Linear.V2:$cdot from linear-1.19.1.3, A

Percentage Accurate: 99.2% → 99.6%

Time: 4.3s

Alternatives: 4

Speedup: 1.2×

• 23.8% of points produce a very large (infinite) output. You may want to add a precondition.

Specification

Math

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

FPCore

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

C

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

Fortran

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

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

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

Initial Program: 99.2% accurate, 1.0× speedup

Math

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

FPCore

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

C

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

Fortran

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

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

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

Alternative 1: 99.6% accurate, 1.2× speedup

Math

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

FPCore

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

C

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

Julia

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

Wolfram

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

Derivation

1. Initial program 98.4%

   \[x \cdot y + z \cdot t \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-+.f64 N/A

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

   2. lift-*.f64 N/A

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

   3. *-commutative N/A

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

   4. lower-fma.f64 99.6

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

   5. lift-*.f64 N/A

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

   6. *-commutative N/A

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

   7. lower-*.f64 99.6

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

4. Applied rewrites 99.6%

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

Alternative 2: 75.5% accurate, 0.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;z \cdot t \leq -0.002:\\ \;\;\;\;t \cdot z\\ \mathbf{elif}\;z \cdot t \leq 2 \cdot 10^{+59}:\\ \;\;\;\;y \cdot x\\ \mathbf{else}:\\ \;\;\;\;\left|t \cdot z\right|\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (if (<= (* z t) -0.002)
   (* t z)
   (if (<= (* z t) 2e+59) (* y x) (fabs (* t z)))))

C

double code(double x, double y, double z, double t) {
	double tmp;
	if ((z * t) <= -0.002) {
		tmp = t * z;
	} else if ((z * t) <= 2e+59) {
		tmp = y * x;
	} else {
		tmp = fabs((t * z));
	}
	return tmp;
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8) :: tmp
    if ((z * t) <= (-0.002d0)) then
        tmp = t * z
    else if ((z * t) <= 2d+59) then
        tmp = y * x
    else
        tmp = abs((t * z))
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t) {
	double tmp;
	if ((z * t) <= -0.002) {
		tmp = t * z;
	} else if ((z * t) <= 2e+59) {
		tmp = y * x;
	} else {
		tmp = Math.abs((t * z));
	}
	return tmp;
}

Python

def code(x, y, z, t):
	tmp = 0
	if (z * t) <= -0.002:
		tmp = t * z
	elif (z * t) <= 2e+59:
		tmp = y * x
	else:
		tmp = math.fabs((t * z))
	return tmp

Julia

function code(x, y, z, t)
	tmp = 0.0
	if (Float64(z * t) <= -0.002)
		tmp = Float64(t * z);
	elseif (Float64(z * t) <= 2e+59)
		tmp = Float64(y * x);
	else
		tmp = abs(Float64(t * z));
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t)
	tmp = 0.0;
	if ((z * t) <= -0.002)
		tmp = t * z;
	elseif ((z * t) <= 2e+59)
		tmp = y * x;
	else
		tmp = abs((t * z));
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_] := If[LessEqual[N[(z * t), $MachinePrecision], -0.002], N[(t * z), $MachinePrecision], If[LessEqual[N[(z * t), $MachinePrecision], 2e+59], N[(y * x), $MachinePrecision], N[Abs[N[(t * z), $MachinePrecision]], $MachinePrecision]]]

Derivation

1. Split input into 3 regimes

2. if (*.f64 z t) < -2e-3

   1. Initial program 95.7%

      \[x \cdot y + z \cdot t \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{t \cdot z} \]
   4. Step-by-step derivation

      1. lower-*.f64 83.9

         \[\leadsto \color{blue}{t \cdot z} \]

   5. Applied rewrites 83.9%

      \[\leadsto \color{blue}{t \cdot z} \]

   if -2e-3 < (*.f64 z t) < 1.99999999999999994e59

   1. Initial program 100.0%

      \[x \cdot y + z \cdot t \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{t \cdot z} \]
   4. Step-by-step derivation

      1. lower-*.f64 27.7

         \[\leadsto \color{blue}{t \cdot z} \]

   5. Applied rewrites 27.7%

      \[\leadsto \color{blue}{t \cdot z} \]
   6. Step-by-step derivation

   7. Applied rewrites 20.6%

      \[\leadsto \left|t \cdot z\right| \]

   8. Taylor expanded in x around inf

      \[\leadsto \color{blue}{x \cdot y} \]
   9. Step-by-step derivation

      1. *-commutative N/A

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

      2. lower-*.f64 78.8

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

   10. Applied rewrites 78.8%

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

   if 1.99999999999999994e59 < (*.f64 z t)

   1. Initial program 98.1%

      \[x \cdot y + z \cdot t \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{t \cdot z} \]
   4. Step-by-step derivation

      1. lower-*.f64 86.2

         \[\leadsto \color{blue}{t \cdot z} \]

   5. Applied rewrites 86.2%

      \[\leadsto \color{blue}{t \cdot z} \]
   6. Step-by-step derivation

   7. Applied rewrites 86.2%

      \[\leadsto \left|t \cdot z\right| \]
3. Recombined 3 regimes into one program.
4. Add Preprocessing

