| /external/eigen/unsupported/Eigen/src/IterativeSolvers/ |
| ConstrainedConjGrad.h | 58 Scalar rho, rho_1, alpha; local
67 rho = 1.0;
72 while (rho >= 1e-38)
77 alpha = rho / p.dot(q);
80 rho_1 = rho;
81 rho = r.dot(r);
82 p = (rho/rho_1) * p + r;
114 Scalar rho = 1.0, rho_1, lambda, gamma; local
157 rho_1 = rho;
158 rho = r.dot(z)
[all...] |
| /bionic/libm/upstream-freebsd/lib/msun/src/ |
| s_ctanh.c | 41 * rho = cosh(x)
55 * beta rho s + I t
78 double t, beta, s, rho, denom; local
133 rho = sqrt(1 + s * s); /* = cosh(x) */
135 return (CMPLX((beta * rho * s) / denom, t / denom));
|
| s_ctanhf.c | 43 float t, beta, s, rho, denom; local
73 rho = sqrtf(1 + s * s);
75 return (CMPLXF((beta * rho * s) / denom, t / denom));
|
| /external/iproute2/netem/ |
| stats.c | 24 double mu=0.0, sigma=0.0, sumsquare=0.0, sum=0.0, top=0.0, rho=0.0; local
51 rho = top/sigma2;
55 printf("rho = %12.6f\n", rho);
57 /*printf("correlation rho = %10.6f\n", top/((double)(n-1)*sigma*sigma));*/
|
| maketable.c | 51 arraystats(double *x, int limit, double *mu, double *sigma, double *rho)
70 *rho = top/sigma2;
202 double mu, sigma, rho; local
221 arraystats(x, limit, &mu, &sigma, &rho);
223 fprintf(stderr, "%d values, mu %10.4f, sigma %10.4f, rho %10.4f\n",
224 limit, mu, sigma, rho);
|
| /external/apache-commons-math/src/main/java/org/apache/commons/math/optimization/direct/ |
| NelderMead.java | 36 private final double rho; field in class:NelderMead
48 * <p>The default coefficients are 1.0 for rho, 2.0 for khi and 0.5
52 this.rho = 1.0;
59 * @param rho reflection coefficient
64 public NelderMead(final double rho, final double khi,
66 this.rho = rho;
105 xR[j] = centroid[j] + rho * (centroid[j] - xWorst[j]);
|
| /external/libcxx/test/std/numerics/complex.number/complex.value.ops/ |
| polar.pass.cpp | 14 // polar(const T& rho, const T& theta = T()); // changed from '0' by LWG#2870
23 test(const T& rho, std::complex<T> x)
25 assert(std::polar(rho) == x);
30 test(const T& rho, const T& theta, std::complex<T> x)
32 assert(std::polar(rho, theta) == x);
|
| /prebuilts/ndk/r16/sources/cxx-stl/llvm-libc++/test/std/numerics/complex.number/complex.value.ops/ |
| polar.pass.cpp | 14 // polar(const T& rho, const T& theta = 0);
23 test(const T& rho, std::complex<T> x)
25 assert(std::polar(rho) == x);
30 test(const T& rho, const T& theta, std::complex<T> x)
32 assert(std::polar(rho, theta) == x);
|
| /external/eigen/Eigen/src/IterativeLinearSolvers/ |
| BiCGSTAB.h | 52 Scalar rho = 1; local
69 Scalar rho_old = rho;
71 rho = r0.dot(r);
72 if (abs(rho) < eps2*r0_sqnorm)
78 rho = r0_sqnorm = r.squaredNorm();
82 Scalar beta = (rho/rho_old) * (alpha / w);
89 alpha = rho / r0.dot(v);
|
| /external/eigen/bench/ |
| eig33.cpp | 76 Scalar rho = std::sqrt(-a_over_3); local
80 roots(2) = c2_over_3 + Scalar(2)*rho*cos_theta;
81 roots(0) = c2_over_3 - rho*(cos_theta + s_sqrt3*sin_theta);
82 roots(1) = c2_over_3 - rho*(cos_theta - s_sqrt3*sin_theta);
|
| /external/tensorflow/tensorflow/python/training/ |
| adadelta_test.py | 58 rho = 0.95
60 adadelta_opt = adadelta.AdadeltaOptimizer(lr, rho, epsilon)
105 accum = accum * rho + (grad**2) * (1 - rho)
108 accum_update = (accum_update * rho + (update[step]**2) *
109 (1.0 - rho))
|
| adadelta.py | 36 def __init__(self, learning_rate=0.001, rho=0.95, epsilon=1e-8,
43 rho: A `Tensor` or a floating point value. The decay rate.
52 self._rho = rho
67 self._rho_t = ops.convert_to_tensor(self._rho, name="rho")
|
| /external/opencv/cv/src/ |
| cvhough.cpp | 64 float rho; member in struct:CvLinePolar
78 rho and theta are discretization steps (in pixels and radians correspondingly).
81 array of (rho, theta) pairs. linesMax is the buffer size (number of pairs).
