/dports/math/py-iohexperimenter/IOHexperimenter-0.2.9.2/src/ |
H A D | coco_transformation_objs.hpp | 18 double log_y; in transform_obj_oscillate_evaluate() local 19 log_y = log(fabs(y[i])) / factor; in transform_obj_oscillate_evaluate() 21 y[i] = pow(exp(log_y + 0.49 * (sin(log_y) + sin(0.79 * log_y))), factor); in transform_obj_oscillate_evaluate() 23 y[i] = -pow(exp(log_y + 0.49 * (sin(0.55 * log_y) + sin(0.31 * log_y))), factor); in transform_obj_oscillate_evaluate()
|
/dports/science/py-chainer/chainer-7.8.0/chainer/functions/loss/ |
H A D | softmax_cross_entropy.py | 148 log_y = log_softmax._log_softmax(x) 150 self.y = numpy.exp(log_y) 158 log_yd = numpy.rollaxis(log_y, 1) 200 log_y = log_softmax._log_softmax(x) 202 self.y = cupy.exp(log_y) 211 log_y = cupy.rollaxis(log_y, 1, log_y.ndim) 231 )(t, log_y.reduced_view(), log_y.shape[-1], 233 ret = ret.astype(log_y.dtype, copy=False) 245 )(t, log_y.reduced_view(), log_y.shape[-1], self.ignore_label) 249 def _soft_target_loss(self, xp, x, t, log_y): argument [all …]
|
/dports/math/stanmath/math-4.2.0/stan/math/prim/prob/ |
H A D | dirichlet_rng.hpp | 52 VectorXd log_y(alpha.size()); in dirichlet_rng() local 57 log_y(i) = log(gamma_rng()) + log_u / alpha(i); in dirichlet_rng() 59 double log_sum_y = log_sum_exp(log_y); in dirichlet_rng() 62 theta(i) = exp(log_y(i) - log_sum_y); in dirichlet_rng()
|
H A D | chi_square_lpdf.hpp | 72 const auto& log_y = to_ref_if<!is_constant_all<T_dof>::value>(log(y_val)); in chi_square_lpdf() local 79 logp += sum((half_nu - 1.0) * log_y); in chi_square_lpdf() 92 (log_y - digamma(half_nu)) * 0.5 - HALF_LOG_TWO); in chi_square_lpdf() 95 = sum(log_y - digamma(half_nu)) * 0.5 - HALF_LOG_TWO * N; in chi_square_lpdf()
|
H A D | pareto_lpdf.hpp | 61 const auto& log_y = to_ref_if<!is_constant_all<T_shape>::value>(log(y_val)); in pareto_lpdf() local 69 logp -= sum(alpha_val * log_y + log_y) * N / max_size(alpha, y); in pareto_lpdf() 88 ops_partials.edge3_.partials_ = inv(alpha_val) + log_y_min - log_y; in pareto_lpdf()
|
H A D | beta_proportion_lpdf.hpp | 83 const auto& log_y in beta_proportion_lpdf() local 98 logp += sum((mukappa - 1) * log_y + (kappa_val - mukappa - 1) * log1m_y); in beta_proportion_lpdf() 116 * (digamma_kappa_mukappa - digamma_mukappa + log_y - log1m_y); in beta_proportion_lpdf() 120 = digamma(kappa_val) + mu_val * (log_y - digamma_mukappa) in beta_proportion_lpdf()
|
H A D | inv_chi_square_lpdf.hpp | 79 const auto& log_y = to_ref_if<!is_constant_all<T_dof>::value>(log(y_val)); in inv_chi_square_lpdf() local 83 T_partials_return logp = -sum((half_nu + 1.0) * log_y); in inv_chi_square_lpdf() 97 = -HALF_LOG_TWO - (digamma(half_nu) + log_y) * 0.5; in inv_chi_square_lpdf()
|