Alternative 3: 75.5% accurate, 0.5× speedup

Math

\[\begin{array}{l} \\ \begin{array}{l} \mathbf{if}\;z \cdot t \leq -0.002 \lor \neg \left(z \cdot t \leq 2 \cdot 10^{+59}\right):\\ \;\;\;\;t \cdot z\\ \mathbf{else}:\\ \;\;\;\;y \cdot x\\ \end{array} \end{array} \]

FPCore

(FPCore (x y z t)
 :precision binary64
 (if (or (<= (* z t) -0.002) (not (<= (* z t) 2e+59))) (* t z) (* y x)))

C

double code(double x, double y, double z, double t) {
	double tmp;
	if (((z * t) <= -0.002) || !((z * t) <= 2e+59)) {
		tmp = t * z;
	} else {
		tmp = y * x;
	}
	return tmp;
}

Fortran

real(8) function code(x, y, z, t)
    real(8), intent (in) :: x
    real(8), intent (in) :: y
    real(8), intent (in) :: z
    real(8), intent (in) :: t
    real(8) :: tmp
    if (((z * t) <= (-0.002d0)) .or. (.not. ((z * t) <= 2d+59))) then
        tmp = t * z
    else
        tmp = y * x
    end if
    code = tmp
end function

Java

public static double code(double x, double y, double z, double t) {
	double tmp;
	if (((z * t) <= -0.002) || !((z * t) <= 2e+59)) {
		tmp = t * z;
	} else {
		tmp = y * x;
	}
	return tmp;
}

Python

def code(x, y, z, t):
	tmp = 0
	if ((z * t) <= -0.002) or not ((z * t) <= 2e+59):
		tmp = t * z
	else:
		tmp = y * x
	return tmp

Julia

function code(x, y, z, t)
	tmp = 0.0
	if ((Float64(z * t) <= -0.002) || !(Float64(z * t) <= 2e+59))
		tmp = Float64(t * z);
	else
		tmp = Float64(y * x);
	end
	return tmp
end

MATLAB

function tmp_2 = code(x, y, z, t)
	tmp = 0.0;
	if (((z * t) <= -0.002) || ~(((z * t) <= 2e+59)))
		tmp = t * z;
	else
		tmp = y * x;
	end
	tmp_2 = tmp;
end

Wolfram

code[x_, y_, z_, t_] := If[Or[LessEqual[N[(z * t), $MachinePrecision], -0.002], N[Not[LessEqual[N[(z * t), $MachinePrecision], 2e+59]], $MachinePrecision]], N[(t * z), $MachinePrecision], N[(y * x), $MachinePrecision]]

Derivation

1. Split input into 2 regimes

2. if (*.f64 z t) < -2e-3 or 1.99999999999999994e59 < (*.f64 z t)

   1. Initial program 96.7%

      \[x \cdot y + z \cdot t \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{t \cdot z} \]
   4. Step-by-step derivation

      1. lower-*.f64 84.9

         \[\leadsto \color{blue}{t \cdot z} \]

   5. Applied rewrites 84.9%

      \[\leadsto \color{blue}{t \cdot z} \]

   if -2e-3 < (*.f64 z t) < 1.99999999999999994e59

   1. Initial program 100.0%

      \[x \cdot y + z \cdot t \]
   2. Add Preprocessing

   3. Taylor expanded in x around 0

      \[\leadsto \color{blue}{t \cdot z} \]
   4. Step-by-step derivation

      1. lower-*.f64 27.7

         \[\leadsto \color{blue}{t \cdot z} \]

   5. Applied rewrites 27.7%

      \[\leadsto \color{blue}{t \cdot z} \]
   6. Step-by-step derivation

   7. Applied rewrites 20.6%

      \[\leadsto \left|t \cdot z\right| \]

   8. Taylor expanded in x around inf

      \[\leadsto \color{blue}{x \cdot y} \]
   9. Step-by-step derivation

      1. *-commutative N/A

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

      2. lower-*.f64 78.8

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

   10. Applied rewrites 78.8%

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

4. Final simplification 81.7%

   \[\leadsto \begin{array}{l} \mathbf{if}\;z \cdot t \leq -0.002 \lor \neg \left(z \cdot t \leq 2 \cdot 10^{+59}\right):\\ \;\;\;\;t \cdot z\\ \mathbf{else}:\\ \;\;\;\;y \cdot x\\ \end{array} \]
5. Add Preprocessing

Alternative 4: 51.9% accurate, 2.3× speedup

Math

\[\begin{array}{l} \\ t \cdot z \end{array} \]

FPCore

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

C

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

Fortran

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

Java

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

Python

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

Julia

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

MATLAB

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

Wolfram

code[x_, y_, z_, t_] := N[(t * z), $MachinePrecision]

Derivation

1. Initial program 98.4%

   \[x \cdot y + z \cdot t \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \color{blue}{t \cdot z} \]
4. Step-by-step derivation

   1. lower-*.f64 55.0

      \[\leadsto \color{blue}{t \cdot z} \]

5. Applied rewrites 55.0%

   \[\leadsto \color{blue}{t \cdot z} \]
6. Add Preprocessing

Reproduce

herbie shell --seed 2024337 
(FPCore (x y z t)
  :name "Linear.V2:$cdot from linear-1.19.1.3, A"
  :precision binary64
  (+ (* x y) (* z t)))