
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* (* (- 1.0 ux) maxCos) ux))
(t_1 (sqrt (- 1.0 (* t_0 t_0))))
(t_2 (* (* uy 2.0) PI)))
(+ (+ (* (* (cos t_2) t_1) xi) (* (* (sin t_2) t_1) yi)) (* t_0 zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((1.0f - ux) * maxCos) * ux;
float t_1 = sqrtf((1.0f - (t_0 * t_0)));
float t_2 = (uy * 2.0f) * ((float) M_PI);
return (((cosf(t_2) * t_1) * xi) + ((sinf(t_2) * t_1) * yi)) + (t_0 * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) t_1 = sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) t_2 = Float32(Float32(uy * Float32(2.0)) * Float32(pi)) return Float32(Float32(Float32(Float32(cos(t_2) * t_1) * xi) + Float32(Float32(sin(t_2) * t_1) * yi)) + Float32(t_0 * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = ((single(1.0) - ux) * maxCos) * ux; t_1 = sqrt((single(1.0) - (t_0 * t_0))); t_2 = (uy * single(2.0)) * single(pi); tmp = (((cos(t_2) * t_1) * xi) + ((sin(t_2) * t_1) * yi)) + (t_0 * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\\
t_1 := \sqrt{1 - t\_0 \cdot t\_0}\\
t_2 := \left(uy \cdot 2\right) \cdot \pi\\
\left(\left(\cos t\_2 \cdot t\_1\right) \cdot xi + \left(\sin t\_2 \cdot t\_1\right) \cdot yi\right) + t\_0 \cdot zi
\end{array}
\end{array}
Herbie found 18 alternatives:
| Alternative | Accuracy | Speedup |
|---|
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* (* (- 1.0 ux) maxCos) ux))
(t_1 (sqrt (- 1.0 (* t_0 t_0))))
(t_2 (* (* uy 2.0) PI)))
(+ (+ (* (* (cos t_2) t_1) xi) (* (* (sin t_2) t_1) yi)) (* t_0 zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((1.0f - ux) * maxCos) * ux;
float t_1 = sqrtf((1.0f - (t_0 * t_0)));
float t_2 = (uy * 2.0f) * ((float) M_PI);
return (((cosf(t_2) * t_1) * xi) + ((sinf(t_2) * t_1) * yi)) + (t_0 * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) t_1 = sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) t_2 = Float32(Float32(uy * Float32(2.0)) * Float32(pi)) return Float32(Float32(Float32(Float32(cos(t_2) * t_1) * xi) + Float32(Float32(sin(t_2) * t_1) * yi)) + Float32(t_0 * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = ((single(1.0) - ux) * maxCos) * ux; t_1 = sqrt((single(1.0) - (t_0 * t_0))); t_2 = (uy * single(2.0)) * single(pi); tmp = (((cos(t_2) * t_1) * xi) + ((sin(t_2) * t_1) * yi)) + (t_0 * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\\
t_1 := \sqrt{1 - t\_0 \cdot t\_0}\\
t_2 := \left(uy \cdot 2\right) \cdot \pi\\
\left(\left(\cos t\_2 \cdot t\_1\right) \cdot xi + \left(\sin t\_2 \cdot t\_1\right) \cdot yi\right) + t\_0 \cdot zi
\end{array}
\end{array}
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* (- 1.0 ux) (* maxCos ux)))
(t_1 (sqrt (- 1.0 (* t_0 t_0))))
(t_2 (* (* uy 2.0) PI)))
(+ (+ (* (* (cos t_2) t_1) xi) (* (* (sin t_2) t_1) yi)) (* t_0 zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = (1.0f - ux) * (maxCos * ux);
float t_1 = sqrtf((1.0f - (t_0 * t_0)));
float t_2 = (uy * 2.0f) * ((float) M_PI);
return (((cosf(t_2) * t_1) * xi) + ((sinf(t_2) * t_1) * yi)) + (t_0 * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(Float32(1.0) - ux) * Float32(maxCos * ux)) t_1 = sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) t_2 = Float32(Float32(uy * Float32(2.0)) * Float32(pi)) return Float32(Float32(Float32(Float32(cos(t_2) * t_1) * xi) + Float32(Float32(sin(t_2) * t_1) * yi)) + Float32(t_0 * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = (single(1.0) - ux) * (maxCos * ux); t_1 = sqrt((single(1.0) - (t_0 * t_0))); t_2 = (uy * single(2.0)) * single(pi); tmp = (((cos(t_2) * t_1) * xi) + ((sin(t_2) * t_1) * yi)) + (t_0 * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \left(1 - ux\right) \cdot \left(maxCos \cdot ux\right)\\
t_1 := \sqrt{1 - t\_0 \cdot t\_0}\\
t_2 := \left(uy \cdot 2\right) \cdot \pi\\
\left(\left(\cos t\_2 \cdot t\_1\right) \cdot xi + \left(\sin t\_2 \cdot t\_1\right) \cdot yi\right) + t\_0 \cdot zi
\end{array}
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* (* (- 1.0 ux) maxCos) ux))
(t_1 (sqrt (- 1.0 (* t_0 t_0))))
