Hyperbolic arc-(co)tangent

Percentage Accurate: 8.4% → 100.0%

Time: 9.5s

Alternatives: 6

Speedup: 14.9×

Specification

Math

\[\begin{array}{l} \\ \frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \end{array} \]

FPCore

(FPCore (x) :precision binary64 (* (/ 1.0 2.0) (log (/ (+ 1.0 x) (- 1.0 x)))))

C

double code(double x) {
	return (1.0 / 2.0) * log(((1.0 + x) / (1.0 - x)));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = (1.0d0 / 2.0d0) * log(((1.0d0 + x) / (1.0d0 - x)))
end function

Java

public static double code(double x) {
	return (1.0 / 2.0) * Math.log(((1.0 + x) / (1.0 - x)));
}

Python

def code(x):
	return (1.0 / 2.0) * math.log(((1.0 + x) / (1.0 - x)))

Julia

function code(x)
	return Float64(Float64(1.0 / 2.0) * log(Float64(Float64(1.0 + x) / Float64(1.0 - x))))
end

MATLAB

function tmp = code(x)
	tmp = (1.0 / 2.0) * log(((1.0 + x) / (1.0 - x)));
end

Wolfram

code[x_] := N[(N[(1.0 / 2.0), $MachinePrecision] * N[Log[N[(N[(1.0 + x), $MachinePrecision] / N[(1.0 - x), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]

Initial Program: 8.4% accurate, 1.0× speedup

Math

\[\begin{array}{l} \\ \frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \end{array} \]

FPCore

(FPCore (x) :precision binary64 (* (/ 1.0 2.0) (log (/ (+ 1.0 x) (- 1.0 x)))))

C

double code(double x) {
	return (1.0 / 2.0) * log(((1.0 + x) / (1.0 - x)));
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = (1.0d0 / 2.0d0) * log(((1.0d0 + x) / (1.0d0 - x)))
end function

Java

public static double code(double x) {
	return (1.0 / 2.0) * Math.log(((1.0 + x) / (1.0 - x)));
}

Python

def code(x):
	return (1.0 / 2.0) * math.log(((1.0 + x) / (1.0 - x)))

Julia

function code(x)
	return Float64(Float64(1.0 / 2.0) * log(Float64(Float64(1.0 + x) / Float64(1.0 - x))))
end

MATLAB

function tmp = code(x)
	tmp = (1.0 / 2.0) * log(((1.0 + x) / (1.0 - x)));
end

Wolfram

code[x_] := N[(N[(1.0 / 2.0), $MachinePrecision] * N[Log[N[(N[(1.0 + x), $MachinePrecision] / N[(1.0 - x), $MachinePrecision]), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]

Alternative 1: 100.0% accurate, 0.6× speedup

Math

\[\begin{array}{l} \\ 0.5 \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (* 0.5 (- (* 2.0 (log1p x)) (log1p (* (- x) x)))))

C

double code(double x) {
	return 0.5 * ((2.0 * log1p(x)) - log1p((-x * x)));
}

Java

public static double code(double x) {
	return 0.5 * ((2.0 * Math.log1p(x)) - Math.log1p((-x * x)));
}

Python

def code(x):
	return 0.5 * ((2.0 * math.log1p(x)) - math.log1p((-x * x)))

Julia

function code(x)
	return Float64(0.5 * Float64(Float64(2.0 * log1p(x)) - log1p(Float64(Float64(-x) * x))))
end

Wolfram

code[x_] := N[(0.5 * N[(N[(2.0 * N[Log[1 + x], $MachinePrecision]), $MachinePrecision] - N[Log[1 + N[((-x) * x), $MachinePrecision]], $MachinePrecision]), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

4. Applied rewrites 100.0%

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right)} \]
5. Step-by-step derivation

   1. lift-/.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{2}} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]

   2. metadata-eval 100.0

      \[\leadsto \color{blue}{0.5} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]

