xgboost::common Namespace Reference

Namespaces
	aft
	base64
	namespace of base64 decoding and encoding table
	detail

Classes
struct	AFTLoss
	The AFT loss function. More...
struct	AFTParam
	Parameter structure for AFT loss and metric. More...
class	Base64InStream
	the stream that reads from base64, note we take from file pointers More...
class	Base64OutStream
	the stream that write to base64, note we take from file pointers More...
class	BlockedSpace2d
class	Column
	a column storage, to be used with ApplySplit. Note that each bin id is stored as index[i] + index_base. Different types of column index for each column allow to reduce the memory usage. More...
class	ColumnMatrix
	a collection of columns, with support for construction from GHistIndexMatrix. More...
class	ColumnSampler
	Handles selection of columns due to colsample_bytree, colsample_bylevel and colsample_bynode parameters. Should be initialised before tree construction and to reset when tree construction is completed. More...
class	CompressedBufferWriter
	Writes bit compressed symbols to a memory buffer. Use CompressedIterator to read symbols back from buffer. Currently limited to a maximum symbol size of 28 bits. More...
class	CompressedIterator
	Read symbols from a bit compressed memory buffer. Usable on device and host. More...
class	ConfigParser
	Implementation of config reader. More...
class	DenseColumn
struct	ExtremeDistribution
class	FixedSizeStream
	A simple class used to consume ‘dmlc::Stream’ all at once. More...
class	GHistBuilder
	builder for histograms of gradient statistics More...
class	HistCollection
	histogram of gradient statistics for multiple nodes More...
class	HistogramCuts
class	HostSketchContainer
struct	Index
	Optionally compressed gradient index. The compression works only with dense data. More...
class	IndexTransformIter
	Transform iterator that takes an index and calls transform operator. More...
struct	IsCatOp
struct	LogisticDistribution
class	MemStackAllocator
	A C-style array with in-stack allocation. As long as the array is smaller than MaxStackSize, it will be allocated inside the stack. Otherwise, it will be heap-allocated. More...
struct	Monitor
	Timing utility used to measure total method execution time over the lifetime of the containing object. More...
struct	NormalDistribution
struct	OptionalWeights
class	ParallelGHistBuilder
	Stores temporary histograms to compute them in parallel Supports processing multiple tree-nodes for nested parallelism Able to reduce histograms across threads in efficient way. More...
class	ParallelGroupBuilder
	multi-thread version of group builder More...
class	PartitionBuilder
class	PeekableInStream
	Input stream that support additional PeekRead operation, besides read. More...
class	QuantileSketchTemplate
	template for all quantile sketch algorithm that uses merge/prune scheme More...
class	Range
class	Range1d
class	RowSetCollection
	collection of rowset More...
struct	Sched
class	SketchContainerImpl
struct	SortedQuantile
	Quantile structure accepts sorted data, extracted from histmaker. More...
class	SortedSketchContainer
class	Span
	span class implementation, based on ISO++20 span<T>. The interface should be the same. More...
class	SparseColumn
class	StreamBufferReader
	buffer reader of the stream that allows you to get More...
struct	Timer
class	Transform
	Do Transformation on HostDeviceVectors. More...
struct	WQSummary
	experimental wsummary More...
class	WQuantileSketch
	Quantile sketch use WQSummary. More...
struct	WXQSummary
	try to do efficient pruning More...
class	WXQuantileSketch
	Quantile sketch use WXQSummary. More...

Typedefs
using	CatBitField = LBitField32
using	KCatBitField = CLBitField32
using	CompressedByteT = unsigned char
using	GHistIndexRow = Span< uint32_t const >
	A single row in global histogram index. Directly represent the global index in the histogram entry. More...
template<typename GradientSumT >
using	GHistRow = Span< xgboost::detail::GradientPairInternal< GradientSumT > >
using	MemoryFixSizeBuffer = rabit::utils::MemoryFixSizeBuffer
using	MemoryBufferStream = rabit::utils::MemoryBufferStream
using	RandomEngine = std::mt19937
	Define mt19937 as default type Random Engine. More...
using	GlobalRandomEngine = RandomEngine
	global random engine More...

