
(FPCore (x) :precision binary64 (/ (- x (sin x)) (tan x)))
double code(double x) {
return (x - sin(x)) / tan(x);
}
real(8) function code(x)
real(8), intent (in) :: x
code = (x - sin(x)) / tan(x)
end function
public static double code(double x) {
return (x - Math.sin(x)) / Math.tan(x);
}
def code(x): return (x - math.sin(x)) / math.tan(x)
function code(x) return Float64(Float64(x - sin(x)) / tan(x)) end
function tmp = code(x) tmp = (x - sin(x)) / tan(x); end
code[x_] := N[(N[(x - N[Sin[x], $MachinePrecision]), $MachinePrecision] / N[Tan[x], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
\frac{x - \sin x}{\tan x}
\end{array}
Sampling outcomes in binary64 precision:
Herbie found 6 alternatives:
| Alternative | Accuracy | Speedup |
|---|
(FPCore (x) :precision binary64 (/ (- x (sin x)) (tan x)))
double code(double x) {
return (x - sin(x)) / tan(x);
}
real(8) function code(x)
real(8), intent (in) :: x
code = (x - sin(x)) / tan(x)
end function
public static double code(double x) {
return (x - Math.sin(x)) / Math.tan(x);
}
def code(x): return (x - math.sin(x)) / math.tan(x)
function code(x) return Float64(Float64(x - sin(x)) / tan(x)) end
function tmp = code(x) tmp = (x - sin(x)) / tan(x); end
code[x_] := N[(N[(x - N[Sin[x], $MachinePrecision]), $MachinePrecision] / N[Tan[x], $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
\frac{x - \sin x}{\tan x}
\end{array}
(FPCore (x)
:precision binary64
(let* ((t_0 (fma (pow x 2.0) -0.0007275132275132275 -0.06388888888888888)))
(*
(- 0.027777777777777776 (* (pow x 4.0) (pow t_0 2.0)))
(/ (pow x 2.0) (- 0.16666666666666666 (* (pow x 2.0) t_0))))))
double code(double x) {
double t_0 = fma(pow(x, 2.0), -0.0007275132275132275, -0.06388888888888888);
return (0.027777777777777776 - (pow(x, 4.0) * pow(t_0, 2.0))) * (pow(x, 2.0) / (0.16666666666666666 - (pow(x, 2.0) * t_0)));
}
function code(x) t_0 = fma((x ^ 2.0), -0.0007275132275132275, -0.06388888888888888) return Float64(Float64(0.027777777777777776 - Float64((x ^ 4.0) * (t_0 ^ 2.0))) * Float64((x ^ 2.0) / Float64(0.16666666666666666 - Float64((x ^ 2.0) * t_0)))) end
code[x_] := Block[{t$95$0 = N[(N[Power[x, 2.0], $MachinePrecision] * -0.0007275132275132275 + -0.06388888888888888), $MachinePrecision]}, N[(N[(0.027777777777777776 - N[(N[Power[x, 4.0], $MachinePrecision] * N[Power[t$95$0, 2.0], $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[(N[Power[x, 2.0], $MachinePrecision] / N[(0.16666666666666666 - N[(N[Power[x, 2.0], $MachinePrecision] * t$95$0), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]]
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \mathsf{fma}\left({x}^{2}, -0.0007275132275132275, -0.06388888888888888\right)\\
\left(0.027777777777777776 - {x}^{4} \cdot {t\_0}^{2}\right) \cdot \frac{{x}^{2}}{0.16666666666666666 - {x}^{2} \cdot t\_0}
\end{array}
\end{array}
Initial program 53.2%
Taylor expanded in x around 0 99.4%
flip-+99.4%
associate-*r/99.4%
Applied egg-rr99.4%
*-commutative99.4%
associate-/l*99.5%
Simplified99.5%
(FPCore (x)
:precision binary64
(*
(/
(pow x 2.0)
(-
0.16666666666666666
(*
(pow x 2.0)
(fma (pow x 2.0) -0.0007275132275132275 -0.06388888888888888))))
(- 0.027777777777777776 (* (pow x 4.0) 0.00408179012345679))))
double code(double x) {
