mirror of
https://gitlab.com/libeigen/eigen.git
synced 2025-06-04 18:54:00 +08:00
Merge with upstream
This commit is contained in:
commit
8ba799805b
@ -66,6 +66,7 @@ namespace Eigen
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(sinh,scalar_sinh_op,hyperbolic sine,\sa ArrayBase::sinh)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(cosh,scalar_cosh_op,hyperbolic cosine,\sa ArrayBase::cosh)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(tanh,scalar_tanh_op,hyperbolic tangent,\sa ArrayBase::tanh)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(logistic,scalar_logistic_op,logistic function,\sa ArrayBase::logistic)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(lgamma,scalar_lgamma_op,natural logarithm of the gamma function,\sa ArrayBase::lgamma)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(digamma,scalar_digamma_op,derivative of lgamma,\sa ArrayBase::digamma)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(erf,scalar_erf_op,error function,\sa ArrayBase::erf)
|
||||
@ -89,7 +90,7 @@ namespace Eigen
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(isinf,scalar_isinf_op,infinite value test,\sa Eigen::isnan DOXCOMMA Eigen::isfinite DOXCOMMA ArrayBase::isinf)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(isfinite,scalar_isfinite_op,finite value test,\sa Eigen::isinf DOXCOMMA Eigen::isnan DOXCOMMA ArrayBase::isfinite)
|
||||
EIGEN_ARRAY_DECLARE_GLOBAL_UNARY(sign,scalar_sign_op,sign (or 0),\sa ArrayBase::sign)
|
||||
|
||||
|
||||
/** \returns an expression of the coefficient-wise power of \a x to the given constant \a exponent.
|
||||
*
|
||||
* \tparam ScalarExponent is the scalar type of \a exponent. It must be compatible with the scalar type of the given expression (\c Derived::Scalar).
|
||||
@ -124,21 +125,21 @@ namespace Eigen
|
||||
*
|
||||
* Example: \include Cwise_array_power_array.cpp
|
||||
* Output: \verbinclude Cwise_array_power_array.out
|
||||
*
|
||||
*
|
||||
* \sa ArrayBase::pow()
|
||||
*
|
||||
* \relates ArrayBase
|
||||
*/
|
||||
template<typename Derived,typename ExponentDerived>
|
||||
inline const Eigen::CwiseBinaryOp<Eigen::internal::scalar_pow_op<typename Derived::Scalar, typename ExponentDerived::Scalar>, const Derived, const ExponentDerived>
|
||||
pow(const Eigen::ArrayBase<Derived>& x, const Eigen::ArrayBase<ExponentDerived>& exponents)
|
||||
pow(const Eigen::ArrayBase<Derived>& x, const Eigen::ArrayBase<ExponentDerived>& exponents)
|
||||
{
|
||||
return Eigen::CwiseBinaryOp<Eigen::internal::scalar_pow_op<typename Derived::Scalar, typename ExponentDerived::Scalar>, const Derived, const ExponentDerived>(
|
||||
x.derived(),
|
||||
exponents.derived()
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
/** \returns an expression of the coefficient-wise power of the scalar \a x to the given array of \a exponents.
|
||||
*
|
||||
* This function computes the coefficient-wise power between a scalar and an array of exponents.