Enumerations
enum	byte : unsigned char
enum	ColumnType : uint8_t { kDenseColumn, kSparseColumn }
	column type More...
enum	BinTypeSize : uint32_t { kUint8BinsTypeSize = 1, kUint16BinsTypeSize = 2, kUint32BinsTypeSize = 4 }
enum	ProbabilityDistributionType : int { ProbabilityDistributionType::kNormal = 0, ProbabilityDistributionType::kLogistic = 1, ProbabilityDistributionType::kExtreme = 2 }
	Enum encoding possible choices of probability distribution. More...
enum	CensoringType : uint8_t { CensoringType::kUncensored, CensoringType::kRightCensored, CensoringType::kLeftCensored, CensoringType::kIntervalCensored }

Functions
template<class T , std::size_t X, class U , std::size_t Y>
XGBOOST_DEVICE bool	operator== (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t X, class U , std::size_t Y>
constexpr XGBOOST_DEVICE bool	operator!= (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t X, class U , std::size_t Y>
constexpr XGBOOST_DEVICE bool	operator< (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t X, class U , std::size_t Y>
constexpr XGBOOST_DEVICE bool	operator<= (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t X, class U , std::size_t Y>
constexpr XGBOOST_DEVICE bool	operator> (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t X, class U , std::size_t Y>
constexpr XGBOOST_DEVICE bool	operator>= (Span< T, X > l, Span< U, Y > r)
template<class T , std::size_t E>
XGBOOST_DEVICE auto	as_bytes (Span< T, E > s) __span_noexcept -> Span< const byte, detail::ExtentAsBytesValue< T, E >::value >
template<class T , std::size_t E>
XGBOOST_DEVICE auto	as_writable_bytes (Span< T, E > s) __span_noexcept -> Span< byte, detail::ExtentAsBytesValue< T, E >::value >
template<typename T >
XGBOOST_DEVICE bst_cat_t	AsCat (T const &v)
XGBOOST_DEVICE bool	IsCat (Span< FeatureType const > ft, bst_feature_t fidx)
constexpr bst_cat_t	OutOfRangeCat ()
XGBOOST_DEVICE bool	InvalidCat (float cat)
template<bool validate = true>
XGBOOST_DEVICE bool	Decision (common::Span< uint32_t const > cats, float cat, bool dft_left)
void	InvalidCategory ()
void	CheckMaxCat (float max_cat, size_t n_categories)
XGBOOST_DEVICE bool	UseOneHot (uint32_t n_cats, uint32_t max_cat_to_onehot)
	Whether should we use onehot encoding for categorical data. More...
std::vector< std::string >	Split (const std::string &s, char delim)
	Split a string by delimiter. More...
template<typename T >
XGBOOST_DEVICE T	Max (T a, T b)
template<typename T >
std::string	ToString (const T &data)
template<typename T1 , typename T2 >
XGBOOST_DEVICE T1	DivRoundUp (const T1 a, const T2 b)
template<class T , std::size_t N>
constexpr auto	UnpackArr (std::array< T, N > &&arr)
template<typename Fn >
auto	MakeIndexTransformIter (Fn &&fn)
int	AllVisibleGPUs ()
void	AssertGPUSupport ()
void	AssertOneAPISupport ()
template<typename Idx , typename Container , typename V = typename Container::value_type, typename Comp = std::less<V>>
std::vector< Idx >	ArgSort (Container const &array, Comp comp=std::less< V >{})
HistogramCuts	SketchOnDMatrix (DMatrix *m, int32_t max_bins, int32_t n_threads, bool use_sorted=false, Span< float > const hessian={})