(t_2 (* (* uy 2.0) PI)))
(+ (+ (* (* (cos t_2) t_1) xi) (* (* (sin t_2) t_1) yi)) (* t_0 zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((1.0f - ux) * maxCos) * ux;
float t_1 = sqrtf((1.0f - (t_0 * t_0)));
float t_2 = (uy * 2.0f) * ((float) M_PI);
return (((cosf(t_2) * t_1) * xi) + ((sinf(t_2) * t_1) * yi)) + (t_0 * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) t_1 = sqrt(Float32(Float32(1.0) - Float32(t_0 * t_0))) t_2 = Float32(Float32(uy * Float32(2.0)) * Float32(pi)) return Float32(Float32(Float32(Float32(cos(t_2) * t_1) * xi) + Float32(Float32(sin(t_2) * t_1) * yi)) + Float32(t_0 * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = ((single(1.0) - ux) * maxCos) * ux; t_1 = sqrt((single(1.0) - (t_0 * t_0))); t_2 = (uy * single(2.0)) * single(pi); tmp = (((cos(t_2) * t_1) * xi) + ((sin(t_2) * t_1) * yi)) + (t_0 * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\\
t_1 := \sqrt{1 - t\_0 \cdot t\_0}\\
t_2 := \left(uy \cdot 2\right) \cdot \pi\\
\left(\left(\cos t\_2 \cdot t\_1\right) \cdot xi + \left(\sin t\_2 \cdot t\_1\right) \cdot yi\right) + t\_0 \cdot zi
\end{array}
\end{array}
Initial program 98.9%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(+
(*
(*
(sqrt
(- 1.0 (* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(+ (cos t_0) (* yi (/ (sin t_0) xi))))
xi)
(* (* (* (- 1.0 ux) maxCos) ux) zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
return ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (cosf(t_0) + (yi * (sinf(t_0) / xi)))) * xi) + ((((1.0f - ux) * maxCos) * ux) * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) return Float32(Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * Float32(cos(t_0) + Float32(yi * Float32(sin(t_0) / xi)))) * xi) + Float32(Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = single(pi) * (uy + uy); tmp = ((sqrt((single(1.0) - ((((single(1.0) - ux) * (single(1.0) - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (cos(t_0) + (yi * (sin(t_0) / xi)))) * xi) + ((((single(1.0) - ux) * maxCos) * ux) * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\left(\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \left(\cos t\_0 + yi \cdot \frac{\sin t\_0}{xi}\right)\right) \cdot xi + \left(\left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\right) \cdot zi
\end{array}
\end{array}
Initial program 98.9%
Taylor expanded in xi around inf
Applied rewrites98.7%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
(sqrt
(- 1.0 (* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(fma (cos t_0) xi (* (sin t_0) yi)))
zi))
zi)))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
return fmaf((maxCos * ux), (1.0f - ux), ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * fmaf(cosf(t_0), xi, (sinf(t_0) * yi))) / zi)) * zi;
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) return Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * fma(cos(t_0), xi, Float32(sin(t_0) * yi))) / zi)) * zi) end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \mathsf{fma}\left(\cos t\_0, xi, \sin t\_0 \cdot yi\right)}{zi}\right) \cdot zi
\end{array}
\end{array}
Initial program 98.9%
Taylor expanded in zi around inf
Applied rewrites98.1%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* (* uy 2.0) PI)))
(+
(+ (* (* (cos t_0) 1.0) xi) (* (* (sin t_0) 1.0) yi))
(* (* (* (- 1.0 ux) maxCos) ux) zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = (uy * 2.0f) * ((float) M_PI);
return (((cosf(t_0) * 1.0f) * xi) + ((sinf(t_0) * 1.0f) * yi)) + ((((1.0f - ux) * maxCos) * ux) * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(uy * Float32(2.0)) * Float32(pi)) return Float32(Float32(Float32(Float32(cos(t_0) * Float32(1.0)) * xi) + Float32(Float32(sin(t_0) * Float32(1.0)) * yi)) + Float32(Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = (uy * single(2.0)) * single(pi); tmp = (((cos(t_0) * single(1.0)) * xi) + ((sin(t_0) * single(1.0)) * yi)) + ((((single(1.0) - ux) * maxCos) * ux) * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \left(uy \cdot 2\right) \cdot \pi\\