|
||||
@ -147,7 +148,7 @@ namespace Eigen
|
||||
*
|
||||
* Example: \include Cwise_scalar_power_array.cpp
|
||||
* Output: \verbinclude Cwise_scalar_power_array.out
|
||||
*
|
||||
*
|
||||
* \sa ArrayBase::pow()
|
||||
*
|
||||
* \relates ArrayBase
|
||||
|
@ -701,7 +701,7 @@ template<typename Scalar> struct scalar_isnan_op {
|
||||
EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE result_type operator() (const Scalar& a) const {
|
||||
#if defined(__SYCL_DEVICE_ONLY__)
|
||||
return numext::isnan(a);
|
||||
#else
|
||||
#else
|
||||
return (numext::isnan)(a);
|
||||
#endif
|
||||
}
|
||||
@ -815,7 +815,7 @@ struct scalar_sign_op<Scalar,true> {
|
||||
template<typename Scalar>
|
||||
struct functor_traits<scalar_sign_op<Scalar> >
|
||||
{ enum {
|
||||
Cost =
|
||||
Cost =
|
||||
NumTraits<Scalar>::IsComplex
|
||||
? ( 8*NumTraits<Scalar>::MulCost ) // roughly
|
||||
: ( 3*NumTraits<Scalar>::AddCost),
|
||||
@ -823,6 +823,34 @@ struct functor_traits<scalar_sign_op<Scalar> >
|
||||
};
|
||||
};
|
||||
|
||||
/** \internal
|
||||
* \brief Template functor to compute the logistic function of a scalar
|
||||
* \sa class CwiseUnaryOp, ArrayBase::logistic()
|
||||
*/
|
||||
template <typename T>
|
||||
struct scalar_logistic_op {
|
||||
EIGEN_EMPTY_STRUCT_CTOR(scalar_logistic_op)
|
||||
EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE T operator()(const T& x) const {
|
||||
const T one = T(1);
|
||||
return one / (one + numext::exp(-x));
|
||||
}
|
||||
|
||||
template <typename Packet> EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE
|
||||
Packet packetOp(const Packet& x) const {
|
||||
const Packet one = pset1<Packet>(T(1));
|
||||
return pdiv(one, padd(one, pexp(pnegate(x))));
|
||||
}
|
||||
};
|
||||
template <typename T>
|
||||
struct functor_traits<scalar_logistic_op<T> > {
|
||||
enum {
|
||||
Cost = NumTraits<T>::AddCost * 2 + NumTraits<T>::MulCost * 6,
|
||||
PacketAccess = packet_traits<T>::HasAdd && packet_traits<T>::HasDiv &&
|
||||
packet_traits<T>::HasNegate && packet_traits<T>::HasExp
|
||||
};
|
||||
};
|
||||
|
||||
|
||||
} // end namespace internal
|
||||
|
||||
} // end namespace Eigen
|
||||
|
@ -21,6 +21,7 @@ typedef CwiseUnaryOp<internal::scalar_acos_op<Scalar>, const Derived> AcosReturn
|
||||
typedef CwiseUnaryOp<internal::scalar_asin_op<Scalar>, const Derived> AsinReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_atan_op<Scalar>, const Derived> AtanReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_tanh_op<Scalar>, const Derived> TanhReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_logistic_op<Scalar>, const Derived> LogisticReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_sinh_op<Scalar>, const Derived> SinhReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_cosh_op<Scalar>, const Derived> CoshReturnType;
|
||||
typedef CwiseUnaryOp<internal::scalar_square_op<Scalar>, const Derived> SquareReturnType;
|
||||
@ -335,6 +336,15 @@ cosh() const
|
||||
return CoshReturnType(derived());
|
||||
}
|
||||
|
||||
/** \returns an expression of the coefficient-wise logistic of *this.
|
||||
*/
|
||||
EIGEN_DEVICE_FUNC
|
||||
inline const LogisticReturnType
|
||||
logistic() const
|
||||
{
|
||||
return LogisticReturnType(derived());
|
||||
}
|
||||
|
||||
/** \returns an expression of the coefficient-wise inverse of *this.
