sqrt C (should all be same)

Percentage Accurate: 53.7% → 99.4%

Time: 9.1s

Alternatives: 5

Speedup: 1.3×

Specification

Math

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

FPCore

(FPCore (x) :precision binary64 (sqrt (* 2.0 (* x x))))

C

double code(double x) {
	return sqrt((2.0 * (x * x)));
}

Fortran

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

Java

public static double code(double x) {
	return Math.sqrt((2.0 * (x * x)));
}

Python

def code(x):
	return math.sqrt((2.0 * (x * x)))

Julia

function code(x)
	return sqrt(Float64(2.0 * Float64(x * x)))
end

MATLAB

function tmp = code(x)
	tmp = sqrt((2.0 * (x * x)));
end

Wolfram

code[x_] := N[Sqrt[N[(2.0 * N[(x * x), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]

Initial Program: 53.7% accurate, 1.0× speedup

Math

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

FPCore

(FPCore (x) :precision binary64 (sqrt (* 2.0 (* x x))))

C

double code(double x) {
	return sqrt((2.0 * (x * x)));
}

Fortran

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

Java

public static double code(double x) {
	return Math.sqrt((2.0 * (x * x)));
}

Python

def code(x):
	return math.sqrt((2.0 * (x * x)))

Julia

function code(x)
	return sqrt(Float64(2.0 * Float64(x * x)))
end

MATLAB

function tmp = code(x)
	tmp = sqrt((2.0 * (x * x)));
end

Wolfram

code[x_] := N[Sqrt[N[(2.0 * N[(x * x), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]

Alternative 1: 99.4% accurate, 0.1× speedup

Math

\[\begin{array}{l} x_m = \left|x\right| \\ {x\_m}^{0.75} \cdot {\left(x\_m \cdot 4\right)}^{0.25} \end{array} \]

FPCore

x_m = (fabs.f64 x)
(FPCore (x_m) :precision binary64 (* (pow x_m 0.75) (pow (* x_m 4.0) 0.25)))

C

x_m = fabs(x);
double code(double x_m) {
	return pow(x_m, 0.75) * pow((x_m * 4.0), 0.25);
}

Fortran

x_m = abs(x)
real(8) function code(x_m)
    real(8), intent (in) :: x_m
    code = (x_m ** 0.75d0) * ((x_m * 4.0d0) ** 0.25d0)
end function

Java

x_m = Math.abs(x);
public static double code(double x_m) {
	return Math.pow(x_m, 0.75) * Math.pow((x_m * 4.0), 0.25);
}

Python

x_m = math.fabs(x)
def code(x_m):
	return math.pow(x_m, 0.75) * math.pow((x_m * 4.0), 0.25)

Julia

x_m = abs(x)
function code(x_m)
	return Float64((x_m ^ 0.75) * (Float64(x_m * 4.0) ^ 0.25))
end

MATLAB

x_m = abs(x);
function tmp = code(x_m)
	tmp = (x_m ^ 0.75) * ((x_m * 4.0) ^ 0.25);
end

Wolfram

x_m = N[Abs[x], $MachinePrecision]
code[x$95$m_] := N[(N[Power[x$95$m, 0.75], $MachinePrecision] * N[Power[N[(x$95$m * 4.0), $MachinePrecision], 0.25], $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 53.9%

   \[\sqrt{2 \cdot \left(x \cdot x\right)} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-sqrt.f64 N/A

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

   2. pow1/2 N/A

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

   3. lift-*.f64 N/A

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

   4. lift-*.f64 N/A

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

   5. pow2 N/A

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

   6. metadata-eval N/A

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

   7. metadata-eval N/A

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

   8. metadata-eval N/A

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

   9. pow-prod-up N/A

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

   10. metadata-eval N/A

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

   11. pow-plus N/A

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

   12. associate-*r* N/A

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

   13. associate-*r* N/A

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

   14. unpow-prod-down N/A

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

   15. lower-*.f64 N/A

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

4. Applied rewrites 49.4%

   \[\leadsto \color{blue}{{\left(4 \cdot x\right)}^{0.25} \cdot {x}^{0.75}} \]

5. Final simplification 49.4%

   \[\leadsto {x}^{0.75} \cdot {\left(x \cdot 4\right)}^{0.25} \]
6. Add Preprocessing

Alternative 2: 99.5% accurate, 0.2× speedup

Math

\[\begin{array}{l} x_m = \left|x\right| \\ {x\_m}^{0.5} \cdot \sqrt{2 \cdot x\_m} \end{array} \]

FPCore

x_m = (fabs.f64 x)
(FPCore (x_m) :precision binary64 (* (pow x_m 0.5) (sqrt (* 2.0 x_m))))

C

x_m = fabs(x);
double code(double x_m) {
	return pow(x_m, 0.5) * sqrt((2.0 * x_m));
}