\left(\left(\cos t\_0 \cdot 1\right) \cdot xi + \left(\sin t\_0 \cdot 1\right) \cdot yi\right) + \left(\left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\right) \cdot zi
\end{array}
\end{array}
Initial program 98.9%
Taylor expanded in ux around 0
Applied rewrites98.8%
Taylor expanded in ux around 0
Applied rewrites98.7%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(+
(* (* 1.0 (+ (cos t_0) (* yi (/ (sin t_0) xi)))) xi)
(* (* (* (- 1.0 ux) maxCos) ux) zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
return ((1.0f * (cosf(t_0) + (yi * (sinf(t_0) / xi)))) * xi) + ((((1.0f - ux) * maxCos) * ux) * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) return Float32(Float32(Float32(Float32(1.0) * Float32(cos(t_0) + Float32(yi * Float32(sin(t_0) / xi)))) * xi) + Float32(Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) t_0 = single(pi) * (uy + uy); tmp = ((single(1.0) * (cos(t_0) + (yi * (sin(t_0) / xi)))) * xi) + ((((single(1.0) - ux) * maxCos) * ux) * zi); end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\left(1 \cdot \left(\cos t\_0 + yi \cdot \frac{\sin t\_0}{xi}\right)\right) \cdot xi + \left(\left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\right) \cdot zi
\end{array}
\end{array}
Initial program 98.9%
Taylor expanded in xi around inf
Applied rewrites98.7%
Taylor expanded in ux around 0
Applied rewrites98.5%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(if (<= uy 0.012000000104308128)
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
(sqrt
(-
1.0
(* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(fma
uy
(fma
-1.3333333333333333
(* (* uy uy) (* yi (* (* PI PI) PI)))
(* 2.0 (* yi PI)))
(* xi (+ 1.0 (* -2.0 (* (* uy uy) (* PI PI)))))))
zi))
zi)
(fma (cos t_0) xi (* (sin t_0) yi)))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
float tmp;
if (uy <= 0.012000000104308128f) {
tmp = fmaf((maxCos * ux), (1.0f - ux), ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * fmaf(uy, fmaf(-1.3333333333333333f, ((uy * uy) * (yi * ((((float) M_PI) * ((float) M_PI)) * ((float) M_PI)))), (2.0f * (yi * ((float) M_PI)))), (xi * (1.0f + (-2.0f * ((uy * uy) * (((float) M_PI) * ((float) M_PI)))))))) / zi)) * zi;
} else {
tmp = fmaf(cosf(t_0), xi, (sinf(t_0) * yi));
}
return tmp;
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) tmp = Float32(0.0) if (uy <= Float32(0.012000000104308128)) tmp = Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * fma(uy, fma(Float32(-1.3333333333333333), Float32(Float32(uy * uy) * Float32(yi * Float32(Float32(Float32(pi) * Float32(pi)) * Float32(pi)))), Float32(Float32(2.0) * Float32(yi * Float32(pi)))), Float32(xi * Float32(Float32(1.0) + Float32(Float32(-2.0) * Float32(Float32(uy * uy) * Float32(Float32(pi) * Float32(pi)))))))) / zi)) * zi); else tmp = fma(cos(t_0), xi, Float32(sin(t_0) * yi)); end return tmp end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\mathbf{if}\;uy \leq 0.012000000104308128:\\
\;\;\;\;\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \mathsf{fma}\left(uy, \mathsf{fma}\left(-1.3333333333333333, \left(uy \cdot uy\right) \cdot \left(yi \cdot \left(\left(\pi \cdot \pi\right) \cdot \pi\right)\right), 2 \cdot \left(yi \cdot \pi\right)\right), xi \cdot \left(1 + -2 \cdot \left(\left(uy \cdot uy\right) \cdot \left(\pi \cdot \pi\right)\right)\right)\right)}{zi}\right) \cdot zi\\
\mathbf{else}:\\
\;\;\;\;\mathsf{fma}\left(\cos t\_0, xi, \sin t\_0 \cdot yi\right)\\
\end{array}
\end{array}
if uy < 0.0120000001Initial program 99.2%
Taylor expanded in zi around inf
Applied rewrites98.4%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
lower-*.f32N/A
lift-PI.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
Applied rewrites98.4%
Taylor expanded in xi around 0
lower-fma.f32N/A
Applied rewrites98.3%
if 0.0120000001 < uy Initial program 97.7%
Taylor expanded in ux around 0
*-commutativeN/A
lower-fma.f32N/A
Applied rewrites90.5%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(if (<= uy 0.012000000104308128)