|
||||
*
|
||||
* Example: \include Cwise_inverse.cpp
|
||||
|
@ -231,6 +231,7 @@ template<typename ArrayType> void array_real(const ArrayType& m)
|
||||
VERIFY_IS_APPROX(m1.sinh(), sinh(m1));
|
||||
VERIFY_IS_APPROX(m1.cosh(), cosh(m1));
|
||||
VERIFY_IS_APPROX(m1.tanh(), tanh(m1));
|
||||
VERIFY_IS_APPROX(m1.logistic(), logistic(m1));
|
||||
|
||||
VERIFY_IS_APPROX(m1.arg(), arg(m1));
|
||||
VERIFY_IS_APPROX(m1.round(), round(m1));
|
||||
@ -266,6 +267,7 @@ template<typename ArrayType> void array_real(const ArrayType& m)
|
||||
VERIFY_IS_APPROX(sinh(m1), 0.5*(exp(m1)-exp(-m1)));
|
||||
VERIFY_IS_APPROX(cosh(m1), 0.5*(exp(m1)+exp(-m1)));
|
||||
VERIFY_IS_APPROX(tanh(m1), (0.5*(exp(m1)-exp(-m1)))/(0.5*(exp(m1)+exp(-m1))));
|
||||
VERIFY_IS_APPROX(logistic(m1), (1.0/(1.0+exp(-m1))));
|
||||
VERIFY_IS_APPROX(arg(m1), ((m1<0).template cast<Scalar>())*std::acos(-1.0));
|
||||
VERIFY((round(m1) <= ceil(m1) && round(m1) >= floor(m1)).all());
|
||||
VERIFY((Eigen::isnan)((m1*0.0)/0.0).all());
|
||||
@ -345,6 +347,7 @@ template<typename ArrayType> void array_complex(const ArrayType& m)
|
||||
VERIFY_IS_APPROX(m1.sinh(), sinh(m1));
|
||||
VERIFY_IS_APPROX(m1.cosh(), cosh(m1));
|
||||
VERIFY_IS_APPROX(m1.tanh(), tanh(m1));
|
||||
VERIFY_IS_APPROX(m1.logistic(), logistic(m1));
|
||||
VERIFY_IS_APPROX(m1.arg(), arg(m1));
|
||||
VERIFY((m1.isNaN() == (Eigen::isnan)(m1)).all());
|
||||
VERIFY((m1.isInf() == (Eigen::isinf)(m1)).all());
|
||||
@ -368,6 +371,7 @@ template<typename ArrayType> void array_complex(const ArrayType& m)
|
||||
VERIFY_IS_APPROX(sinh(m1), 0.5*(exp(m1)-exp(-m1)));
|
||||
VERIFY_IS_APPROX(cosh(m1), 0.5*(exp(m1)+exp(-m1)));
|
||||
VERIFY_IS_APPROX(tanh(m1), (0.5*(exp(m1)-exp(-m1)))/(0.5*(exp(m1)+exp(-m1))));
|
||||
VERIFY_IS_APPROX(logistic(m1), (1.0/(1.0 + exp(-m1))));
|
||||
|
||||
for (Index i = 0; i < m.rows(); ++i)
|
||||
for (Index j = 0; j < m.cols(); ++j)
|
||||
|
@ -187,7 +187,7 @@ struct TensorEvaluator<const TensorAssignOp<LeftArgType, RightArgType>, Device>
|
||||
|
||||
EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE void evalBlock(TensorBlock* block) {
|
||||
if (TensorEvaluator<LeftArgType, Device>::RawAccess &&
|
||||
m_leftImpl.data() != nullptr) {
|
||||
m_leftImpl.data() != NULL) {
|
||||
TensorBlock left_block(block->first_coeff_index(), block->block_sizes(),
|
||||
block->tensor_strides(), block->tensor_strides(),
|
||||
m_leftImpl.data() + block->first_coeff_index());
|
||||
|
@ -200,9 +200,9 @@ class TensorBase<Derived, ReadOnlyAccessors>
|
||||
}
|
||||
|
||||
EIGEN_DEVICE_FUNC
|
||||
EIGEN_STRONG_INLINE const TensorCwiseUnaryOp<internal::scalar_sigmoid_op<Scalar>, const Derived>
|
||||
EIGEN_STRONG_INLINE const TensorCwiseUnaryOp<internal::scalar_logistic_op<Scalar>, const Derived>
|
||||
sigmoid() const {
|
||||
return unaryExpr(internal::scalar_sigmoid_op<Scalar>());
|
||||
return unaryExpr(internal::scalar_logistic_op<Scalar>());
|
||||
}
|
||||
|
||||
EIGEN_DEVICE_FUNC
