static void apply(Alpha a, const V1 &x, const V2 &y, Beta b, V3 &z)
{
if (!math::is_zero(b))
z.array() = a * x.array() * y.array() + b * z.array();
else
z.array() = a * x.array() * y.array();
}
void foo( D *p )
{
for( int i = 0; i < 10; ++i ) {
V1 *q = p; // assignment to 'q' should be brought
// outside the loop
q->bar( p->V1::v1 + p->V2::v1, p->V1::v2 + p->V2::v2 );
}
}
void bi::renormalise(V1 lws) {
thrust::replace_if(lws.begin(), lws.end(), is_not_finite_functor<real>(),
bi::log(0.0));
real mx = max_reduce(lws);
if (is_finite(mx)) {
sub_elements(lws, mx, lws);
}
}
void bi::cov(const M1 X, const V1 mu, M2 Sigma) {
/* pre-conditions */
BI_ASSERT(X.size2() == mu.size());
BI_ASSERT(Sigma.size1() == mu.size() && Sigma.size2() == mu.size());
const int N = X.size1();
typename sim_temp_matrix<M2>::type Y(X.size1(), X.size2());
Y = X;
sub_rows(Y, mu);
syrk(1.0/(N - 1.0), Y, 0.0, Sigma, 'U', 'T');
}
ExecStatus
Int<V0,V1,Idx,Val>::post(Home home, IntSharedArray& c, V0 x0, V1 x1) {
if (x0.assigned()) {
GECODE_ME_CHECK(x1.eq(home,c[x0.val()]));
} else if (x1.assigned()) {
GECODE_ES_CHECK(assigned_val(home,c,x0,x1));
} else {
(void) new (home) Int<V0,V1,Idx,Val>(home,c,x0,x1);
}
return ES_OK;
}
void
MiscCheckTheParallelEnvironment(const V1& vec1, const V2& vec2)
{
const BaseEnvironment& env = vec1.env();
if (env.numSubEnvironments() == (unsigned int) env.fullComm().NumProc()) {
UQ_FATAL_TEST_MACRO(env.subRank() != 0,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"there should exist only one processor per sub environment");
UQ_FATAL_TEST_MACRO((vec1.numOfProcsForStorage() != 1) ||
(vec2.numOfProcsForStorage() != 1),
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"only 1 processor (per sub environment) should be necessary for the storage of a parameter vector");
}
else if (env.numSubEnvironments() < (unsigned int) env.fullComm().NumProc()) {
UQ_FATAL_TEST_MACRO(env.fullComm().NumProc()%env.numSubEnvironments() != 0,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"total number of processors should be a multiple of the number of sub environments");
unsigned int numProcsPerSubEnvironment = env.fullComm().NumProc()/env.numSubEnvironments();
UQ_FATAL_TEST_MACRO(env.subComm().NumProc() != (int) numProcsPerSubEnvironment,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"inconsistent number of processors per sub environment");
if ((vec1.numOfProcsForStorage() == 1) &&
(vec2.numOfProcsForStorage() == 1)) {
// Ok
}
else if ((vec1.numOfProcsForStorage() == numProcsPerSubEnvironment) &&
(vec2.numOfProcsForStorage() == numProcsPerSubEnvironment)) {
UQ_FATAL_TEST_MACRO(true,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"parallel vectors are not supported yet");
}
else {