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
(sqrt
(-
1.0
(* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(+
xi
(*
uy
(fma
2.0
(* yi PI)
(*
uy
(fma
-2.0
(* xi (* PI PI))
(* -1.3333333333333333 (* uy (* yi (* (* PI PI) PI))))))))))
zi))
zi)
(fma (cos t_0) xi (* (sin t_0) yi)))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
float tmp;
if (uy <= 0.012000000104308128f) {
tmp = fmaf((maxCos * ux), (1.0f - ux), ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (xi + (uy * fmaf(2.0f, (yi * ((float) M_PI)), (uy * fmaf(-2.0f, (xi * (((float) M_PI) * ((float) M_PI))), (-1.3333333333333333f * (uy * (yi * ((((float) M_PI) * ((float) M_PI)) * ((float) M_PI))))))))))) / zi)) * zi;
} else {
tmp = fmaf(cosf(t_0), xi, (sinf(t_0) * yi));
}
return tmp;
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) tmp = Float32(0.0) if (uy <= Float32(0.012000000104308128)) tmp = Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * Float32(xi + Float32(uy * fma(Float32(2.0), Float32(yi * Float32(pi)), Float32(uy * fma(Float32(-2.0), Float32(xi * Float32(Float32(pi) * Float32(pi))), Float32(Float32(-1.3333333333333333) * Float32(uy * Float32(yi * Float32(Float32(Float32(pi) * Float32(pi)) * Float32(pi))))))))))) / zi)) * zi); else tmp = fma(cos(t_0), xi, Float32(sin(t_0) * yi)); end return tmp end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\mathbf{if}\;uy \leq 0.012000000104308128:\\
\;\;\;\;\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \left(xi + uy \cdot \mathsf{fma}\left(2, yi \cdot \pi, uy \cdot \mathsf{fma}\left(-2, xi \cdot \left(\pi \cdot \pi\right), -1.3333333333333333 \cdot \left(uy \cdot \left(yi \cdot \left(\left(\pi \cdot \pi\right) \cdot \pi\right)\right)\right)\right)\right)\right)}{zi}\right) \cdot zi\\
\mathbf{else}:\\
\;\;\;\;\mathsf{fma}\left(\cos t\_0, xi, \sin t\_0 \cdot yi\right)\\
\end{array}
\end{array}
if uy < 0.0120000001Initial program 99.2%
Taylor expanded in zi around inf
Applied rewrites98.4%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
lower-*.f32N/A
lift-PI.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
Applied rewrites98.4%
if 0.0120000001 < uy Initial program 97.7%
Taylor expanded in ux around 0
*-commutativeN/A
lower-fma.f32N/A
Applied rewrites90.5%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(let* ((t_0 (* PI (+ uy uy))))
(if (<= uy 0.012000000104308128)
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
1.0
(+
xi
(*
uy
(fma
2.0
(* yi PI)
(*
uy
(fma
-2.0
(* xi (* PI PI))
(* -1.3333333333333333 (* uy (* yi (* (* PI PI) PI))))))))))
zi))
zi)
(fma (cos t_0) xi (* (sin t_0) yi)))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
float t_0 = ((float) M_PI) * (uy + uy);
float tmp;
if (uy <= 0.012000000104308128f) {
tmp = fmaf((maxCos * ux), (1.0f - ux), ((1.0f * (xi + (uy * fmaf(2.0f, (yi * ((float) M_PI)), (uy * fmaf(-2.0f, (xi * (((float) M_PI) * ((float) M_PI))), (-1.3333333333333333f * (uy * (yi * ((((float) M_PI) * ((float) M_PI)) * ((float) M_PI))))))))))) / zi)) * zi;
} else {
tmp = fmaf(cosf(t_0), xi, (sinf(t_0) * yi));
}
return tmp;
}
function code(xi, yi, zi, ux, uy, maxCos) t_0 = Float32(Float32(pi) * Float32(uy + uy)) tmp = Float32(0.0) if (uy <= Float32(0.012000000104308128)) tmp = Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(Float32(1.0) * Float32(xi + Float32(uy * fma(Float32(2.0), Float32(yi * Float32(pi)), Float32(uy * fma(Float32(-2.0), Float32(xi * Float32(Float32(pi) * Float32(pi))), Float32(Float32(-1.3333333333333333) * Float32(uy * Float32(yi * Float32(Float32(Float32(pi) * Float32(pi)) * Float32(pi))))))))))) / zi)) * zi); else tmp = fma(cos(t_0), xi, Float32(sin(t_0) * yi)); end return tmp end
\begin{array}{l}
\\
\begin{array}{l}
t_0 := \pi \cdot \left(uy + uy\right)\\
\mathbf{if}\;uy \leq 0.012000000104308128:\\
\;\;\;\;\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{1 \cdot \left(xi + uy \cdot \mathsf{fma}\left(2, yi \cdot \pi, uy \cdot \mathsf{fma}\left(-2, xi \cdot \left(\pi \cdot \pi\right), -1.3333333333333333 \cdot \left(uy \cdot \left(yi \cdot \left(\left(\pi \cdot \pi\right) \cdot \pi\right)\right)\right)\right)\right)\right)}{zi}\right) \cdot zi\\