|
||||
|
@ -155,8 +155,8 @@ struct TensorBlockCopyOp {
|
||||
typedef const Eigen::Array<Scalar, Dynamic, 1> Src;
|
||||
typedef Eigen::Array<Scalar, Dynamic, 1> Dst;
|
||||
|
||||
typedef Eigen::Map<Src, 0, InnerStride<>> SrcMap;
|
||||
typedef Eigen::Map<Dst, 0, InnerStride<>> DstMap;
|
||||
typedef Eigen::Map<Src, 0, InnerStride<> > SrcMap;
|
||||
typedef Eigen::Map<Dst, 0, InnerStride<> > DstMap;
|
||||
|
||||
const SrcMap src(src_base, num_coeff_to_copy, InnerStride<>(src_stride));
|
||||
DstMap dst(dst_base, num_coeff_to_copy, InnerStride<>(dst_stride));
|
||||
@ -405,9 +405,9 @@ struct TensorBlockCwiseBinaryOp {
|
||||
typedef const Eigen::Array<RightScalar, Dynamic, 1> Rhs;
|
||||
typedef Eigen::Array<OutputScalar, Dynamic, 1> Out;
|
||||
|
||||
typedef Eigen::Map<Lhs, 0, InnerStride<>> LhsMap;
|
||||
typedef Eigen::Map<Rhs, 0, InnerStride<>> RhsMap;
|
||||
typedef Eigen::Map<Out, 0, InnerStride<>> OutMap;
|
||||
typedef Eigen::Map<Lhs, 0, InnerStride<> > LhsMap;
|
||||
typedef Eigen::Map<Rhs, 0, InnerStride<> > RhsMap;
|
||||
typedef Eigen::Map<Out, 0, InnerStride<> > OutMap;
|
||||
|
||||
const LeftScalar* lhs_base = &left_data[left_index];
|
||||
const RightScalar* rhs_base = &right_data[right_index];
|
||||
@ -501,7 +501,7 @@ struct TensorBlockCwiseBinaryIO {
|
||||
if (size == 1) {
|
||||
continue;
|
||||
}
|
||||
auto& state = block_iter_state[num_squeezed_dims];
|
||||
BlockIteratorState& state = block_iter_state[num_squeezed_dims];
|
||||
state.output_stride = block_strides[dim];
|
||||
state.left_stride = left_strides[dim];
|
||||
state.right_stride = right_strides[dim];
|
||||
@ -523,7 +523,7 @@ struct TensorBlockCwiseBinaryIO {
|
||||
right_stride, right_data);
|
||||
// Update index.
|
||||
for (int j = 0; j < num_squeezed_dims; ++j) {
|
||||
auto& state = block_iter_state[j];
|
||||
BlockIteratorState& state = block_iter_state[j];
|
||||
if (++state.count < state.size) {
|
||||
output_index += state.output_stride;
|
||||
left_index += state.left_stride;
|
||||
|
@ -102,7 +102,7 @@ class Allocator {
|
||||
// Build a thread pool device on top the an existing pool of threads.
|
||||
struct ThreadPoolDevice {
|
||||
// The ownership of the thread pool remains with the caller.
|
||||
ThreadPoolDevice(ThreadPoolInterface* pool, int num_cores, Allocator* allocator = nullptr)
|
||||
ThreadPoolDevice(ThreadPoolInterface* pool, int num_cores, Allocator* allocator = NULL)
|
||||
: pool_(pool), num_threads_(num_cores), allocator_(allocator) { }
|
||||
|
||||
EIGEN_STRONG_INLINE void* allocate(size_t num_bytes) const {
|
||||
@ -282,7 +282,7 @@ struct ThreadPoolDevice {
|
||||
// Convenience wrapper for parallelFor that does not align blocks.
|
||||
void parallelFor(Index n, const TensorOpCost& cost,
|
||||
std::function<void(Index, Index)> f) const {
|
||||
parallelFor(n, cost, nullptr, std::move(f));
|
||||
parallelFor(n, cost, NULL, std::move(f));
|
||||
}
|
||||
|
||||
// Thread pool accessor.