6. Applied rewrites 100.0%

   \[\leadsto \color{blue}{0.5} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]
7. Add Preprocessing

Alternative 2: 99.8% accurate, 0.8× speedup

Math

\[\begin{array}{l} \\ \left(\mathsf{log1p}\left(x\right) \cdot 2 - \left(\left(\left(\left(\left(-0.25 \cdot x\right) \cdot x - 0.3333333333333333\right) \cdot x\right) \cdot x - 0.5\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \cdot 0.5 \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (*
  (-
   (* (log1p x) 2.0)
   (*
    (-
     (* (- (* (* (- (* (* -0.25 x) x) 0.3333333333333333) x) x) 0.5) (* x x))
     1.0)
    (* x x)))
  0.5))

C

double code(double x) {
	return ((log1p(x) * 2.0) - (((((((((-0.25 * x) * x) - 0.3333333333333333) * x) * x) - 0.5) * (x * x)) - 1.0) * (x * x))) * 0.5;
}

Java

public static double code(double x) {
	return ((Math.log1p(x) * 2.0) - (((((((((-0.25 * x) * x) - 0.3333333333333333) * x) * x) - 0.5) * (x * x)) - 1.0) * (x * x))) * 0.5;
}

Python

def code(x):
	return ((math.log1p(x) * 2.0) - (((((((((-0.25 * x) * x) - 0.3333333333333333) * x) * x) - 0.5) * (x * x)) - 1.0) * (x * x))) * 0.5

Julia

function code(x)
	return Float64(Float64(Float64(log1p(x) * 2.0) - Float64(Float64(Float64(Float64(Float64(Float64(Float64(Float64(Float64(-0.25 * x) * x) - 0.3333333333333333) * x) * x) - 0.5) * Float64(x * x)) - 1.0) * Float64(x * x))) * 0.5)
end

Wolfram

code[x_] := N[(N[(N[(N[Log[1 + x], $MachinePrecision] * 2.0), $MachinePrecision] - N[(N[(N[(N[(N[(N[(N[(N[(N[(-0.25 * x), $MachinePrecision] * x), $MachinePrecision] - 0.3333333333333333), $MachinePrecision] * x), $MachinePrecision] * x), $MachinePrecision] - 0.5), $MachinePrecision] * N[(x * x), $MachinePrecision]), $MachinePrecision] - 1.0), $MachinePrecision] * N[(x * x), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * 0.5), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

4. Applied rewrites 100.0%

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right)} \]

5. Taylor expanded in x around 0

   \[\leadsto \frac{1}{2} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \color{blue}{{x}^{2} \cdot \left({x}^{2} \cdot \left({x}^{2} \cdot \left(\frac{-1}{4} \cdot {x}^{2} - \frac{1}{3}\right) - \frac{1}{2}\right) - 1\right)}\right) \]
6. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \color{blue}{\left({x}^{2} \cdot \left({x}^{2} \cdot \left(\frac{-1}{4} \cdot {x}^{2} - \frac{1}{3}\right) - \frac{1}{2}\right) - 1\right) \cdot {x}^{2}}\right) \]

   2. lower-*.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \color{blue}{\left({x}^{2} \cdot \left({x}^{2} \cdot \left(\frac{-1}{4} \cdot {x}^{2} - \frac{1}{3}\right) - \frac{1}{2}\right) - 1\right) \cdot {x}^{2}}\right) \]

7. Applied rewrites 99.6%

   \[\leadsto \frac{1}{2} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \color{blue}{\left(\left(\left(\left(-0.25 \cdot x\right) \cdot x - 0.3333333333333333\right) \cdot \left(x \cdot x\right) - 0.5\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)}\right) \]
8. Step-by-step derivation

   1. lift-*.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{2} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \left(\left(\left(\left(\frac{-1}{4} \cdot x\right) \cdot x - \frac{1}{3}\right) \cdot \left(x \cdot x\right) - \frac{1}{2}\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right)} \]