H A D | frechet_lpdf.hpp | 64 const auto& log_y in frechet_lpdf() local 78 logp -= sum((alpha_val + 1.0) * log_y) * N / max_size(y, alpha); in frechet_lpdf() 86 = inv(alpha_val) + (1 - sigma_div_y_pow_alpha) * (log_sigma - log_y); in frechet_lpdf()
|
/dports/math/py-pystan/pystan-2.19.0.0/pystan/stan/lib/stan_math/stan/math/prim/mat/prob/ |
H A D | dirichlet_rng.hpp | 57 VectorXd log_y(alpha.size()); in dirichlet_rng() local 62 log_y(i) = log(gamma_rng()) + log_u / alpha(i); in dirichlet_rng() 64 double log_sum_y = log_sum_exp(log_y); in dirichlet_rng() 67 theta(i) = exp(log_y(i) - log_sum_y); in dirichlet_rng()
|
/dports/math/py-pystan/pystan-2.19.0.0/pystan/stan/lib/stan_math/stan/math/prim/scal/prob/ |
H A D | pareto_lpdf.hpp | 65 log_y(length(y)); in pareto_lpdf() local 68 log_y[n] = log(value_of(y_vec[n])); in pareto_lpdf() 102 logp -= alpha_dbl * log_y[n] + log_y[n]; in pareto_lpdf() 110 += 1 / alpha_dbl + log_y_min[n] - log_y[n]; in pareto_lpdf()
|
H A D | inv_chi_square_lpdf.hpp | 79 log_y(length(y)); in inv_chi_square_lpdf() local 82 log_y[i] = log(value_of(y_vec[i])); in inv_chi_square_lpdf() 110 logp -= (half_nu + 1.0) * log_y[n]; in inv_chi_square_lpdf() 121 - 0.5 * log_y[n]; in inv_chi_square_lpdf()
|
H A D | chi_square_lpdf.hpp | 81 log_y(length(y)); in chi_square_lpdf() local 84 log_y[i] = log(value_of(y_vec[i])); in chi_square_lpdf() 115 logp += (half_nu - 1.0) * log_y[n]; in chi_square_lpdf() 125 + log_y[n] * 0.5; in chi_square_lpdf()
|
H A D | beta_proportion_lpdf.hpp | 92 log_y(length(y)); in beta_proportion_lpdf() local 99 log_y[n] = log(value_of(y_vec[n])); in beta_proportion_lpdf() 159 logp += (mukappa_dbl - 1) * log_y[n] in beta_proportion_lpdf() 172 * (digamma_kappa_mukappa[n] - digamma_mukappa[n] + log_y[n] in beta_proportion_lpdf() 176 += digamma_kappa[n] + mu_dbl * (log_y[n] - digamma_mukappa[n]) in beta_proportion_lpdf()
|
H A D | gamma_lpdf.hpp | 92 log_y(length(y)); in gamma_lpdf() local 96 log_y[n] = log(value_of(y_vec[n])); in gamma_lpdf() 130 logp += (alpha_dbl - 1.0) * log_y[n]; in gamma_lpdf() 138 += -digamma_alpha[n] + log_beta[n] + log_y[n]; in gamma_lpdf()
|
H A D | inv_gamma_lpdf.hpp | 88 log_y(length(y)); in inv_gamma_lpdf() local 95 log_y[n] = log(value_of(y_vec[n])); in inv_gamma_lpdf() 129 logp -= (alpha_dbl + 1.0) * log_y[n]; in inv_gamma_lpdf() 138 += -digamma_alpha[n] + log_beta[n] - log_y[n]; in inv_gamma_lpdf()
|
H A D | frechet_lpdf.hpp | 67 log_y(length(y)); in frechet_lpdf() local 70 log_y[i] = log(value_of(y_vec[i])); in frechet_lpdf() 102 logp -= (alpha_dbl + 1.0) * log_y[n]; in frechet_lpdf() 117 + (1.0 - sigma_div_y_pow_alpha[n]) * (log_sigma[n] - log_y[n]); in frechet_lpdf()
|
H A D | weibull_lpdf.hpp | 83 log_y(length(y)); in weibull_lpdf() local 86 log_y[i] = log(value_of(y_vec[i])); in weibull_lpdf() 118 logp += (alpha_dbl - 1.0) * log_y[n]; in weibull_lpdf() 133 + (1.0 - y_div_sigma_pow_alpha[n]) * (log_y[n] - log_sigma[n]); in weibull_lpdf()