UQ_FATAL_TEST_MACRO(true,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"number of processors required for a vector storage should be equal to either 1 or to the number of processors in the sub environment");
}
}
else {
UQ_FATAL_TEST_MACRO(true,
env.worldRank(),
"MiscCheckTheParallelEnvironment<V1,V2>()",
"number of processors per sub environment is less than 1!");
}
return;
}
void bi::var(const M1 X, const V1 mu, V2 sigma) {
/* pre-conditions */
BI_ASSERT(X.size2() == mu.size());
BI_ASSERT(sigma.size() == mu.size());
const int N = X.size1();
typename sim_temp_matrix<M1>::type Z(X.size2(), X.size1());
Z = X;
sub_rows(Z, mu);
dot_columns(Z, sigma);
scal(1.0/(N - 1.0), sigma);
}
void bi::cross(const M1 X, const M2 Y, const V1 muX, const V2 muY,
M3 SigmaXY) {
/* pre-conditions */
BI_ASSERT(X.size2() == muX.size());
BI_ASSERT(Y.size2() == muY.size());
BI_ASSERT(X.size1() == Y.size1());
BI_ASSERT(SigmaXY.size1() == muX.size() && SigmaXY.size2() == muY.size());
const int N = X.size1();
gemm(1.0/(N - 1.0), X, Y, 0.0, SigmaXY, 'T', 'N');
ger(-N/(N - 1.0), muX, muY, SigmaXY);
}
void bi::inverse_gamma_log_densities(const M1 Z, const T1 alpha, const T1 beta,
V1 p, const bool clear) {
/* pre-condition */
BI_ASSERT(Z.size1() == p.size());
op_elements(vec(Z), vec(Z), inverse_gamma_log_density_functor<T1>(alpha, beta));
if (clear) {
sum_columns(Z, p);
} else {
typename sim_temp_vector<V1>::type p1(p.size());
sum_columns(Z, p1);
add_elements(p, p1, p);
}
}
void bi::MetropolisResamplerHost::ancestors(Random& rng, const V1 lws,
V2 as, int B) {
const int P1 = lws.size(); // number of particles
const int P2 = as.size(); // number of ancestors to draw
#pragma omp parallel
{
real alpha, lw1, lw2;
int k, p1, p2, p;
#pragma omp for
for (p = 0; p < P2; ++p) {
p1 = p;
lw1 = lws(p);
for (k = 0; k < B; ++k) {
p2 = rng.uniformInt(0, P1 - 1);
lw2 = lws(p2);
alpha = rng.uniform<real>();
if (bi::log(alpha) < lw2 - lw1) {
/* accept */
p1 = p2;
lw1 = lw2;
}
}
/* write result */
as(p) = p1;
}
}
}
result_type apply_incompatible( const V1& src, const V2& dst) const
{
//copy_pixels( color_converted_view<typename V2::value_type>( src, _cc),dst);
tbb::parallel_for( tbb::blocked_range<std::size_t>( 0, src.height()),
detail::make_tbb_copy_pixels_fun( color_converted_view<typename V2::value_type>( src, _cc), dst),
GIL_TBB_ALGORITHMS_DEFAULT_PARTITIONER());
}
void bi::gaussian_log_densities(const M1 Z, const T1 logZ, V1 p,
const bool clear) {
/* pre-condition */
BI_ASSERT(Z.size1() == p.size());
typedef typename V1::value_type T2;
if (clear) {
dot_rows(Z, p);
op_elements(p, p, gaussian_log_density_functor<T2>(logZ));
} else {
typename sim_temp_vector<V1>::type p1(p.size());
dot_rows(Z, p1);
op_elements(p1, p, p, gaussian_log_density_update_functor<T2>(logZ));
}
}
typename viennacl::enable_if< viennacl::is_any_dense_nonstructured_matrix<M1>::value
&& viennacl::is_any_dense_nonstructured_vector<V1>::value