	Run CPU sketching on DMatrix. More...
template<typename GradientIndex >
int32_t XGBOOST_HOST_DEV_INLINE	BinarySearchBin (size_t begin, size_t end, GradientIndex const &data, uint32_t const fidx_begin, uint32_t const fidx_end)
template<typename GradientSumT >
void	InitilizeHistByZeroes (GHistRow< GradientSumT > hist, size_t begin, size_t end)
	fill a histogram by zeros More...
template<typename GradientSumT >
void	IncrementHist (GHistRow< GradientSumT > dst, const GHistRow< GradientSumT > add, size_t begin, size_t end)
	Increment hist as dst += add in range [begin, end) More...
template<typename GradientSumT >
void	CopyHist (GHistRow< GradientSumT > dst, const GHistRow< GradientSumT > src, size_t begin, size_t end)
	Copy hist from src to dst in range [begin, end) More...
template<typename GradientSumT >
void	SubtractionHist (GHistRow< GradientSumT > dst, const GHistRow< GradientSumT > src1, const GHistRow< GradientSumT > src2, size_t begin, size_t end)
	Compute Subtraction: dst = src1 - src2 in range [begin, end) More...
std::string	LoadSequentialFile (std::string uri, bool stream=false)
	Helper function for loading consecutive file to avoid dmlc Stream when possible. More...
std::string	FileExtension (std::string fname, bool lower=true)
	Get file extension from file name. More...
std::string	ReadAll (dmlc::Stream *fi, PeekableInStream *fp)
	Read the whole buffer from dmlc stream. More...
XGBOOST_DEVICE float	Sigmoid (float x)
	calculate the sigmoid of the input. More...
template<typename T , typename U >
constexpr XGBOOST_DEVICE bool	CloseTo (T a, U b)
	Equality test for both integer and floating point. More...
template<typename Iterator >
XGBOOST_DEVICE void	Softmax (Iterator start, Iterator end)
	Do inplace softmax transformaton on start to end. More...
template<typename Iterator >
XGBOOST_DEVICE Iterator	FindMaxIndex (Iterator begin, Iterator end)
	Find the maximum iterator within the iterators. More...
float	LogSum (float x, float y)
	perform numerically safe logsum More...
template<typename Iterator >
float	LogSum (Iterator begin, Iterator end)
	perform numerically safe logsum More...
template<typename T >
XGBOOST_DEVICE std::enable_if< std::numeric_limits< T >::is_integer, bool >::type	CheckNAN (T)
XGBOOST_DEVICE bool	CheckNAN (float x)
XGBOOST_DEVICE bool	CheckNAN (double x)
template<typename T >
XGBOOST_DEVICE T	LogGamma (T v)
GlobalRandomEngine &	GlobalRandom ()
	global singleton of a random engine. This random engine is thread-local and only visible to current thread. More...
template<typename T >
std::vector< T >	WeightedSamplingWithoutReplacement (std::vector< T > const &array, std::vector< float > const &weights, size_t n)
template<typename Func >
void	ParallelFor2d (const BlockedSpace2d &space, int nthreads, Func func)
template<typename Index , typename Func >
void	ParallelFor (Index size, int32_t n_threads, Sched sched, Func fn)
template<typename Index , typename Func >
void	ParallelFor (Index size, int32_t n_threads, Func fn)
int32_t	OmpGetThreadLimit ()
int32_t	GetCfsCPUCount () noexcept
int32_t	OmpGetNumThreads (int32_t n_threads)