return (pow(x, 2.0) / (0.16666666666666666 - (pow(x, 2.0) * fma(pow(x, 2.0), -0.0007275132275132275, -0.06388888888888888)))) * (0.027777777777777776 - (pow(x, 4.0) * 0.00408179012345679));
}
function code(x) return Float64(Float64((x ^ 2.0) / Float64(0.16666666666666666 - Float64((x ^ 2.0) * fma((x ^ 2.0), -0.0007275132275132275, -0.06388888888888888)))) * Float64(0.027777777777777776 - Float64((x ^ 4.0) * 0.00408179012345679))) end
code[x_] := N[(N[(N[Power[x, 2.0], $MachinePrecision] / N[(0.16666666666666666 - N[(N[Power[x, 2.0], $MachinePrecision] * N[(N[Power[x, 2.0], $MachinePrecision] * -0.0007275132275132275 + -0.06388888888888888), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision] * N[(0.027777777777777776 - N[(N[Power[x, 4.0], $MachinePrecision] * 0.00408179012345679), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
\frac{{x}^{2}}{0.16666666666666666 - {x}^{2} \cdot \mathsf{fma}\left({x}^{2}, -0.0007275132275132275, -0.06388888888888888\right)} \cdot \left(0.027777777777777776 - {x}^{4} \cdot 0.00408179012345679\right)
\end{array}
Initial program 53.2%
Taylor expanded in x around 0 99.4%
flip-+99.4%
associate-*r/99.4%
Applied egg-rr99.4%
*-commutative99.4%
associate-/l*99.5%
Simplified99.5%
Taylor expanded in x around 0 99.5%
Final simplification99.5%
(FPCore (x)
:precision binary64
(*
(- 0.027777777777777776 (* (pow x 4.0) 0.00408179012345679))
(*
x
(/
x
(-
0.16666666666666666
(*
(pow x 2.0)
(fma (pow x 2.0) -0.0007275132275132275 -0.06388888888888888)))))))
double code(double x) {
return (0.027777777777777776 - (pow(x, 4.0) * 0.00408179012345679)) * (x * (x / (0.16666666666666666 - (pow(x, 2.0) * fma(pow(x, 2.0), -0.0007275132275132275, -0.06388888888888888)))));
}
function code(x) return Float64(Float64(0.027777777777777776 - Float64((x ^ 4.0) * 0.00408179012345679)) * Float64(x * Float64(x / Float64(0.16666666666666666 - Float64((x ^ 2.0) * fma((x ^ 2.0), -0.0007275132275132275, -0.06388888888888888)))))) end
code[x_] := N[(N[(0.027777777777777776 - N[(N[Power[x, 4.0], $MachinePrecision] * 0.00408179012345679), $MachinePrecision]), $MachinePrecision] * N[(x * N[(x / N[(0.16666666666666666 - N[(N[Power[x, 2.0], $MachinePrecision] * N[(N[Power[x, 2.0], $MachinePrecision] * -0.0007275132275132275 + -0.06388888888888888), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
\left(0.027777777777777776 - {x}^{4} \cdot 0.00408179012345679\right) \cdot \left(x \cdot \frac{x}{0.16666666666666666 - {x}^{2} \cdot \mathsf{fma}\left({x}^{2}, -0.0007275132275132275, -0.06388888888888888\right)}\right)
\end{array}
Initial program 53.2%
Taylor expanded in x around 0 99.4%
flip-+99.4%
associate-*r/99.4%
Applied egg-rr99.4%
*-commutative99.4%
associate-/l*99.5%
Simplified99.5%
Taylor expanded in x around 0 99.5%
pow299.5%
*-un-lft-identity99.5%
times-frac99.4%
Applied egg-rr99.4%
Final simplification99.4%
(FPCore (x) :precision binary64 (* (pow x 2.0) (+ 0.16666666666666666 (* (* x x) (- (* -0.0007275132275132275 (* x x)) 0.06388888888888888)))))
double code(double x) {
return pow(x, 2.0) * (0.16666666666666666 + ((x * x) * ((-0.0007275132275132275 * (x * x)) - 0.06388888888888888)));
}
real(8) function code(x)
real(8), intent (in) :: x
code = (x ** 2.0d0) * (0.16666666666666666d0 + ((x * x) * (((-0.0007275132275132275d0) * (x * x)) - 0.06388888888888888d0)))
end function
public static double code(double x) {