>::type
inplace_solve(const matrix_expression< const M1, const M1, op_trans> & proxy,
V1 & vec,
SOLVERTAG)
{
assert( (proxy.lhs().size1() == vec.size()) && bool("Size check failed in inplace_solve(): size1(A) != size(b)"));
assert( (proxy.lhs().size2() == vec.size()) && bool("Size check failed in inplace_solve(): size2(A) != size(b)"));
switch (viennacl::traits::handle(proxy.lhs()).get_active_handle_id())
{
case viennacl::MAIN_MEMORY:
viennacl::linalg::host_based::inplace_solve(proxy, vec, SOLVERTAG());
break;
#ifdef VIENNACL_WITH_OPENCL
case viennacl::OPENCL_MEMORY:
viennacl::linalg::opencl::inplace_solve(proxy, vec, SOLVERTAG());
break;
#endif
#ifdef VIENNACL_WITH_CUDA
case viennacl::CUDA_MEMORY:
viennacl::linalg::cuda::inplace_solve(proxy, vec, SOLVERTAG());
break;
#endif
default:
throw "not implemented";
}
}
void rot (const T1 &t1, V1 &v1, const T2 &t2, V2 &v2)
{
typedef typename promote_traits<typename V1::value_type, typename V2::value_type>::promote_type promote_type;
vector<promote_type> vt (t1 * v1 + t2 * v2);
v2.assign (- t2 * v1 + t1 * v2);
v1.assign (vt);
}
double inner_prod( const V1& v1, const V2& v2 ) {
double inner_res;
inner_res = 0;
for (int i=0; i<v1.size(); ++i) {
inner_res += v1[i]*v2[i] ;
}
return inner_res;
}
void bi::MultinomialResamplerHost::ancestors(Random& rng, const V1 lws, V2 as,
MultinomialPrecompute<ON_HOST>& pre)
throw (ParticleFilterDegeneratedException) {
typedef typename V1::value_type T1;
const int P = as.size();
const int lwsSize = lws.size();
T1 lW;
/* weights */
if (pre.W > 0) {
lW = bi::log(pre.W);
#pragma omp parallel
{
int Q = P/bi_omp_max_threads;
int start = bi_omp_tid*Q + bi::min(bi_omp_tid, P % bi_omp_max_threads); // min() handles leftovers
if (bi_omp_tid < P % bi_omp_max_threads) {
++Q; // pick up a leftover
}
int i, j = lwsSize;
T1 lMax = 0.0, lu;
for (i = Q; i > 0; --i) {
lMax += bi::log(rng.uniform<T1>())/i;
lu = lW + lMax;
while (j > 0 && lu < bi::log(pre.Ws(j - 1))) {
--j;
}
if (pre.sort) {
as(start + i - 1) = pre.ps(j);
} else {
as(start + i - 1) = j;
}
}
}
} else {
throw ParticleFilterDegeneratedException();
}
/* post-condition */
BI_ASSERT(max_reduce(as) < lws.size());
}
void bi::mean(const M1 X, V1 mu) {
/* pre-condition */
BI_ASSERT(X.size2() == mu.size());
const int N = X.size1();
typename sim_temp_vector<V1>::type w(N);
set_elements(w, 1.0);
gemv(1.0/N, X, w, 0.0, mu, 'T');
}
inline void bi::mean(const GammaPdf& q, V1 mu) {
/* pre-condition */
BI_ASSERT(mu.size() == q.size());
real alpha = q.shape();
real beta = q.scale();
set_elements(mu, alpha*beta);
}
void bi::mean(const M1 X, const V1 w, V2 mu) {
/* pre-conditions */
BI_ASSERT(X.size2() == mu.size());
BI_ASSERT(X.size1() == w.size());
typedef typename V1::value_type T;
T Wt = sum_reduce(w);
gemv(1.0/Wt, X, w, 0.0, mu, 'T');
}
inline void bi::mean(const InverseGammaPdf& q, V1 mu) {
