
(FPCore (cosTheta_i u1 u2) :precision binary32 (* (sqrt (/ u1 (- 1.0 u1))) (cos (* 6.28318530718 u2))))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf((u1 / (1.0f - u1))) * cosf((6.28318530718f * u2));
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = sqrt((u1 / (1.0e0 - u1))) * cos((6.28318530718e0 * u2))
end function
function code(cosTheta_i, u1, u2) return Float32(sqrt(Float32(u1 / Float32(Float32(1.0) - u1))) * cos(Float32(Float32(6.28318530718) * u2))) end
function tmp = code(cosTheta_i, u1, u2) tmp = sqrt((u1 / (single(1.0) - u1))) * cos((single(6.28318530718) * u2)); end
\begin{array}{l}
\\
\sqrt{\frac{u1}{1 - u1}} \cdot \cos \left(6.28318530718 \cdot u2\right)
\end{array}
Sampling outcomes in binary32 precision:
Herbie found 8 alternatives:
| Alternative | Accuracy | Speedup |
|---|
(FPCore (cosTheta_i u1 u2) :precision binary32 (* (sqrt (/ u1 (- 1.0 u1))) (cos (* 6.28318530718 u2))))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf((u1 / (1.0f - u1))) * cosf((6.28318530718f * u2));
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = sqrt((u1 / (1.0e0 - u1))) * cos((6.28318530718e0 * u2))
end function
function code(cosTheta_i, u1, u2) return Float32(sqrt(Float32(u1 / Float32(Float32(1.0) - u1))) * cos(Float32(Float32(6.28318530718) * u2))) end
function tmp = code(cosTheta_i, u1, u2) tmp = sqrt((u1 / (single(1.0) - u1))) * cos((single(6.28318530718) * u2)); end
\begin{array}{l}
\\
\sqrt{\frac{u1}{1 - u1}} \cdot \cos \left(6.28318530718 \cdot u2\right)
\end{array}
(FPCore (cosTheta_i u1 u2) :precision binary32 (* (sqrt (/ u1 (- 1.0 u1))) (cos (cbrt (* (* 6.28318530718 u2) (* 39.47841760436263 (* u2 u2)))))))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf((u1 / (1.0f - u1))) * cosf(cbrtf(((6.28318530718f * u2) * (39.47841760436263f * (u2 * u2)))));
}
function code(cosTheta_i, u1, u2) return Float32(sqrt(Float32(u1 / Float32(Float32(1.0) - u1))) * cos(cbrt(Float32(Float32(Float32(6.28318530718) * u2) * Float32(Float32(39.47841760436263) * Float32(u2 * u2)))))) end
\begin{array}{l}
\\
\sqrt{\frac{u1}{1 - u1}} \cdot \cos \left(\sqrt[3]{\left(6.28318530718 \cdot u2\right) \cdot \left(39.47841760436263 \cdot \left(u2 \cdot u2\right)\right)}\right)
\end{array}
Initial program 99.2%
add-cbrt-cube_binary3299.2%
Applied rewrite-once99.2%
*-commutative99.2%
swap-sqr99.2%
metadata-eval99.2%
Simplified99.2%
Final simplification99.2%
(FPCore (cosTheta_i u1 u2)
:precision binary32
(if (<= (* 6.28318530718 u2) 0.11999999731779099)
(*
(+ 1.0 (* (* u2 u2) -19.739208802181317))
(sqrt (/ 1.0 (+ -1.0 (/ 1.0 u1)))))
(* (cos (* 6.28318530718 u2)) (sqrt u1))))
float code(float cosTheta_i, float u1, float u2) {
float tmp;
if ((6.28318530718f * u2) <= 0.11999999731779099f) {
tmp = (1.0f + ((u2 * u2) * -19.739208802181317f)) * sqrtf((1.0f / (-1.0f + (1.0f / u1))));
} else {
tmp = cosf((6.28318530718f * u2)) * sqrtf(u1);
}
return tmp;
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
real(4) :: tmp
if ((6.28318530718e0 * u2) <= 0.11999999731779099e0) then
tmp = (1.0e0 + ((u2 * u2) * (-19.739208802181317e0))) * sqrt((1.0e0 / ((-1.0e0) + (1.0e0 / u1))))
else
tmp = cos((6.28318530718e0 * u2)) * sqrt(u1)
end if
code = tmp
end function