|
||||
|
@ -227,7 +227,7 @@ class TensorExecutor<Expression, ThreadPoolDevice, Vectorizable, Tileable> {
|
||||
typedef EvalRange<Evaluator, StorageIndex, Vectorizable> EvalRange;
|
||||
|
||||
Evaluator evaluator(expr, device);
|
||||
const bool needs_assign = evaluator.evalSubExprsIfNeeded(nullptr);
|
||||
const bool needs_assign = evaluator.evalSubExprsIfNeeded(NULL);
|
||||
if (needs_assign) {
|
||||
const StorageIndex PacketSize =
|
||||
Vectorizable
|
||||
@ -257,7 +257,6 @@ class TensorExecutor<Expression, ThreadPoolDevice, Vectorizable, /*Tileable*/ tr
|
||||
|
||||
static EIGEN_STRONG_INLINE void run(const Expression& expr,
|
||||
const ThreadPoolDevice& device) {
|
||||
typedef TensorBlock<ScalarNoConst, StorageIndex, NumDims, Evaluator::Layout> TensorBlock;
|
||||
typedef TensorBlockMapper<ScalarNoConst, StorageIndex, NumDims, Evaluator::Layout> TensorBlockMapper;
|
||||
|
||||
Evaluator evaluator(expr, device);
|
||||
@ -271,7 +270,7 @@ class TensorExecutor<Expression, ThreadPoolDevice, Vectorizable, /*Tileable*/ tr
|
||||
return;
|
||||
}
|
||||
|
||||
const bool needs_assign = evaluator.evalSubExprsIfNeeded(nullptr);
|
||||
const bool needs_assign = evaluator.evalSubExprsIfNeeded(NULL);
|
||||
if (needs_assign) {
|
||||
TensorBlockShapeType block_shape = TensorBlockShapeType::kSkewedInnerDims;
|
||||
Index block_total_size = 0;
|
||||
|
@ -54,36 +54,6 @@ struct functor_traits<scalar_fmod_op<Scalar> > {
|
||||
PacketAccess = false };
|
||||
};
|
||||
|
||||
|
||||
/** \internal
|
||||
* \brief Template functor to compute the sigmoid of a scalar
|
||||
* \sa class CwiseUnaryOp, ArrayBase::sigmoid()
|
||||
*/
|
||||
template <typename T>
|
||||
struct scalar_sigmoid_op {
|
||||
EIGEN_EMPTY_STRUCT_CTOR(scalar_sigmoid_op)
|
||||
EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE T operator()(const T& x) const {
|
||||
const T one = T(1);
|
||||
return one / (one + numext::exp(-x));
|
||||
}
|
||||
|
||||
template <typename Packet> EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE
|
||||
Packet packetOp(const Packet& x) const {
|
||||
const Packet one = pset1<Packet>(T(1));
|
||||
return pdiv(one, padd(one, pexp(pnegate(x))));
|
||||
}
|
||||
};
|
||||
|
||||
template <typename T>
|
||||
struct functor_traits<scalar_sigmoid_op<T> > {
|
||||
enum {
|
||||
Cost = NumTraits<T>::AddCost * 2 + NumTraits<T>::MulCost * 6,
|
||||
PacketAccess = packet_traits<T>::HasAdd && packet_traits<T>::HasDiv &&
|
||||
packet_traits<T>::HasNegate && packet_traits<T>::HasExp
|
||||
};
|
||||
};
|
||||
|
||||
|
||||
template<typename Reducer, typename Device>
|
||||
struct reducer_traits {
|
||||
enum {
|
||||
|
@ -428,7 +428,7 @@ void test_threadpool_allocate(TestAllocator* allocator)
|
||||
void* ptr = device.allocate(512);
|
||||
device.deallocate(ptr);
|
||||
}
|
||||
VERIFY(allocator != nullptr);
|
||||
VERIFY(allocator != NULL);
|
||||
VERIFY_IS_EQUAL(allocator->alloc_count(), num_allocs);
|
||||
VERIFY_IS_EQUAL(allocator->dealloc_count(), num_allocs);
|
||||
}
|
||||
@ -460,7 +460,7 @@ EIGEN_DECLARE_TEST(cxx11_tensor_thread_pool)
|
||||
CALL_SUBTEST_6(test_multithread_random());
|
||||
|
||||
TestAllocator test_allocator;
|
||||
CALL_SUBTEST_6(test_multithread_shuffle<ColMajor>(nullptr));
|
||||
CALL_SUBTEST_6(test_multithread_shuffle<ColMajor>(NULL));
|
||||
CALL_SUBTEST_6(test_multithread_shuffle<RowMajor>(&test_allocator));
|
||||
CALL_SUBTEST_6(test_threadpool_allocate(&test_allocator));
|
||||
}
|
||||
|
Loading…
x
Reference in New Issue
Block a user