/* pre-condition */
BI_ASSERT(mu.size() == q.size());
BI_ASSERT(q.shape() > 1.0);
real alpha = q.shape();
real beta = q.scale();
set_elements(mu, alpha*std::pow(beta, 2));
}
void bi::cov(const M1 X, const V1 w, const V2 mu, M2 Sigma) {
/* pre-conditions */
BI_ASSERT(X.size2() == mu.size());
BI_ASSERT(X.size1() == w.size());
BI_ASSERT(Sigma.size1() == mu.size() && Sigma.size2() == mu.size());
typedef typename V1::value_type T;
typename sim_temp_matrix<M2>::type Y(X.size1(), X.size2());
typename sim_temp_matrix<M2>::type Z(X.size1(), X.size2());
typename sim_temp_vector<V2>::type v(w.size());
T Wt = sum_reduce(w);
Y = X;
sub_rows(Y, mu);
sqrt_elements(w, v);
gdmm(1.0, v, Y, 0.0, Z);
syrk(1.0/Wt, Z, 0.0, Sigma, 'U', 'T');
// alternative weight: 1.0/(Wt - W2t/Wt)
}
void bi::hist(const V1 x, const V2 w, V3 c, V4 h) {
/* pre-condition */
BI_ASSERT(x.size() == w.size());
BI_ASSERT(c.size() == h.size());
BI_ASSERT(!V3::on_device);
BI_ASSERT(!V4::on_device);
typedef typename V1::value_type T1;
typedef typename V2::value_type T2;
const int P = x.size();
const int B = c.size();
T1 mx, mn;
int i, j;
typename temp_host_vector<T1>::type xSorted(P);
typename temp_host_vector<T2>::type wSorted(P);
xSorted = x;
wSorted = w;
bi::sort_by_key(xSorted, wSorted);
mn = xSorted[0];
mx = xSorted[xSorted.size() - 1];
/* compute bin right edges */
for (j = 0; j < B; ++j) {
c[j] = mn + (j + 1)*(mx - mn)/B;
}
/* compute bin heights */
h.clear();
for (i = 0, j = 0; i < P; ++i) {
if (xSorted[i] >= c[j] && j < B - 1) {
++j;
}
h[j] += wSorted[i];
}
/* compute bin centres */
for (j = B - 1; j > 0; --j) {
c[j] = 0.5*(c[j - 1] + c[j]);
}
c[0] = 0.5*(mn + c[0]);
}
void bi::var(const M1 X, const V1 w, const V2 mu, V3 sigma) {
/* pre-conditions */
BI_ASSERT(X.size2() == mu.size());
BI_ASSERT(X.size1() == w.size());
BI_ASSERT(sigma.size() == mu.size());
typedef typename V1::value_type T1;
typename sim_temp_matrix<M1>::type Z(X.size1(), X.size2());
typename sim_temp_matrix<M1>::type Y(X.size1(), X.size2());
typename sim_temp_vector<V2>::type v(w.size());
T1 Wt = sum_reduce(w);
Z = X;
sub_rows(Z, mu);
sqrt_elements(w, v);
gdmm(1.0, v, Z, 0.0, Y);
dot_columns(Y, sigma);
divscal_elements(sigma, Wt, sigma);
// alternative weight: 1.0/(Wt - W2t/Wt)
}
void checkDenseVectorAssignment(V1& v1, V2 const& v2){
BOOST_REQUIRE_EQUAL(v1.size(),v2.size());
//indexed access
for(std::size_t i = 0; i != v2.size(); ++i){
v1(i) = 0;
BOOST_CHECK_EQUAL(v1(i),0);
v1(i) = v2(i);
BOOST_CHECK_EQUAL(v1(i),v2(i));
v1(i) = 0;
BOOST_CHECK_EQUAL(v1(i),0);
}
//iterator access rows
typedef typename V1::iterator Iter;
BOOST_REQUIRE_EQUAL(v1.end()-v1.begin(), v1.size());
std::size_t k = 0;
for(Iter it = v1.begin(); it != v1.end(); ++it,++k){
BOOST_CHECK_EQUAL(k,it.index());
*it = 0;
BOOST_CHECK_EQUAL(v1(k),0);
*it = v2(k);