   2. lift-/.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{2}} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \left(\left(\left(\left(\frac{-1}{4} \cdot x\right) \cdot x - \frac{1}{3}\right) \cdot \left(x \cdot x\right) - \frac{1}{2}\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \]

   3. metadata-eval N/A

      \[\leadsto \color{blue}{\frac{1}{2}} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \left(\left(\left(\left(\frac{-1}{4} \cdot x\right) \cdot x - \frac{1}{3}\right) \cdot \left(x \cdot x\right) - \frac{1}{2}\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \]

   4. *-commutative N/A

      \[\leadsto \color{blue}{\left(2 \cdot \mathsf{log1p}\left(x\right) - \left(\left(\left(\left(\frac{-1}{4} \cdot x\right) \cdot x - \frac{1}{3}\right) \cdot \left(x \cdot x\right) - \frac{1}{2}\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \cdot \frac{1}{2}} \]

   5. lower-*.f64 99.6

      \[\leadsto \color{blue}{\left(2 \cdot \mathsf{log1p}\left(x\right) - \left(\left(\left(\left(-0.25 \cdot x\right) \cdot x - 0.3333333333333333\right) \cdot \left(x \cdot x\right) - 0.5\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \cdot 0.5} \]

9. Applied rewrites 99.6%

   \[\leadsto \color{blue}{\left(\mathsf{log1p}\left(x\right) \cdot 2 - \left(\left(\left(\left(\left(-0.25 \cdot x\right) \cdot x - 0.3333333333333333\right) \cdot x\right) \cdot x - 0.5\right) \cdot \left(x \cdot x\right) - 1\right) \cdot \left(x \cdot x\right)\right) \cdot 0.5} \]
10. Add Preprocessing

Alternative 3: 99.8% accurate, 3.0× speedup

Math

\[\begin{array}{l} \\ 0.5 \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(0.2857142857142857, x \cdot x, 0.4\right), x \cdot x, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right) \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (*
  0.5
  (*
   (fma
    (fma (fma 0.2857142857142857 (* x x) 0.4) (* x x) 0.6666666666666666)
    (* x x)
    2.0)
   x)))

C

double code(double x) {
	return 0.5 * (fma(fma(fma(0.2857142857142857, (x * x), 0.4), (x * x), 0.6666666666666666), (x * x), 2.0) * x);
}

Julia

function code(x)
	return Float64(0.5 * Float64(fma(fma(fma(0.2857142857142857, Float64(x * x), 0.4), Float64(x * x), 0.6666666666666666), Float64(x * x), 2.0) * x))
end

Wolfram

code[x_] := N[(0.5 * N[(N[(N[(N[(0.2857142857142857 * N[(x * x), $MachinePrecision] + 0.4), $MachinePrecision] * N[(x * x), $MachinePrecision] + 0.6666666666666666), $MachinePrecision] * N[(x * x), $MachinePrecision] + 2.0), $MachinePrecision] * x), $MachinePrecision]), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

4. Applied rewrites 100.0%

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right)} \]
5. Step-by-step derivation

   1. lift-/.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{2}} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]

   2. metadata-eval 100.0

      \[\leadsto \color{blue}{0.5} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]

6. Applied rewrites 100.0%

   \[\leadsto \color{blue}{0.5} \cdot \left(2 \cdot \mathsf{log1p}\left(x\right) - \mathsf{log1p}\left(\left(-x\right) \cdot x\right)\right) \]

7. Taylor expanded in x around 0

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(x \cdot \left(2 + {x}^{2} \cdot \left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right)\right)\right)\right)} \]
8. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(\left(2 + {x}^{2} \cdot \left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right)\right)\right) \cdot x\right)} \]

   2. lower-*.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(\left(2 + {x}^{2} \cdot \left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right)\right)\right) \cdot x\right)} \]