Fortran

x_m = abs(x)
real(8) function code(x_m)
    real(8), intent (in) :: x_m
    code = (x_m ** 0.5d0) * sqrt((2.0d0 * x_m))
end function

Java

x_m = Math.abs(x);
public static double code(double x_m) {
	return Math.pow(x_m, 0.5) * Math.sqrt((2.0 * x_m));
}

Python

x_m = math.fabs(x)
def code(x_m):
	return math.pow(x_m, 0.5) * math.sqrt((2.0 * x_m))

Julia

x_m = abs(x)
function code(x_m)
	return Float64((x_m ^ 0.5) * sqrt(Float64(2.0 * x_m)))
end

MATLAB

x_m = abs(x);
function tmp = code(x_m)
	tmp = (x_m ^ 0.5) * sqrt((2.0 * x_m));
end

Wolfram

x_m = N[Abs[x], $MachinePrecision]
code[x$95$m_] := N[(N[Power[x$95$m, 0.5], $MachinePrecision] * N[Sqrt[N[(2.0 * x$95$m), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 53.9%

   \[\sqrt{2 \cdot \left(x \cdot x\right)} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-sqrt.f64 N/A

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

   2. pow1/2 N/A

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

   3. lift-*.f64 N/A

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

   4. lift-*.f64 N/A

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

   5. associate-*r* N/A

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

   6. unpow-prod-down N/A

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

   7. lower-*.f64 N/A

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

   8. pow1/2 N/A

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

   9. lower-sqrt.f64 N/A

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

   10. *-commutative N/A

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

   11. lower-*.f64 N/A

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

   12. pow1/2 N/A

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

   13. lower-sqrt.f64 49.4

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

4. Applied rewrites 49.4%

   \[\leadsto \color{blue}{\sqrt{x \cdot 2} \cdot \sqrt{x}} \]
5. Step-by-step derivation

   1. lift-sqrt.f64 N/A

      \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

   2. pow1/2 N/A

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

   3. lower-pow.f64 49.4

      \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{{x}^{0.5}} \]

6. Applied rewrites 49.4%

   \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{{x}^{0.5}} \]

7. Final simplification 49.4%

   \[\leadsto {x}^{0.5} \cdot \sqrt{2 \cdot x} \]
8. Add Preprocessing

Alternative 3: 99.5% accurate, 0.7× speedup

Math

\[\begin{array}{l} x_m = \left|x\right| \\ \sqrt{x\_m} \cdot \sqrt{2 \cdot x\_m} \end{array} \]

FPCore

x_m = (fabs.f64 x)
(FPCore (x_m) :precision binary64 (* (sqrt x_m) (sqrt (* 2.0 x_m))))

C

x_m = fabs(x);
double code(double x_m) {
	return sqrt(x_m) * sqrt((2.0 * x_m));
}

Fortran

x_m = abs(x)
real(8) function code(x_m)
    real(8), intent (in) :: x_m
    code = sqrt(x_m) * sqrt((2.0d0 * x_m))
end function

Java

x_m = Math.abs(x);
public static double code(double x_m) {
	return Math.sqrt(x_m) * Math.sqrt((2.0 * x_m));
}

Python

x_m = math.fabs(x)
def code(x_m):
	return math.sqrt(x_m) * math.sqrt((2.0 * x_m))

Julia

x_m = abs(x)
function code(x_m)
	return Float64(sqrt(x_m) * sqrt(Float64(2.0 * x_m)))
end

MATLAB

x_m = abs(x);
function tmp = code(x_m)
	tmp = sqrt(x_m) * sqrt((2.0 * x_m));
end

Wolfram

x_m = N[Abs[x], $MachinePrecision]
code[x$95$m_] := N[(N[Sqrt[x$95$m], $MachinePrecision] * N[Sqrt[N[(2.0 * x$95$m), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 53.9%

   \[\sqrt{2 \cdot \left(x \cdot x\right)} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-sqrt.f64 N/A

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

   2. pow1/2 N/A

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

   3. lift-*.f64 N/A

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

   4. lift-*.f64 N/A

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

   5. associate-*r* N/A

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

   6. unpow-prod-down N/A

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

   7. lower-*.f64 N/A

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

   8. pow1/2 N/A

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

   9. lower-sqrt.f64 N/A

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

   10. *-commutative N/A

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

   11. lower-*.f64 N/A

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

   12. pow1/2 N/A

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

   13. lower-sqrt.f64 49.4

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

4. Applied rewrites 49.4%

   \[\leadsto \color{blue}{\sqrt{x \cdot 2} \cdot \sqrt{x}} \]

5. Final simplification 49.4%

   \[\leadsto \sqrt{x} \cdot \sqrt{2 \cdot x} \]
6. Add Preprocessing

Alternative 4: 99.3% accurate, 1.3× speedup

Math

\[\begin{array}{l} x_m = \left|x\right| \\ \sqrt{2} \cdot x\_m \end{array} \]