BOOST_CHECK_EQUAL(v1(k),v2(k));
*it = 0;
BOOST_CHECK_EQUAL(v1(k),0);
}
//test that the actual iterated length equals the number of elements
BOOST_CHECK_EQUAL(k, v2.size());
}
ExecStatus
Int<V0,V1,Idx,Val>::assigned_val(Space& home, IntSharedArray& c,
V0 x0, V1 x1) {
Region r(home);
int* v = r.alloc<int>(x0.size());
int n = 0;
for (ViewValues<V0> i(x0); i(); ++i)
if (c[i.val()] != x1.val())
v[n++]=i.val();
Iter::Values::Array iv(v,n);
GECODE_ME_CHECK(x0.minus_v(home,iv,false));
return ES_OK;
}
template <typename V, typename V1> void addVectors(V &a, const V1 &b) {
#ifdef FLOAT_REALS
cblas_saxpy
#else
cblas_daxpy
#endif
(a.size(), 1, b.data(), 1, a.data(), 1);
#ifdef FLOAT_REALS
cblas_sscal
#else
cblas_dscal
#endif
(a.size(), 0.5, a.data(), 1);
}
forceinline ExecStatus
post_int(Home home, IntSharedArray& c, V0 x0, V1 x1) {
assert(c.size() > 0);
GECODE_ME_CHECK(x0.gq(home,0));
GECODE_ME_CHECK(x0.le(home,c.size()));
Support::IntType idx_type = Support::s_type(c.size());
int min = c[0];
int max = c[0];
for (int i=1; i<c.size(); i++) {
min = std::min(c[i],min); max = std::max(c[i],max);
}
GECODE_ME_CHECK(x1.gq(home,min));
GECODE_ME_CHECK(x1.lq(home,max));
Support::IntType val_type =
std::max(Support::s_type(min),Support::s_type(max));
switch (idx_type) {
case Support::IT_CHAR:
switch (val_type) {
case Support::IT_CHAR:
return Int<V0,V1,signed char,signed char>::post(home,c,x0,x1);
case Support::IT_SHRT:
return Int<V0,V1,signed char,signed short int>::post(home,c,x0,x1);
default: break;
}
break;
case Support::IT_SHRT:
switch (val_type) {
case Support::IT_CHAR:
case Support::IT_SHRT:
return Int<V0,V1,signed short int,signed short int>::post(home,c,x0,x1);
default: break;
}
break;
default: break;
}
return Int<V0,V1,signed int,signed int>::post(home,c,x0,x1);
}
template <class V1, class V2> double dot_impl(
const V1 &v1, const V2 &v2) {
assert(v1.size() == v2.size());
if(v1.stride() > 0 && v2.stride() > 0){
return ddot(v1.size(),
v1.data(), v1.stride(),
v2.data(), v2.stride());
}else{
double ans = 0;
for(int i = 0; i < v1.size(); ++i){
ans += v1[i] * v2[i];
}
return ans;
}
}
void bi::cross(const M1 X, const M2 Y, const V1 w, const V2 muX,
const V3 muY, M3 SigmaXY) {
/* pre-conditions */
BI_ASSERT(X.size2() == muX.size());
BI_ASSERT(Y.size2() == muY.size());
BI_ASSERT(X.size1() == Y.size1());
BI_ASSERT(X.size1() == w.size());
BI_ASSERT(Y.size1() == w.size());
BI_ASSERT(SigmaXY.size1() == muX.size() && SigmaXY.size2() == muY.size());
typedef typename V1::value_type T;
typename sim_temp_matrix<M3>::type Z(X.size1(), X.size2());
T Wt = sum_reduce(w);
T Wt2 = std::pow(Wt, 2);
T W2t = sumsq_reduce(w);
gdmm(1.0, w, X, 0.0, Z);
gemm(1.0/Wt, Z, Y, 0.0, SigmaXY, 'T', 'N');
ger(-1.0, muX, muY, SigmaXY);
matrix_scal(1.0/(1.0 - W2t/Wt2), SigmaXY);
}
void swap (V1 &v1, V2 &v2)
{
v1.swap (v2);
}