function code(cosTheta_i, u1, u2) tmp = Float32(0.0) if (Float32(Float32(6.28318530718) * u2) <= Float32(0.11999999731779099)) tmp = Float32(Float32(Float32(1.0) + Float32(Float32(u2 * u2) * Float32(-19.739208802181317))) * sqrt(Float32(Float32(1.0) / Float32(Float32(-1.0) + Float32(Float32(1.0) / u1))))); else tmp = Float32(cos(Float32(Float32(6.28318530718) * u2)) * sqrt(u1)); end return tmp end
function tmp_2 = code(cosTheta_i, u1, u2) tmp = single(0.0); if ((single(6.28318530718) * u2) <= single(0.11999999731779099)) tmp = (single(1.0) + ((u2 * u2) * single(-19.739208802181317))) * sqrt((single(1.0) / (single(-1.0) + (single(1.0) / u1)))); else tmp = cos((single(6.28318530718) * u2)) * sqrt(u1); end tmp_2 = tmp; end
\begin{array}{l}
\\
\begin{array}{l}
\mathbf{if}\;6.28318530718 \cdot u2 \leq 0.11999999731779099:\\
\;\;\;\;\left(1 + \left(u2 \cdot u2\right) \cdot -19.739208802181317\right) \cdot \sqrt{\frac{1}{-1 + \frac{1}{u1}}}\\
\mathbf{else}:\\
\;\;\;\;\cos \left(6.28318530718 \cdot u2\right) \cdot \sqrt{u1}\\
\end{array}
\end{array}
if (*.f32 314159265359/50000000000 u2) < 0.119999997Initial program 99.3%
add-exp-log_binary3296.6%
Applied rewrite-once96.6%
rem-exp-log99.3%
clear-num99.1%
inv-pow99.1%
metadata-eval99.1%
sqrt-pow199.2%
div-sub99.2%
sub-neg99.2%
*-inverses99.2%
metadata-eval99.2%
metadata-eval99.2%
metadata-eval99.2%
Applied egg-rr99.2%
Taylor expanded in u2 around 0 98.7%
associate-*r*98.7%
distribute-rgt1-in98.6%
unpow298.6%
sub-neg98.6%
metadata-eval98.6%
+-commutative98.6%
Simplified98.6%
if 0.119999997 < (*.f32 314159265359/50000000000 u2) Initial program 98.6%
Taylor expanded in u1 around 0 79.2%
Final simplification94.9%
(FPCore (cosTheta_i u1 u2) :precision binary32 (* (sqrt (/ u1 (- 1.0 u1))) (cos (* 6.28318530718 u2))))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf((u1 / (1.0f - u1))) * cosf((6.28318530718f * u2));
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = sqrt((u1 / (1.0e0 - u1))) * cos((6.28318530718e0 * u2))
end function
function code(cosTheta_i, u1, u2) return Float32(sqrt(Float32(u1 / Float32(Float32(1.0) - u1))) * cos(Float32(Float32(6.28318530718) * u2))) end
function tmp = code(cosTheta_i, u1, u2) tmp = sqrt((u1 / (single(1.0) - u1))) * cos((single(6.28318530718) * u2)); end
\begin{array}{l}
\\
\sqrt{\frac{u1}{1 - u1}} \cdot \cos \left(6.28318530718 \cdot u2\right)
\end{array}
Initial program 99.2%
Final simplification99.2%
(FPCore (cosTheta_i u1 u2) :precision binary32 (* (+ 1.0 (* (* u2 u2) -19.739208802181317)) (sqrt (/ 1.0 (+ -1.0 (/ 1.0 u1))))))
float code(float cosTheta_i, float u1, float u2) {
return (1.0f + ((u2 * u2) * -19.739208802181317f)) * sqrtf((1.0f / (-1.0f + (1.0f / u1))));
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = (1.0e0 + ((u2 * u2) * (-19.739208802181317e0))) * sqrt((1.0e0 / ((-1.0e0) + (1.0e0 / u1))))
end function
function code(cosTheta_i, u1, u2) return Float32(Float32(Float32(1.0) + Float32(Float32(u2 * u2) * Float32(-19.739208802181317))) * sqrt(Float32(Float32(1.0) / Float32(Float32(-1.0) + Float32(Float32(1.0) / u1))))) end
function tmp = code(cosTheta_i, u1, u2) tmp = (single(1.0) + ((u2 * u2) * single(-19.739208802181317))) * sqrt((single(1.0) / (single(-1.0) + (single(1.0) / u1)))); end