   3. +-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\color{blue}{\left({x}^{2} \cdot \left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right)\right) + 2\right)} \cdot x\right) \]

   4. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\left(\color{blue}{\left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right)\right) \cdot {x}^{2}} + 2\right) \cdot x\right) \]

   5. lower-fma.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\color{blue}{\mathsf{fma}\left(\frac{2}{3} + {x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right), {x}^{2}, 2\right)} \cdot x\right) \]

   6. +-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\color{blue}{{x}^{2} \cdot \left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right) + \frac{2}{3}}, {x}^{2}, 2\right) \cdot x\right) \]

   7. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\color{blue}{\left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}\right) \cdot {x}^{2}} + \frac{2}{3}, {x}^{2}, 2\right) \cdot x\right) \]

   8. lower-fma.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(\frac{2}{5} + \frac{2}{7} \cdot {x}^{2}, {x}^{2}, \frac{2}{3}\right)}, {x}^{2}, 2\right) \cdot x\right) \]

   9. +-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\color{blue}{\frac{2}{7} \cdot {x}^{2} + \frac{2}{5}}, {x}^{2}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   10. lower-fma.f64 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(\frac{2}{7}, {x}^{2}, \frac{2}{5}\right)}, {x}^{2}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   11. unpow2 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{7}, \color{blue}{x \cdot x}, \frac{2}{5}\right), {x}^{2}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   12. lower-*.f64 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{7}, \color{blue}{x \cdot x}, \frac{2}{5}\right), {x}^{2}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   13. unpow2 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{7}, x \cdot x, \frac{2}{5}\right), \color{blue}{x \cdot x}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   14. lower-*.f64 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{7}, x \cdot x, \frac{2}{5}\right), \color{blue}{x \cdot x}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   15. unpow2 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{7}, x \cdot x, \frac{2}{5}\right), x \cdot x, \frac{2}{3}\right), \color{blue}{x \cdot x}, 2\right) \cdot x\right) \]

   16. lower-*.f64 99.6

       \[\leadsto 0.5 \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(0.2857142857142857, x \cdot x, 0.4\right), x \cdot x, 0.6666666666666666\right), \color{blue}{x \cdot x}, 2\right) \cdot x\right) \]

9. Applied rewrites 99.6%

   \[\leadsto 0.5 \cdot \color{blue}{\left(\mathsf{fma}\left(\mathsf{fma}\left(\mathsf{fma}\left(0.2857142857142857, x \cdot x, 0.4\right), x \cdot x, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right)} \]
10. Add Preprocessing

Alternative 4: 99.7% accurate, 4.1× speedup

Math

\[\begin{array}{l} \\ \left(\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, 0.4, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right) \cdot 0.5 \end{array} \]

FPCore

(FPCore (x)
 :precision binary64
 (* (* (fma (fma (* x x) 0.4 0.6666666666666666) (* x x) 2.0) x) 0.5))

C

double code(double x) {
	return (fma(fma((x * x), 0.4, 0.6666666666666666), (x * x), 2.0) * x) * 0.5;
}

Julia

function code(x)
	return Float64(Float64(fma(fma(Float64(x * x), 0.4, 0.6666666666666666), Float64(x * x), 2.0) * x) * 0.5)
end

Wolfram

code[x_] := N[(N[(N[(N[(N[(x * x), $MachinePrecision] * 0.4 + 0.6666666666666666), $MachinePrecision] * N[(x * x), $MachinePrecision] + 2.0), $MachinePrecision] * x), $MachinePrecision] * 0.5), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(x \cdot \left(2 + {x}^{2} \cdot \left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}\right)\right)\right)} \]
4. Step-by-step derivation

   1. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(\left(2 + {x}^{2} \cdot \left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}\right)\right) \cdot x\right)} \]

   2. lower-*.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(\left(2 + {x}^{2} \cdot \left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}\right)\right) \cdot x\right)} \]