|
H A D | lognormal_lpdf.hpp | 88 log_y(length(y)); in lognormal_lpdf() local 91 log_y[n] = log(value_of(y_vec[n])); in lognormal_lpdf() 109 logy_m_mu = log_y[n] - mu_dbl; in lognormal_lpdf() 119 logp -= log_y[n]; in lognormal_lpdf()
|
H A D | scaled_inv_chi_square_lpdf.hpp | 92 log_y(length(y)); in scaled_inv_chi_square_lpdf() local 95 log_y[i] = log(value_of(y_vec[i])); in scaled_inv_chi_square_lpdf() 135 logp -= (half_nu[n] + 1.0) * log_y[n]; in scaled_inv_chi_square_lpdf() 147 - 0.5 * log_y[n] - 0.5 * s_dbl * s_dbl * inv_y[n]; in scaled_inv_chi_square_lpdf()
|
/dports/math/stanmath/math-4.2.0/stan/math/opencl/prim/ |
H A D | pareto_lpdf.hpp | 74 auto log_y = log(y_val); in pareto_lpdf() local 81 logp1 - elt_multiply(alpha_val, log_y) - log_y, logp1); in pareto_lpdf() 88 auto alpha_deriv = elt_divide(1.0, alpha_val) + log_y_min - log_y; in pareto_lpdf()
|
H A D | inv_chi_square_lpdf.hpp | 78 auto log_y = log(y_val); in inv_chi_square_lpdf() local 82 auto logp1 = -elt_multiply(half_nu + 1.0, log_y); in inv_chi_square_lpdf() 90 auto nu_deriv = -HALF_LOG_TWO - (digamma(half_nu) + log_y) * 0.5; in inv_chi_square_lpdf()
|
/dports/graphics/py-plotly/plotly-4.14.3/plotly/express/ |
H A D | _chart_types.py | 51 log_y=False, 95 log_y=False, 164 log_y=False, 238 log_y=False, 282 log_y=False, 338 log_y=False, 433 log_y=False, 496 log_y=False, argument 545 log_y=False, argument 597 log_y=False, argument [all …]
|
/dports/games/uqm/uqm-0.8.0/src/uqm/ |
H A D | hyper.c | 86 GLOBAL_SIS (log_y) -= new_dy; in MoveSIS() 87 if (GLOBAL_SIS (log_y) < 0) in MoveSIS() 89 new_dy += (SIZE)GLOBAL_SIS (log_y); in MoveSIS() 90 GLOBAL_SIS (log_y) = 0; in MoveSIS() 92 else if (GLOBAL_SIS (log_y) > MAX_Y_LOGICAL) in MoveSIS() 95 GLOBAL_SIS (log_y) = MAX_Y_LOGICAL; in MoveSIS() 401 SDWORD log_x, log_y; in ElementToUniverse() local 405 log_y = GLOBAL_SIS (log_y) in ElementToUniverse() 408 pPt->y = LOGY_TO_UNIVERSE (log_y); in ElementToUniverse() 1237 EncounterPtr->log_y -= delta_y; in ProcessEncounter() [all …]
|
/dports/math/octave-forge-statistics/statistics-1.4.3/inst/ |
H A D | iwishpdf.m | 10 ## @deftypefn {Function File} {} @var{y} = iwishpdf (@var{W}, @var{Tau}, @var{df}, @var{log_y}=fal… 14 ## If the flag @var{log_y} is set, return the log probability density -- this helps avoid underflow… 24 function [y] = iwishpdf(W, Tau, df, log_y=false) 50 if ~log_y
|
H A D | wishpdf.m | 10 ## @deftypefn {Function File} {} @var{y} = wishpdf (@var{W}, @var{Sigma}, @var{df}, @var{log_y}=fa… 14 ## If the flag @var{log_y} is set, return the log probability density -- this helps avoid underflow… 24 function [y] = wishpdf(W, Sigma, df, log_y=false) 50 if ~log_y
|