\mathbf{else}:\\
\;\;\;\;\mathsf{fma}\left(\cos t\_0, xi, \sin t\_0 \cdot yi\right)\\
\end{array}
\end{array}
if uy < 0.0120000001Initial program 99.2%
Taylor expanded in zi around inf
Applied rewrites98.4%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
lower-*.f32N/A
lift-PI.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
Applied rewrites98.4%
Taylor expanded in ux around 0
Applied rewrites98.2%
if 0.0120000001 < uy Initial program 97.7%
Taylor expanded in ux around 0
*-commutativeN/A
lower-fma.f32N/A
Applied rewrites90.5%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
1.0
(+
xi
(*
uy
(fma
2.0
(* yi PI)
(*
uy
(fma
-2.0
(* xi (* PI PI))
(* -1.3333333333333333 (* uy (* yi (* (* PI PI) PI))))))))))
zi))
zi))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return fmaf((maxCos * ux), (1.0f - ux), ((1.0f * (xi + (uy * fmaf(2.0f, (yi * ((float) M_PI)), (uy * fmaf(-2.0f, (xi * (((float) M_PI) * ((float) M_PI))), (-1.3333333333333333f * (uy * (yi * ((((float) M_PI) * ((float) M_PI)) * ((float) M_PI))))))))))) / zi)) * zi;
}
function code(xi, yi, zi, ux, uy, maxCos) return Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(Float32(1.0) * Float32(xi + Float32(uy * fma(Float32(2.0), Float32(yi * Float32(pi)), Float32(uy * fma(Float32(-2.0), Float32(xi * Float32(Float32(pi) * Float32(pi))), Float32(Float32(-1.3333333333333333) * Float32(uy * Float32(yi * Float32(Float32(Float32(pi) * Float32(pi)) * Float32(pi))))))))))) / zi)) * zi) end
\begin{array}{l}
\\
\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{1 \cdot \left(xi + uy \cdot \mathsf{fma}\left(2, yi \cdot \pi, uy \cdot \mathsf{fma}\left(-2, xi \cdot \left(\pi \cdot \pi\right), -1.3333333333333333 \cdot \left(uy \cdot \left(yi \cdot \left(\left(\pi \cdot \pi\right) \cdot \pi\right)\right)\right)\right)\right)\right)}{zi}\right) \cdot zi
\end{array}
Initial program 98.9%
Taylor expanded in zi around inf
Applied rewrites98.1%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
lower-*.f32N/A
lift-PI.f32N/A
lower-*.f32N/A
lower-fma.f32N/A
Applied rewrites89.0%
Taylor expanded in ux around 0
Applied rewrites88.9%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(+
(*
(*
(sqrt
(- 1.0 (* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(+ 1.0 (* 2.0 (/ (* uy (* yi PI)) xi))))
xi)
(* (* (* (- 1.0 ux) maxCos) ux) zi)))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (1.0f + (2.0f * ((uy * (yi * ((float) M_PI))) / xi)))) * xi) + ((((1.0f - ux) * maxCos) * ux) * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) return Float32(Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * Float32(Float32(1.0) + Float32(Float32(2.0) * Float32(Float32(uy * Float32(yi * Float32(pi))) / xi)))) * xi) + Float32(Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = ((sqrt((single(1.0) - ((((single(1.0) - ux) * (single(1.0) - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (single(1.0) + (single(2.0) * ((uy * (yi * single(pi))) / xi)))) * xi) + ((((single(1.0) - ux) * maxCos) * ux) * zi); end
\begin{array}{l}
\\
\left(\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \left(1 + 2 \cdot \frac{uy \cdot \left(yi \cdot \pi\right)}{xi}\right)\right) \cdot xi + \left(\left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\right) \cdot zi
\end{array}
Initial program 98.9%
Taylor expanded in xi around inf
Applied rewrites98.7%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-/.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift-PI.f3281.7
Applied rewrites81.7%
(FPCore (xi yi zi ux uy maxCos)
:precision binary32
(*
(fma
(* maxCos ux)
(- 1.0 ux)
(/
(*
(sqrt
(- 1.0 (* (* (* (- 1.0 ux) (- 1.0 ux)) (* ux ux)) (* maxCos maxCos))))
(+ xi (* 2.0 (* uy (* yi PI)))))
zi))