   3. +-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\color{blue}{\left({x}^{2} \cdot \left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}\right) + 2\right)} \cdot x\right) \]

   4. *-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\left(\color{blue}{\left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}\right) \cdot {x}^{2}} + 2\right) \cdot x\right) \]

   5. lower-fma.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\color{blue}{\mathsf{fma}\left(\frac{2}{3} + \frac{2}{5} \cdot {x}^{2}, {x}^{2}, 2\right)} \cdot x\right) \]

   6. +-commutative N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\color{blue}{\frac{2}{5} \cdot {x}^{2} + \frac{2}{3}}, {x}^{2}, 2\right) \cdot x\right) \]

   7. lower-fma.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\color{blue}{\mathsf{fma}\left(\frac{2}{5}, {x}^{2}, \frac{2}{3}\right)}, {x}^{2}, 2\right) \cdot x\right) \]

   8. unpow2 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{5}, \color{blue}{x \cdot x}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   9. lower-*.f64 N/A

      \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{5}, \color{blue}{x \cdot x}, \frac{2}{3}\right), {x}^{2}, 2\right) \cdot x\right) \]

   10. unpow2 N/A

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{5}, x \cdot x, \frac{2}{3}\right), \color{blue}{x \cdot x}, 2\right) \cdot x\right) \]

   11. lower-*.f64 99.4

       \[\leadsto \frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(0.4, x \cdot x, 0.6666666666666666\right), \color{blue}{x \cdot x}, 2\right) \cdot x\right) \]

5. Applied rewrites 99.4%

   \[\leadsto \frac{1}{2} \cdot \color{blue}{\left(\mathsf{fma}\left(\mathsf{fma}\left(0.4, x \cdot x, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right)} \]
6. Step-by-step derivation

   1. lift-*.f64 N/A

      \[\leadsto \color{blue}{\frac{1}{2} \cdot \left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{5}, x \cdot x, \frac{2}{3}\right), x \cdot x, 2\right) \cdot x\right)} \]

   2. *-commutative N/A

      \[\leadsto \color{blue}{\left(\mathsf{fma}\left(\mathsf{fma}\left(\frac{2}{5}, x \cdot x, \frac{2}{3}\right), x \cdot x, 2\right) \cdot x\right) \cdot \frac{1}{2}} \]

   3. lower-*.f64 99.4

      \[\leadsto \color{blue}{\left(\mathsf{fma}\left(\mathsf{fma}\left(0.4, x \cdot x, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right) \cdot \frac{1}{2}} \]

   4. lift-/.f64 N/A

      \[\leadsto \left(\mathsf{fma}\left(\mathsf{Rewrite=>}\left(lower-fma.f64, \left(\mathsf{fma}\left(x \cdot x, \frac{2}{5}, \frac{2}{3}\right)\right)\right), x \cdot x, 2\right) \cdot x\right) \cdot \color{blue}{\frac{1}{2}} \]

   5. metadata-eval N/A

      \[\leadsto \left(\mathsf{fma}\left(\mathsf{Rewrite=>}\left(lower-fma.f64, \left(\mathsf{fma}\left(x \cdot x, \frac{2}{5}, \frac{2}{3}\right)\right)\right), x \cdot x, 2\right) \cdot x\right) \cdot \color{blue}{\frac{1}{2}} \]

7. Applied rewrites 99.4%

   \[\leadsto \color{blue}{\left(\mathsf{fma}\left(\mathsf{fma}\left(x \cdot x, 0.4, 0.6666666666666666\right), x \cdot x, 2\right) \cdot x\right) \cdot 0.5} \]
8. Add Preprocessing

Alternative 5: 99.5% accurate, 7.9× speedup

Math

\[\begin{array}{l} \\ \mathsf{fma}\left(\left(x \cdot x\right) \cdot x, 0.3333333333333333, x\right) \end{array} \]