\begin{array}{l}
\\
\left(1 + \left(u2 \cdot u2\right) \cdot -19.739208802181317\right) \cdot \sqrt{\frac{1}{-1 + \frac{1}{u1}}}
\end{array}
Initial program 99.2%
add-exp-log_binary3296.6%
Applied rewrite-once96.6%
rem-exp-log99.2%
clear-num99.0%
inv-pow99.0%
metadata-eval99.0%
sqrt-pow199.0%
div-sub99.1%
sub-neg99.1%
*-inverses99.1%
metadata-eval99.1%
metadata-eval99.1%
metadata-eval99.1%
Applied egg-rr99.1%
Taylor expanded in u2 around 0 88.2%
associate-*r*88.2%
distribute-rgt1-in88.2%
unpow288.2%
sub-neg88.2%
metadata-eval88.2%
+-commutative88.2%
Simplified88.2%
Final simplification88.2%
(FPCore (cosTheta_i u1 u2) :precision binary32 (sqrt (/ u1 (- 1.0 u1))))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf((u1 / (1.0f - u1)));
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = sqrt((u1 / (1.0e0 - u1)))
end function
function code(cosTheta_i, u1, u2) return sqrt(Float32(u1 / Float32(Float32(1.0) - u1))) end
function tmp = code(cosTheta_i, u1, u2) tmp = sqrt((u1 / (single(1.0) - u1))); end
\begin{array}{l}
\\
\sqrt{\frac{u1}{1 - u1}}
\end{array}
Initial program 99.2%
Taylor expanded in u2 around 0 78.6%
Final simplification78.6%
(FPCore (cosTheta_i u1 u2) :precision binary32 (sqrt u1))
float code(float cosTheta_i, float u1, float u2) {
return sqrtf(u1);
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = sqrt(u1)
end function
function code(cosTheta_i, u1, u2) return sqrt(u1) end
function tmp = code(cosTheta_i, u1, u2) tmp = sqrt(u1); end
\begin{array}{l}
\\
\sqrt{u1}
\end{array}
Initial program 99.2%
Taylor expanded in u2 around 0 78.6%
Taylor expanded in u1 around 0 62.5%
Final simplification62.5%
(FPCore (cosTheta_i u1 u2) :precision binary32 (+ u1 0.5))
float code(float cosTheta_i, float u1, float u2) {
return u1 + 0.5f;
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = u1 + 0.5e0
end function
function code(cosTheta_i, u1, u2) return Float32(u1 + Float32(0.5)) end
function tmp = code(cosTheta_i, u1, u2) tmp = u1 + single(0.5); end
\begin{array}{l}
\\
u1 + 0.5
\end{array}
Initial program 99.2%
Taylor expanded in u2 around 0 78.6%
Taylor expanded in u1 around 0 69.6%
+-commutative69.6%
unpow269.6%
fma-udef69.6%
Simplified69.6%
Taylor expanded in u1 around inf 20.4%
+-commutative20.4%
Simplified20.4%
Final simplification20.4%
(FPCore (cosTheta_i u1 u2) :precision binary32 1.0)
float code(float cosTheta_i, float u1, float u2) {
return 1.0f;
}
real(4) function code(costheta_i, u1, u2)
real(4), intent (in) :: costheta_i
real(4), intent (in) :: u1
real(4), intent (in) :: u2
code = 1.0e0
end function
function code(cosTheta_i, u1, u2) return Float32(1.0) end
function tmp = code(cosTheta_i, u1, u2) tmp = single(1.0); end
\begin{array}{l}
\\
1
\end{array}
Initial program 99.2%
Taylor expanded in u2 around 0 78.6%
Applied egg-rr60.9%
Taylor expanded in u1 around inf 19.4%
Final simplification19.4%
herbie shell --seed 2023297
(FPCore (cosTheta_i u1 u2)
:name "Trowbridge-Reitz Sample, near normal, slope_x"
:precision binary32
:pre (and (and (and (> cosTheta_i 0.9999) (<= cosTheta_i 1.0)) (and (<= 2.328306437e-10 u1) (<= u1 1.0))) (and (<= 2.328306437e-10 u2) (<= u2 1.0)))
(* (sqrt (/ u1 (- 1.0 u1))) (cos (* 6.28318530718 u2))))