zi))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return fmaf((maxCos * ux), (1.0f - ux), ((sqrtf((1.0f - ((((1.0f - ux) * (1.0f - ux)) * (ux * ux)) * (maxCos * maxCos)))) * (xi + (2.0f * (uy * (yi * ((float) M_PI)))))) / zi)) * zi;
}
function code(xi, yi, zi, ux, uy, maxCos) return Float32(fma(Float32(maxCos * ux), Float32(Float32(1.0) - ux), Float32(Float32(sqrt(Float32(Float32(1.0) - Float32(Float32(Float32(Float32(Float32(1.0) - ux) * Float32(Float32(1.0) - ux)) * Float32(ux * ux)) * Float32(maxCos * maxCos)))) * Float32(xi + Float32(Float32(2.0) * Float32(uy * Float32(yi * Float32(pi)))))) / zi)) * zi) end
\begin{array}{l}
\\
\mathsf{fma}\left(maxCos \cdot ux, 1 - ux, \frac{\sqrt{1 - \left(\left(\left(1 - ux\right) \cdot \left(1 - ux\right)\right) \cdot \left(ux \cdot ux\right)\right) \cdot \left(maxCos \cdot maxCos\right)} \cdot \left(xi + 2 \cdot \left(uy \cdot \left(yi \cdot \pi\right)\right)\right)}{zi}\right) \cdot zi
\end{array}
Initial program 98.9%
Taylor expanded in zi around inf
Applied rewrites98.1%
Taylor expanded in uy around 0
lower-+.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift-PI.f3281.2
Applied rewrites81.2%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 (+ (+ xi (* 2.0 (* uy (* yi PI)))) (* (* (* (- 1.0 ux) maxCos) ux) zi)))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return (xi + (2.0f * (uy * (yi * ((float) M_PI))))) + ((((1.0f - ux) * maxCos) * ux) * zi);
}
function code(xi, yi, zi, ux, uy, maxCos) return Float32(Float32(xi + Float32(Float32(2.0) * Float32(uy * Float32(yi * Float32(pi))))) + Float32(Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * ux) * zi)) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = (xi + (single(2.0) * (uy * (yi * single(pi))))) + ((((single(1.0) - ux) * maxCos) * ux) * zi); end
\begin{array}{l}
\\
\left(xi + 2 \cdot \left(uy \cdot \left(yi \cdot \pi\right)\right)\right) + \left(\left(\left(1 - ux\right) \cdot maxCos\right) \cdot ux\right) \cdot zi
\end{array}
Initial program 98.9%
Taylor expanded in uy around 0
associate-*r*N/A
lower-fma.f32N/A
Applied rewrites81.9%
Taylor expanded in ux around 0
lower-+.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift-PI.f3281.7
Applied rewrites81.7%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 (+ xi (* (* (- 1.0 ux) maxCos) (* zi ux))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return xi + (((1.0f - ux) * maxCos) * (zi * ux));
}
module fmin_fmax_functions
implicit none
private
public fmax
public fmin
interface fmax
module procedure fmax88
module procedure fmax44
module procedure fmax84
module procedure fmax48
end interface
interface fmin
module procedure fmin88
module procedure fmin44
module procedure fmin84
module procedure fmin48
end interface
contains
real(8) function fmax88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(4) function fmax44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(8) function fmax84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmax48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
end function
real(8) function fmin88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(4) function fmin44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(8) function fmin84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmin48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
end function
end module
real(4) function code(xi, yi, zi, ux, uy, maxcos)
use fmin_fmax_functions
real(4), intent (in) :: xi
real(4), intent (in) :: yi
real(4), intent (in) :: zi
real(4), intent (in) :: ux
real(4), intent (in) :: uy
real(4), intent (in) :: maxcos
code = xi + (((1.0e0 - ux) * maxcos) * (zi * ux))
end function
function code(xi, yi, zi, ux, uy, maxCos) return Float32(xi + Float32(Float32(Float32(Float32(1.0) - ux) * maxCos) * Float32(zi * ux))) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = xi + (((single(1.0) - ux) * maxCos) * (zi * ux)); end
\begin{array}{l}
\\
xi + \left(\left(1 - ux\right) \cdot maxCos\right) \cdot \left(zi \cdot ux\right)
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lift-*.f32N/A
*-commutativeN/A
lower-*.f3298.9