FPCore

(FPCore (x) :precision binary64 (fma (* (* x x) x) 0.3333333333333333 x))

C

double code(double x) {
	return fma(((x * x) * x), 0.3333333333333333, x);
}

Julia

function code(x)
	return fma(Float64(Float64(x * x) * x), 0.3333333333333333, x)
end

Wolfram

code[x_] := N[(N[(N[(x * x), $MachinePrecision] * x), $MachinePrecision] * 0.3333333333333333 + x), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

3. Taylor expanded in x around 0

   \[\leadsto \color{blue}{x \cdot \left(1 + \frac{1}{3} \cdot {x}^{2}\right)} \]
4. Step-by-step derivation

   1. +-commutative N/A

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

   2. distribute-lft-in N/A

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

   3. *-commutative N/A

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

   4. associate-*r* N/A

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

   5. unpow2 N/A

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

   6. cube-mult N/A

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

   7. *-rgt-identity N/A

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

   8. lower-fma.f64 N/A

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

   9. lower-pow.f64 99.3

      \[\leadsto \mathsf{fma}\left(\color{blue}{{x}^{3}}, 0.3333333333333333, x\right) \]

5. Applied rewrites 99.3%

   \[\leadsto \color{blue}{\mathsf{fma}\left({x}^{3}, 0.3333333333333333, x\right)} \]
6. Step-by-step derivation

7. Applied rewrites 99.3%

   \[\leadsto \mathsf{fma}\left(\left(x \cdot x\right) \cdot x, 0.3333333333333333, x\right) \]
8. Add Preprocessing

Alternative 6: 99.0% accurate, 14.9× speedup

Math

\[\begin{array}{l} \\ \left(x + x\right) \cdot 0.5 \end{array} \]

FPCore

(FPCore (x) :precision binary64 (* (+ x x) 0.5))

C

double code(double x) {
	return (x + x) * 0.5;
}

Fortran

real(8) function code(x)
    real(8), intent (in) :: x
    code = (x + x) * 0.5d0
end function

Java

public static double code(double x) {
	return (x + x) * 0.5;
}

Python

def code(x):
	return (x + x) * 0.5

Julia

function code(x)
	return Float64(Float64(x + x) * 0.5)
end

MATLAB

function tmp = code(x)
	tmp = (x + x) * 0.5;
end

Wolfram

code[x_] := N[(N[(x + x), $MachinePrecision] * 0.5), $MachinePrecision]

Derivation

1. Initial program 9.2%

   \[\frac{1}{2} \cdot \log \left(\frac{1 + x}{1 - x}\right) \]
2. Add Preprocessing

3. Taylor expanded in x around 0

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

   1. lower-*.f64 98.8

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

5. Applied rewrites 98.8%

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

   1. lift-*.f64 N/A

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

   2. *-commutative N/A

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

   3. lower-*.f64 98.8

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

   4. lift-/.f64 N/A

      \[\leadsto \mathsf{Rewrite=>}\left(lower-*.f64, \left(x \cdot 2\right)\right) \cdot \color{blue}{\frac{1}{2}} \]

   5. metadata-eval N/A

      \[\leadsto \mathsf{Rewrite=>}\left(lower-*.f64, \left(x \cdot 2\right)\right) \cdot \color{blue}{\frac{1}{2}} \]

7. Applied rewrites 98.8%

   \[\leadsto \color{blue}{\left(x \cdot 2\right) \cdot 0.5} \]
8. Step-by-step derivation

9. Applied rewrites 98.8%

   \[\leadsto \left(x + \color{blue}{x}\right) \cdot 0.5 \]
10. Add Preprocessing

Reproduce

herbie shell --seed 2024338 
(FPCore (x)
  :name "Hyperbolic arc-(co)tangent"
  :precision binary64
  (* (/ 1.0 2.0) (log (/ (+ 1.0 x) (- 1.0 x)))))