FPCore

x_m = (fabs.f64 x)
(FPCore (x_m) :precision binary64 (* (sqrt 2.0) x_m))

C

x_m = fabs(x);
double code(double x_m) {
	return sqrt(2.0) * x_m;
}

Fortran

x_m = abs(x)
real(8) function code(x_m)
    real(8), intent (in) :: x_m
    code = sqrt(2.0d0) * x_m
end function

Java

x_m = Math.abs(x);
public static double code(double x_m) {
	return Math.sqrt(2.0) * x_m;
}

Python

x_m = math.fabs(x)
def code(x_m):
	return math.sqrt(2.0) * x_m

Julia

x_m = abs(x)
function code(x_m)
	return Float64(sqrt(2.0) * x_m)
end

MATLAB

x_m = abs(x);
function tmp = code(x_m)
	tmp = sqrt(2.0) * x_m;
end

Wolfram

x_m = N[Abs[x], $MachinePrecision]
code[x$95$m_] := N[(N[Sqrt[2.0], $MachinePrecision] * x$95$m), $MachinePrecision]

Derivation

1. Initial program 53.9%

   \[\sqrt{2 \cdot \left(x \cdot x\right)} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-sqrt.f64 N/A

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

   2. lift-*.f64 N/A

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

   3. sqrt-prod N/A

      \[\leadsto \color{blue}{\sqrt{2} \cdot \sqrt{x \cdot x}} \]

   4. pow1/2 N/A

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

   5. lift-*.f64 N/A

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

   6. sqrt-prod N/A

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

   7. rem-square-sqrt N/A

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

   8. lower-*.f64 N/A

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

   9. pow1/2 N/A

      \[\leadsto \color{blue}{\sqrt{2}} \cdot x \]

   10. lower-sqrt.f64 50.4

       \[\leadsto \color{blue}{\sqrt{2}} \cdot x \]

4. Applied rewrites 50.4%

   \[\leadsto \color{blue}{\sqrt{2} \cdot x} \]
5. Add Preprocessing

Alternative 5: 5.4% accurate, 1.9× speedup

Math

\[\begin{array}{l} x_m = \left|x\right| \\ \sqrt{2} \end{array} \]

FPCore

x_m = (fabs.f64 x)
(FPCore (x_m) :precision binary64 (sqrt 2.0))

C

x_m = fabs(x);
double code(double x_m) {
	return sqrt(2.0);
}

Fortran

x_m = abs(x)
real(8) function code(x_m)
    real(8), intent (in) :: x_m
    code = sqrt(2.0d0)
end function

Java

x_m = Math.abs(x);
public static double code(double x_m) {
	return Math.sqrt(2.0);
}

Python

x_m = math.fabs(x)
def code(x_m):
	return math.sqrt(2.0)

Julia

x_m = abs(x)
function code(x_m)
	return sqrt(2.0)
end

MATLAB

x_m = abs(x);
function tmp = code(x_m)
	tmp = sqrt(2.0);
end

Wolfram

x_m = N[Abs[x], $MachinePrecision]
code[x$95$m_] := N[Sqrt[2.0], $MachinePrecision]

Derivation

1. Initial program 53.9%

   \[\sqrt{2 \cdot \left(x \cdot x\right)} \]
2. Add Preprocessing
3. Step-by-step derivation

   1. lift-sqrt.f64 N/A

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

   2. pow1/2 N/A

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

   3. lift-*.f64 N/A

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

   4. lift-*.f64 N/A

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

   5. associate-*r* N/A

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

   6. unpow-prod-down N/A

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

   7. lower-*.f64 N/A

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

   8. pow1/2 N/A

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

   9. lower-sqrt.f64 N/A

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

   10. *-commutative N/A

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

   11. lower-*.f64 N/A

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

   12. pow1/2 N/A

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

   13. lower-sqrt.f64 49.4

       \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

4. Applied rewrites 49.4%

   \[\leadsto \color{blue}{\sqrt{x \cdot 2} \cdot \sqrt{x}} \]
5. Step-by-step derivation

   1. lift-sqrt.f64 N/A

      \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{\sqrt{x}} \]

   2. pow1/2 N/A

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

   3. lower-pow.f64 49.4

      \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{{x}^{0.5}} \]

6. Applied rewrites 49.4%

   \[\leadsto \sqrt{x \cdot 2} \cdot \color{blue}{{x}^{0.5}} \]

8. Applied rewrites 5.4%

   \[\leadsto \color{blue}{\sqrt{2}} \]
9. Add Preprocessing

Reproduce

herbie shell --seed 2024308 
(FPCore (x)
  :name "sqrt C (should all be same)"
  :precision binary64
  (sqrt (* 2.0 (* x x))))