85 icvHoughLinesStandard( const CvMat* img, float rho, float theta,
104 float irho = 1 / rho;
115 numrho = cvRound(((width + height) * 2 + 1) / rho);
165 line.rho = (r - (numrho - 1)*0.5f) * rho;
192 float rho, float theta, int threshold,
211 int rn, tn; /* number of rho and theta discrete values *
[all...] |
| /external/mesa3d/src/gallium/auxiliary/gallivm/ |
| lp_bld_sample.c | 203 * Generate code to compute coordinate gradient (rho).
206 * The resulting rho has bld->levelf format (per quad or per element).
232 LLVMValueRef rho; local
245 * rho calcs are always per quad except for explicit derivs (excluding
264 rho = lp_build_pack_aos_scalars(bld->gallivm, coord_bld->type,
268 rho = lp_build_swizzle_scalar_aos(coord_bld, cube_rho, 0, 4);
273 /* skipping sqrt hence returning rho squared */
275 rho = lp_build_mul(rho_bld, cubesize, rho);
311 rho = lp_build_max(coord_bld, rho_xvec, rho_yvec)
758 LLVMValueRef rho; local
[all...] |
| /external/ImageMagick/MagickCore/ |
| morphology.c | 275 args.rho = args.sigma; /* then width = height */
276 if ( args.rho < 1.0 ) /* if width too small */
277 args.rho = 1.0; /* then width = 1 */
279 args.sigma = args.rho; /* then height = width */
280 kernel->width = (size_t)args.rho;
421 flags, args.rho, args.sigma, args.xi, args.psi );
429 args.rho = 1.0; /* Default scale = 1.0, zero is valid */
446 args.rho = args.sigma; /* then width = height */
447 if ( args.rho < 1.0 ) /* if width too small */
448 args.rho = 3; /* then width = 3 *
[all...] |
| geometry.c | 818 % rho, xi, and psi values. It also returns flags that indicates which
819 % of the four values (sigma, rho, xi, psi) were located in the string, and
874 c=sscanf(geometry,"%lf%*[ ,]%lf%*[ ,]%lf%*[ ,]%lf",&coordinate.rho,
881 geometry_info->rho=coordinate.rho;
981 Parse rho, sigma, xi, psi, and optionally chi.
995 Parse rho.
1005 geometry_info->rho=value;
[all...] |
| geometry.h | 105 rho, member in struct:_GeometryInfo
|
| /external/bouncycastle/bcprov/src/main/java/org/bouncycastle/math/ec/ |
| WTauNafMultiplier.java | 37 ZTauElement rho = Tnaf.partModReduction(k, m, a, s, mu, (byte)10); local
39 return multiplyWTnaf(p, rho, curve.getPreCompInfo(p, PRECOMP_NAME), a, mu);
|
| /external/webrtc/webrtc/modules/audio_processing/intelligibility/ |
| intelligibility_enhancer.h | 48 rho(0.02f) {}
57 float rho; member in struct:webrtc::IntelligibilityEnhancer::Config
|
| /external/tensorflow/tensorflow/core/kernels/ |
| training_ops_gpu.cu.cc | 60 typename TTypes<T>::ConstScalar rho,
67 accum.device(d) = accum * rho.reshape(single).broadcast(bcast) +
69 rho.reshape(single).broadcast(bcast));
75 accum_update * rho.reshape(single).broadcast(bcast) +
77 (grad.constant(T(1)) - rho.reshape(single).broadcast(bcast));
150 typename TTypes<T>::ConstScalar rho,
159 ms + (rho.constant(one) - rho).reshape(single).broadcast(bcast) *
175 typename TTypes<T>::ConstScalar rho,
184 (rho.constant(one) - rho).reshape(single).broadcast(bcast)
[all...] |
| training_ops.h | 43 typename TTypes<T>::ConstScalar rho,
147 typename TTypes<T>::ConstScalar rho,
159 typename TTypes<T>::ConstScalar rho,
|
| /external/ImageMagick/MagickWand/ |
| operation.c | 88 flags, info.rho, info.sigma, info.xi, info.psi )
[all...] |
| mogrify.c | 782 mogrify_image=AdaptiveBlurImage(*image,geometry_info.rho,
806 mogrify_image=AdaptiveSharpenImage(*image,geometry_info.rho,
847 geometry_info.sigma=geometry_info.rho;
858 fmod(geometry_info.rho,360.0)));
860 fmod(geometry_info.rho,360.0)));
930 geometry_info.rho=1.5;
933 mogrify_image=BlueShiftImage(*image,geometry_info.rho,exception);
[all...] |
| /external/tensorflow/tensorflow/python/keras/_impl/keras/ |
| optimizers.py | 232 rho: float >= 0.
238 def __init__(self, lr=0.001, rho=0.9, epsilon=None, decay=0., **kwargs):
242 self.rho = K.variable(rho, name='rho')
264 new_a = self.rho * a + (1. - self.rho) * K.square(g)
278 'rho': float(K.get_value(self.rho)),
356 rho: float >= 0
[all...] |
| /external/tensorflow/tensorflow/contrib/opt/python/training/ |
| elastic_average_optimizer.py | 126 rho=None,
139 rho: the amount of exploration we allow ine the model. The default
156 if rho is None:
159 self._rho = rho
177 Add rho*elastic_difference to loss to control the exploration
|