Applied rewrites98.9%
Taylor expanded in uy around 0
pow2N/A
pow2N/A
pow2N/A
lower-sqrt.f32N/A
lift-*.f32N/A
lift--.f32N/A
lift--.f32N/A
lift-*.f32N/A
Applied rewrites51.6%
Taylor expanded in ux around 0
Applied rewrites51.5%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 (+ xi (* maxCos (* ux (* zi (- 1.0 ux))))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return xi + (maxCos * (ux * (zi * (1.0f - ux))));
}
module fmin_fmax_functions
implicit none
private
public fmax
public fmin
interface fmax
module procedure fmax88
module procedure fmax44
module procedure fmax84
module procedure fmax48
end interface
interface fmin
module procedure fmin88
module procedure fmin44
module procedure fmin84
module procedure fmin48
end interface
contains
real(8) function fmax88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(4) function fmax44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(8) function fmax84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmax48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
end function
real(8) function fmin88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(4) function fmin44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(8) function fmin84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmin48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
end function
end module
real(4) function code(xi, yi, zi, ux, uy, maxcos)
use fmin_fmax_functions
real(4), intent (in) :: xi
real(4), intent (in) :: yi
real(4), intent (in) :: zi
real(4), intent (in) :: ux
real(4), intent (in) :: uy
real(4), intent (in) :: maxcos
code = xi + (maxcos * (ux * (zi * (1.0e0 - ux))))
end function
function code(xi, yi, zi, ux, uy, maxCos) return Float32(xi + Float32(maxCos * Float32(ux * Float32(zi * Float32(Float32(1.0) - ux))))) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = xi + (maxCos * (ux * (zi * (single(1.0) - ux)))); end
\begin{array}{l}
\\
xi + maxCos \cdot \left(ux \cdot \left(zi \cdot \left(1 - ux\right)\right)\right)
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
Taylor expanded in uy around 0
lower-fma.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f32N/A
lower-sqrt.f32N/A
lower--.f32N/A
Applied rewrites51.6%
Taylor expanded in maxCos around 0
lower-+.f32N/A
lower-*.f32N/A
lift--.f32N/A
lift-*.f32N/A
lift-*.f3251.5
Applied rewrites51.5%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 (fma maxCos (* ux (* zi (- 1.0 ux))) xi))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return fmaf(maxCos, (ux * (zi * (1.0f - ux))), xi);
}
function code(xi, yi, zi, ux, uy, maxCos) return fma(maxCos, Float32(ux * Float32(zi * Float32(Float32(1.0) - ux))), xi) end
\begin{array}{l}
\\
\mathsf{fma}\left(maxCos, ux \cdot \left(zi \cdot \left(1 - ux\right)\right), xi\right)
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
Taylor expanded in uy around 0
lower-fma.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f32N/A
lower-sqrt.f32N/A
lower--.f32N/A
Applied rewrites51.6%
Taylor expanded in ux around 0
Applied rewrites51.5%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 (+ xi (* maxCos (* ux zi))))
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return xi + (maxCos * (ux * zi));
}
module fmin_fmax_functions
implicit none
private
public fmax
public fmin
interface fmax
module procedure fmax88
module procedure fmax44
module procedure fmax84
module procedure fmax48
end interface
interface fmin
module procedure fmin88
module procedure fmin44
module procedure fmin84
module procedure fmin48
end interface
contains
real(8) function fmax88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(4) function fmax44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(8) function fmax84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmax48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
end function
real(8) function fmin88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(4) function fmin44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(8) function fmin84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmin48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
