FPCore (x y) :precision binary64 (+ (+ (* x 2.0) (* x x)) (* y y)))

C

double code(double x, double y) {
	return ((x * 2.0) + (x * x)) + (y * y);
}

Fortran

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

Java

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

Python

def code(x, y):
	return ((x * 2.0) + (x * x)) + (y * y)

Julia

function code(x, y)
	return Float64(Float64(Float64(x * 2.0) + Float64(x * x)) + Float64(y * y))
end

MATLAB

function tmp = code(x, y)
	tmp = ((x * 2.0) + (x * x)) + (y * y);
end

Wolfram

code[x_, y_] := N[(N[(N[(x * 2.0), $MachinePrecision] + N[(x * x), $MachinePrecision]), $MachinePrecision] + N[(y * y), $MachinePrecision]), $MachinePrecision]

Initial Program: 100.0% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \left(x \cdot 2 + x \cdot x\right) + y \cdot y \end{array} \]

FPCore

(FPCore (x y) :precision binary64 (+ (+ (* x 2.0) (* x x)) (* y y)))

C

double code(double x, double y) {
	return ((x * 2.0) + (x * x)) + (y * y);
}

Fortran

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

Java

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

Python

def code(x, y):
	return ((x * 2.0) + (x * x)) + (y * y)

Julia

function code(x, y)
	return Float64(Float64(Float64(x * 2.0) + Float64(x * x)) + Float64(y * y))
end

MATLAB

function tmp = code(x, y)
	tmp = ((x * 2.0) + (x * x)) + (y * y);
end

Wolfram

code[x_, y_] := N[(N[(N[(x * 2.0), $MachinePrecision] + N[(x * x), $MachinePrecision]), $MachinePrecision] + N[(y * y), $MachinePrecision]), $MachinePrecision]

Alternative 1: 100.0% accurate, 1.2× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(x, x, \mathsf{fma}\left(x, 2, y \cdot y\right)\right) \end{array} \]

FPCore

(FPCore (x y) :precision binary64 (fma x x (fma x 2.0 (* y y))))

C

double code(double x, double y) {
	return fma(x, x, fma(x, 2.0, (y * y)));
}

Julia

function code(x, y)
	return fma(x, x, fma(x, 2.0, Float64(y * y)))
end

Wolfram

code[x_, y_] := N[(x * x + N[(x * 2.0 + N[(y * y), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 100.0%

   \[\left(x \cdot 2 + x \cdot x\right) + y \cdot y \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-+.f64 N/A

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

   2. lift-+.f64 N/A

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

   3. +-commutative N/A

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

   4. associate-+l+ N/A

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

   5. lift-*.f64 N/A

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

   6. lower-fma.f64 N/A

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

   7. lift-*.f64 N/A

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

   8. lower-fma.f64 100.0

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

4. Applied rewrites 100.0%

   \[\leadsto \color{blue}{\mathsf{fma}\left(x, x, \mathsf{fma}\left(x, 2, y \cdot y\right)\right)} \]
5. Add Preprocessing

Alternative 2: 70.1% accurate, 0.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} t_0 := x \cdot 2 + x \cdot x\\ \mathbf{if}\;t\_0 \leq 2 \cdot 10^{-162}:\\ \;\;\;\;y \cdot y\\ \mathbf{elif}\;t\_0 \leq 0.5:\\ \;\;\;\;x \cdot 2\\ \mathbf{else}:\\ \;\;\;\;x \cdot x\\ \end{array} \end{array} \]

FPCore

(FPCore (x y)
 :precision binary64
 (let* ((t_0 (+ (* x 2.0) (* x x))))
   (if (<= t_0 2e-162) (* y y) (if (<= t_0 0.5) (* x 2.0) (* x x)))))

C

double code(double x, double y) {
	double t_0 = (x * 2.0) + (x * x);
	double tmp;
	if (t_0 <= 2e-162) {
		tmp = y * y;
	} else if (t_0 <= 0.5) {
		tmp = x * 2.0;
	} else {
		tmp = x * x;
	}
	return tmp;
}

Fortran

real(8) function code(x, y)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8) :: t_0
    real(8) :: tmp
    t_0 = (x * 2.0d0) + (x * x)
    if (t_0 <= 2d-162) then
        tmp = y * y
    else if (t_0 <= 0.5d0) then
        tmp = x * 2.0d0
    else
        tmp = x * x
    end if
    code = tmp
end function

Java

public static double code(double x, double y) {
	double t_0 = (x * 2.0) + (x * x);
	double tmp;
	if (t_0 <= 2e-162) {
		tmp = y * y;
	} else if (t_0 <= 0.5) {
		tmp = x * 2.0;
	} else {
		tmp = x * x;
	}
	return tmp;
}