Variables
constexpr std::size_t	dynamic_extent = std::numeric_limits<std::size_t>::max()
constexpr double	kPI = 3.14159265358979323846
	Constant PI. More...
constexpr double	kEulerMascheroni = 0.57721566490153286060651209008240243104215933593992
	The Euler-Mascheroni_constant. More...
constexpr size_t	kBlockThreads = 256

Typedef Documentation

CatBitField

using xgboost::common::CatBitField = typedef LBitField32

CompressedByteT

using xgboost::common::CompressedByteT = typedef unsigned char

GHistIndexRow

using xgboost::common::GHistIndexRow = typedef Span<uint32_t const>

A single row in global histogram index. Directly represent the global index in the histogram entry.

GHistRow

template<typename GradientSumT >

using xgboost::common::GHistRow = typedef Span<xgboost::detail::GradientPairInternal<GradientSumT> >

GlobalRandomEngine

using xgboost::common::GlobalRandomEngine = typedef RandomEngine

global random engine

KCatBitField

using xgboost::common::KCatBitField = typedef CLBitField32

MemoryBufferStream

using xgboost::common::MemoryBufferStream = typedef rabit::utils::MemoryBufferStream

MemoryFixSizeBuffer

using xgboost::common::MemoryFixSizeBuffer = typedef rabit::utils::MemoryFixSizeBuffer

RandomEngine

using xgboost::common::RandomEngine = typedef std::mt19937

Define mt19937 as default type Random Engine.

Enumeration Type Documentation

BinTypeSize

enum xgboost::common::BinTypeSize : uint32_t

Enumerator
kUint8BinsTypeSize
kUint16BinsTypeSize
kUint32BinsTypeSize

byte

enum xgboost::common::byte : unsigned char

strong

CensoringType

enum xgboost::common::CensoringType : uint8_t

strong

Enumerator
kUncensored
kRightCensored
kLeftCensored
kIntervalCensored

ColumnType

enum xgboost::common::ColumnType : uint8_t

column type

Enumerator
kDenseColumn
kSparseColumn

ProbabilityDistributionType

enum xgboost::common::ProbabilityDistributionType : int

strong

Enum encoding possible choices of probability distribution.

Enumerator
kNormal
kLogistic
kExtreme

Function Documentation

AllVisibleGPUs()

int xgboost::common::AllVisibleGPUs

(

)

ArgSort()

template<typename Idx , typename Container , typename V = typename Container::value_type, typename Comp = std::less<V>>

std::vector<Idx> xgboost::common::ArgSort	(	Container const &	array,
		Comp	comp = std::less<V>{}
	)

as_bytes()

template<class T , std::size_t E>

XGBOOST_DEVICE auto xgboost::common::as_bytes

(

Span< T, E >

s

)

-> Span<const byte, detail::ExtentAsBytesValue<T, E>::value>

as_writable_bytes()

template<class T , std::size_t E>

XGBOOST_DEVICE auto xgboost::common::as_writable_bytes

(

Span< T, E >

s

)

-> Span<byte, detail::ExtentAsBytesValue<T, E>::value>

AsCat()

template<typename T >

XGBOOST_DEVICE bst_cat_t xgboost::common::AsCat

(

T const &

v

)

AssertGPUSupport()

void xgboost::common::AssertGPUSupport ( )

inline

AssertOneAPISupport()

void xgboost::common::AssertOneAPISupport ( )

inline

BinarySearchBin()

template<typename GradientIndex >

int32_t XGBOOST_HOST_DEV_INLINE xgboost::common::BinarySearchBin	(	size_t	begin,
		size_t	end,
		GradientIndex const &	data,
		uint32_t const	fidx_begin,
		uint32_t const	fidx_end
	)

CheckMaxCat()

void xgboost::common::CheckMaxCat ( float  max_cat, size_t  n_categories  )

inline

CheckNAN() [1/3]

XGBOOST_DEVICE bool xgboost::common::CheckNAN ( double  x )

inline

CheckNAN() [2/3]

XGBOOST_DEVICE bool xgboost::common::CheckNAN ( float  x )

inline

CheckNAN() [3/3]

template<typename T >

XGBOOST_DEVICE std::enable_if< std::numeric_limits<T>::is_integer, bool>::type xgboost::common::CheckNAN

(

T

)

CloseTo()

template<typename T , typename U >

constexpr XGBOOST_DEVICE bool xgboost::common::CloseTo ( T  a, U  b  )

constexpr

Equality test for both integer and floating point.

CopyHist()

template<typename GradientSumT >

void xgboost::common::CopyHist	(	GHistRow< GradientSumT >	dst,
		const GHistRow< GradientSumT >	src,
		size_t	begin,
		size_t	end
	)

Copy hist from src to dst in range [begin, end)

Decision()

template<bool validate = true>

XGBOOST_DEVICE bool xgboost::common::Decision ( common::Span< uint32_t const >  cats, float  cat, bool  dft_left  )

inline

DivRoundUp()

template<typename T1 , typename T2 >

XGBOOST_DEVICE T1 xgboost::common::DivRoundUp	(	const T1	a,
		const T2	b
	)

FileExtension()

std::string xgboost::common::FileExtension	(	std::string	fname,
		bool	lower = true
	)

Get file extension from file name.