return Math.pow(x, 2.0) * (0.16666666666666666 + ((x * x) * ((-0.0007275132275132275 * (x * x)) - 0.06388888888888888)));
}
def code(x): return math.pow(x, 2.0) * (0.16666666666666666 + ((x * x) * ((-0.0007275132275132275 * (x * x)) - 0.06388888888888888)))
function code(x) return Float64((x ^ 2.0) * Float64(0.16666666666666666 + Float64(Float64(x * x) * Float64(Float64(-0.0007275132275132275 * Float64(x * x)) - 0.06388888888888888)))) end
function tmp = code(x) tmp = (x ^ 2.0) * (0.16666666666666666 + ((x * x) * ((-0.0007275132275132275 * (x * x)) - 0.06388888888888888))); end
code[x_] := N[(N[Power[x, 2.0], $MachinePrecision] * N[(0.16666666666666666 + N[(N[(x * x), $MachinePrecision] * N[(N[(-0.0007275132275132275 * N[(x * x), $MachinePrecision]), $MachinePrecision] - 0.06388888888888888), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
{x}^{2} \cdot \left(0.16666666666666666 + \left(x \cdot x\right) \cdot \left(-0.0007275132275132275 \cdot \left(x \cdot x\right) - 0.06388888888888888\right)\right)
\end{array}
Initial program 53.2%
Taylor expanded in x around 0 99.4%
unpow299.4%
Applied egg-rr99.4%
unpow299.4%
Applied egg-rr99.4%
(FPCore (x) :precision binary64 (* 0.16666666666666666 (* x x)))
double code(double x) {
return 0.16666666666666666 * (x * x);
}
real(8) function code(x)
real(8), intent (in) :: x
code = 0.16666666666666666d0 * (x * x)
end function
public static double code(double x) {
return 0.16666666666666666 * (x * x);
}
def code(x): return 0.16666666666666666 * (x * x)
function code(x) return Float64(0.16666666666666666 * Float64(x * x)) end
function tmp = code(x) tmp = 0.16666666666666666 * (x * x); end
code[x_] := N[(0.16666666666666666 * N[(x * x), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
0.16666666666666666 \cdot \left(x \cdot x\right)
\end{array}
Initial program 53.2%
Taylor expanded in x around 0 99.4%
Taylor expanded in x around 0 98.7%
*-commutative98.7%
Simplified98.7%
unpow299.4%
Applied egg-rr98.7%
Final simplification98.7%
(FPCore (x) :precision binary64 1.0)
double code(double x) {
return 1.0;
}
real(8) function code(x)
real(8), intent (in) :: x
code = 1.0d0
end function
public static double code(double x) {
return 1.0;
}
def code(x): return 1.0
function code(x) return 1.0 end
function tmp = code(x) tmp = 1.0; end
code[x_] := 1.0
\begin{array}{l}
\\
1
\end{array}
Initial program 53.2%
Taylor expanded in x around inf 4.2%
Taylor expanded in x around 0 4.2%
(FPCore (x) :precision binary64 (* 0.16666666666666666 (* x x)))
double code(double x) {
return 0.16666666666666666 * (x * x);
}
real(8) function code(x)
real(8), intent (in) :: x
code = 0.16666666666666666d0 * (x * x)
end function
public static double code(double x) {
return 0.16666666666666666 * (x * x);
}
def code(x): return 0.16666666666666666 * (x * x)
function code(x) return Float64(0.16666666666666666 * Float64(x * x)) end
function tmp = code(x) tmp = 0.16666666666666666 * (x * x); end
code[x_] := N[(0.16666666666666666 * N[(x * x), $MachinePrecision]), $MachinePrecision]
\begin{array}{l}
\\
0.16666666666666666 \cdot \left(x \cdot x\right)
\end{array}
herbie shell --seed 2024114
(FPCore (x)
:name "ENA, Section 1.4, Exercise 4a"
:precision binary64
:pre (and (<= -1.0 x) (<= x 1.0))
:alt
(! :herbie-platform default (* 1/6 (* x x)))
(/ (- x (sin x)) (tan x)))