Python

def code(x, y):
	t_0 = (x * 2.0) + (x * x)
	tmp = 0
	if t_0 <= 2e-162:
		tmp = y * y
	elif t_0 <= 0.5:
		tmp = x * 2.0
	else:
		tmp = x * x
	return tmp

Julia

function code(x, y)
	t_0 = Float64(Float64(x * 2.0) + Float64(x * x))
	tmp = 0.0
	if (t_0 <= 2e-162)
		tmp = Float64(y * y);
	elseif (t_0 <= 0.5)
		tmp = Float64(x * 2.0);
	else
		tmp = Float64(x * x);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y)
	t_0 = (x * 2.0) + (x * x);
	tmp = 0.0;
	if (t_0 <= 2e-162)
		tmp = y * y;
	elseif (t_0 <= 0.5)
		tmp = x * 2.0;
	else
		tmp = x * x;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_] := Block[{t$95$0 = N[(N[(x * 2.0), $MachinePrecision] + N[(x * x), $MachinePrecision]), $MachinePrecision]}, If[LessEqual[t$95$0, 2e-162], N[(y * y), $MachinePrecision], If[LessEqual[t$95$0, 0.5], N[(x * 2.0), $MachinePrecision], N[(x * x), $MachinePrecision]]]]

Derivation

1. Split input into 3 regimes

2. if (+.f64 (*.f64 x #s(literal 2 binary64)) (*.f64 x x)) < 1.99999999999999991e-162

   1. Initial program 100.0%

      \[\left(x \cdot 2 + x \cdot x\right) + y \cdot y \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{{y}^{2}} \]
   4. Step-by-step derivation

      1. unpow2 N/A

         \[\leadsto \color{blue}{y \cdot y} \]

      2. lower-*.f64 66.1

         \[\leadsto \color{blue}{y \cdot y} \]

   5. Applied rewrites 66.1%

      \[\leadsto \color{blue}{y \cdot y} \]

   if 1.99999999999999991e-162 < (+.f64 (*.f64 x #s(literal 2 binary64)) (*.f64 x x)) < 0.5

   1. Initial program 100.0%

      \[\left(x \cdot 2 + x \cdot x\right) + y \cdot y \]
   2. Add Preprocessing

   3. Taylor expanded in x around inf

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

      1. unpow2 N/A

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

      2. associate-*l* N/A

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

      3. +-commutative N/A

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

      4. distribute-rgt-in N/A

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

      5. associate-*l* N/A

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

      6. lft-mult-inverse N/A

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

      7. metadata-eval N/A

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

      8. *-lft-identity N/A

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

      9. lower-*.f64 N/A

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

      10. lower-+.f64 41.9

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

   5. Applied rewrites 41.9%

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

   6. Taylor expanded in x around 0

      \[\leadsto x \cdot 2 \]
   7. Step-by-step derivation

   8. Applied rewrites 40.0%

      \[\leadsto x \cdot 2 \]

   if 0.5 < (+.f64 (*.f64 x #s(literal 2 binary64)) (*.f64 x x))

   1. Initial program 99.9%

      \[\left(x \cdot 2 + x \cdot x\right) + y \cdot y \]
   2. Add Preprocessing

   3. Taylor expanded in x around inf

      \[\leadsto \color{blue}{{x}^{2}} \]
   4. Step-by-step derivation

      1. unpow2 N/A

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

      2. lower-*.f64 80.9

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

   5. Applied rewrites 80.9%

      \[\leadsto \color{blue}{x \cdot x} \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Developer Target 1: 100.0% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ y \cdot y + \left(2 \cdot x + x \cdot x\right) \end{array} \]

FPCore

(FPCore (x y) :precision binary64 (+ (* y y) (+ (* 2.0 x) (* x x))))

C

double code(double x, double y) {
	return (y * y) + ((2.0 * x) + (x * x));
}

Fortran

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

Java

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

Python

def code(x, y):
	return (y * y) + ((2.0 * x) + (x * x))

Julia

function code(x, y)
	return Float64(Float64(y * y) + Float64(Float64(2.0 * x) + Float64(x * x)))
end

MATLAB

function tmp = code(x, y)
	tmp = (y * y) + ((2.0 * x) + (x * x));
end

Wolfram

code[x_, y_] := N[(N[(y * y), $MachinePrecision] + N[(N[(2.0 * x), $MachinePrecision] + N[(x * x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Reproduce

herbie shell --seed 2024228 
(FPCore (x y)
  :name "Numeric.Log:$clog1p from log-domain-0.10.2.1, A"
  :precision binary64

  :alt
  (! :herbie-platform default (+ (* y y) (+ (* 2 x) (* x x))))

  (+ (+ (* x 2.0) (* x x)) (* y y)))