Parameters

lower

Return in lower case.

Returns

File extension without the .

FindMaxIndex()

template<typename Iterator >

XGBOOST_DEVICE Iterator xgboost::common::FindMaxIndex ( Iterator  begin, Iterator  end  )

inline

Find the maximum iterator within the iterators.

Parameters

begin	The beginning iterator.
end	The end iterator.

Returns

the iterator point to the maximum value.

Template Parameters

Iterator

The type of the iterator.

GetCfsCPUCount()

int32_t xgboost::common::GetCfsCPUCount ( )

noexcept

GlobalRandom()

GlobalRandomEngine& xgboost::common::GlobalRandom

(

)

global singleton of a random engine. This random engine is thread-local and only visible to current thread.

IncrementHist()

template<typename GradientSumT >

void xgboost::common::IncrementHist	(	GHistRow< GradientSumT >	dst,
		const GHistRow< GradientSumT >	add,
		size_t	begin,
		size_t	end
	)

Increment hist as dst += add in range [begin, end)

InitilizeHistByZeroes()

template<typename GradientSumT >

void xgboost::common::InitilizeHistByZeroes	(	GHistRow< GradientSumT >	hist,
		size_t	begin,
		size_t	end
	)

fill a histogram by zeros

InvalidCat()

XGBOOST_DEVICE bool xgboost::common::InvalidCat ( float  cat )

inline

InvalidCategory()

void xgboost::common::InvalidCategory ( )

inline

IsCat()

XGBOOST_DEVICE bool xgboost::common::IsCat ( Span< FeatureType const >  ft, bst_feature_t  fidx  )

inline

LoadSequentialFile()

std::string xgboost::common::LoadSequentialFile	(	std::string	uri,
		bool	stream = false
	)

Helper function for loading consecutive file to avoid dmlc Stream when possible.

Parameters

uri	URI or file name to file.
stream	Use dmlc Stream unconditionally if set to true. Used for running test without remote filesystem.

Returns

File content.

LogGamma()

template<typename T >

XGBOOST_DEVICE T xgboost::common::LogGamma ( T  v )

inline

LogSum() [1/2]

float xgboost::common::LogSum ( float  x, float  y  )

inline

perform numerically safe logsum

Parameters

x	left input operand
y	right input operand

Returns

log(exp(x) + exp(y))

LogSum() [2/2]

template<typename Iterator >

float xgboost::common::LogSum ( Iterator  begin, Iterator  end  )

inline

perform numerically safe logsum

Parameters

begin	The beginning iterator.
end	The end iterator.

Returns

the iterator point to the maximum value.

Template Parameters

Iterator

The type of the iterator.

MakeIndexTransformIter()

template<typename Fn >

auto xgboost::common::MakeIndexTransformIter

(

Fn &&

fn

)

Max()

template<typename T >

XGBOOST_DEVICE T xgboost::common::Max	(	T	a,
		T	b
	)

OmpGetNumThreads()

int32_t xgboost::common::OmpGetNumThreads ( int32_t  n_threads )

inline

OmpGetThreadLimit()

int32_t xgboost::common::OmpGetThreadLimit ( )

inline

operator!=()

template<class T , std::size_t X, class U , std::size_t Y>

constexpr XGBOOST_DEVICE bool xgboost::common::operator!= ( Span< T, X >  l, Span< U, Y >  r  )

constexpr

operator<()

template<class T , std::size_t X, class U , std::size_t Y>

constexpr XGBOOST_DEVICE bool xgboost::common::operator< ( Span< T, X >  l, Span< U, Y >  r  )

constexpr

operator<=()

template<class T , std::size_t X, class U , std::size_t Y>

constexpr XGBOOST_DEVICE bool xgboost::common::operator<= ( Span< T, X >  l, Span< U, Y >  r  )

constexpr

operator==()

template<class T , std::size_t X, class U , std::size_t Y>

XGBOOST_DEVICE bool xgboost::common::operator==	(	Span< T, X >	l,
		Span< U, Y >	r
	)