end function
end module
real(4) function code(xi, yi, zi, ux, uy, maxcos)
use fmin_fmax_functions
real(4), intent (in) :: xi
real(4), intent (in) :: yi
real(4), intent (in) :: zi
real(4), intent (in) :: ux
real(4), intent (in) :: uy
real(4), intent (in) :: maxcos
code = xi + (maxcos * (ux * zi))
end function
function code(xi, yi, zi, ux, uy, maxCos) return Float32(xi + Float32(maxCos * Float32(ux * zi))) end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = xi + (maxCos * (ux * zi)); end
\begin{array}{l}
\\
xi + maxCos \cdot \left(ux \cdot zi\right)
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
Taylor expanded in uy around 0
lower-fma.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f32N/A
lower-sqrt.f32N/A
lower--.f32N/A
Applied rewrites51.6%
Taylor expanded in ux around 0
lower-+.f32N/A
lower-*.f32N/A
lower-*.f3249.7
Applied rewrites49.7%
(FPCore (xi yi zi ux uy maxCos) :precision binary32 xi)
float code(float xi, float yi, float zi, float ux, float uy, float maxCos) {
return xi;
}
module fmin_fmax_functions
implicit none
private
public fmax
public fmin
interface fmax
module procedure fmax88
module procedure fmax44
module procedure fmax84
module procedure fmax48
end interface
interface fmin
module procedure fmin88
module procedure fmin44
module procedure fmin84
module procedure fmin48
end interface
contains
real(8) function fmax88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(4) function fmax44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, max(x, y), y /= y), x /= x)
end function
real(8) function fmax84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, max(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmax48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), max(dble(x), y), y /= y), x /= x)
end function
real(8) function fmin88(x, y) result (res)
real(8), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(4) function fmin44(x, y) result (res)
real(4), intent (in) :: x
real(4), intent (in) :: y
res = merge(y, merge(x, min(x, y), y /= y), x /= x)
end function
real(8) function fmin84(x, y) result(res)
real(8), intent (in) :: x
real(4), intent (in) :: y
res = merge(dble(y), merge(x, min(x, dble(y)), y /= y), x /= x)
end function
real(8) function fmin48(x, y) result(res)
real(4), intent (in) :: x
real(8), intent (in) :: y
res = merge(y, merge(dble(x), min(dble(x), y), y /= y), x /= x)
end function
end module
real(4) function code(xi, yi, zi, ux, uy, maxcos)
use fmin_fmax_functions
real(4), intent (in) :: xi
real(4), intent (in) :: yi
real(4), intent (in) :: zi
real(4), intent (in) :: ux
real(4), intent (in) :: uy
real(4), intent (in) :: maxcos
code = xi
end function
function code(xi, yi, zi, ux, uy, maxCos) return xi end
function tmp = code(xi, yi, zi, ux, uy, maxCos) tmp = xi; end
\begin{array}{l}
\\
xi
\end{array}
Initial program 98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
lift-*.f32N/A
lift-*.f32N/A
lift--.f32N/A
associate-*l*N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f3298.9
Applied rewrites98.9%
Taylor expanded in uy around 0
lower-fma.f32N/A
lower-*.f32N/A
lower-*.f32N/A
lift--.f32N/A
lower-*.f32N/A
lower-sqrt.f32N/A
lower--.f32N/A
Applied rewrites51.6%
Taylor expanded in ux around 0
Applied rewrites45.8%
herbie shell --seed 2025122
(FPCore (xi yi zi ux uy maxCos)
:name "UniformSampleCone 2"
:precision binary32
:pre (and (and (and (and (and (and (<= -10000.0 xi) (<= xi 10000.0)) (and (<= -10000.0 yi) (<= yi 10000.0))) (and (<= -10000.0 zi) (<= zi 10000.0))) (and (<= 2.328306437e-10 ux) (<= ux 1.0))) (and (<= 2.328306437e-10 uy) (<= uy 1.0))) (and (<= 0.0 maxCos) (<= maxCos 1.0)))
(+ (+ (* (* (cos (* (* uy 2.0) PI)) (sqrt (- 1.0 (* (* (* (- 1.0 ux) maxCos) ux) (* (* (- 1.0 ux) maxCos) ux))))) xi) (* (* (sin (* (* uy 2.0) PI)) (sqrt (- 1.0 (* (* (* (- 1.0 ux) maxCos) ux) (* (* (- 1.0 ux) maxCos) ux))))) yi)) (* (* (* (- 1.0 ux) maxCos) ux) zi)))