operator>()

template<class T , std::size_t X, class U , std::size_t Y>

constexpr XGBOOST_DEVICE bool xgboost::common::operator> ( Span< T, X >  l, Span< U, Y >  r  )

constexpr

operator>=()

template<class T , std::size_t X, class U , std::size_t Y>

constexpr XGBOOST_DEVICE bool xgboost::common::operator>= ( Span< T, X >  l, Span< U, Y >  r  )

constexpr

OutOfRangeCat()

constexpr bst_cat_t xgboost::common::OutOfRangeCat ( )

inlineconstexpr

ParallelFor() [1/2]

template<typename Index , typename Func >

void xgboost::common::ParallelFor	(	Index	size,
		int32_t	n_threads,
		Func	fn
	)

ParallelFor() [2/2]

template<typename Index , typename Func >

void xgboost::common::ParallelFor	(	Index	size,
		int32_t	n_threads,
		Sched	sched,
		Func	fn
	)

ParallelFor2d()

template<typename Func >

void xgboost::common::ParallelFor2d	(	const BlockedSpace2d &	space,
		int	nthreads,
		Func	func
	)

ReadAll()

std::string xgboost::common::ReadAll ( dmlc::Stream *  fi, PeekableInStream *  fp  )

inline

Read the whole buffer from dmlc stream.

Sigmoid()

XGBOOST_DEVICE float xgboost::common::Sigmoid ( float  x )

inline

calculate the sigmoid of the input.

Parameters

x	input parameter

Returns

the transformed value.

SketchOnDMatrix()

HistogramCuts xgboost::common::SketchOnDMatrix ( DMatrix *  m, int32_t  max_bins, int32_t  n_threads, bool  use_sorted = false, Span< float > const  hessian = {}  )

inline

Run CPU sketching on DMatrix.

Parameters

use_sorted

Whether should we use SortedCSC for sketching, it's more efficient but consumes more memory.

Softmax()

template<typename Iterator >

XGBOOST_DEVICE void xgboost::common::Softmax ( Iterator  start, Iterator  end  )

inline

Do inplace softmax transformaton on start to end.

Template Parameters

Iterator

Input iterator type

Parameters

start	Start iterator of input
end	end iterator of input

Split()

std::vector<std::string> xgboost::common::Split ( const std::string &  s, char  delim  )

inline

Split a string by delimiter.

Parameters

s	String to be split.
delim	The delimiter.

SubtractionHist()

template<typename GradientSumT >

void xgboost::common::SubtractionHist	(	GHistRow< GradientSumT >	dst,
		const GHistRow< GradientSumT >	src1,
		const GHistRow< GradientSumT >	src2,
		size_t	begin,
		size_t	end
	)

Compute Subtraction: dst = src1 - src2 in range [begin, end)

ToString()

template<typename T >

std::string xgboost::common::ToString ( const T &  data )

inline

UnpackArr()

template<class T , std::size_t N>

constexpr auto xgboost::common::UnpackArr ( std::array< T, N > &&  arr )

constexpr

UseOneHot()

XGBOOST_DEVICE bool xgboost::common::UseOneHot ( uint32_t  n_cats, uint32_t  max_cat_to_onehot  )

inline

Whether should we use onehot encoding for categorical data.

WeightedSamplingWithoutReplacement()

template<typename T >

std::vector<T> xgboost::common::WeightedSamplingWithoutReplacement	(	std::vector< T > const &	array,
		std::vector< float > const &	weights,
		size_t	n
	)

Variable Documentation

dynamic_extent

constexpr std::size_t xgboost::common::dynamic_extent = std::numeric_limits<std::size_t>::max()

constexpr

kBlockThreads

constexpr size_t xgboost::common::kBlockThreads = 256

constexpr

kEulerMascheroni

constexpr double xgboost::common::kEulerMascheroni = 0.57721566490153286060651209008240243104215933593992

constexpr

The Euler-Mascheroni_constant.

kPI

constexpr double xgboost::common::kPI = 3.14159265358979323846

constexpr

Constant PI.