The precise version of two-level interval-equality index. More...

#include <irelic.h>

Inheritance diagram for ibis::fuzz:

Public Member Functions
virtual long	append (const char dt, const char df, uint32_t nnew)
	Create an index based on data in df and append the result to the index in dt. More...

virtual uint32_t	estimate (const ibis::qContinuousRange &expr) const
	Return the number of hits satisfying the given continuous range expression. More...

virtual double	estimateCost (const ibis::qContinuousRange &expr) const
	Estimate the cost of resolving the continuous range expression. More...

virtual long	evaluate (const ibis::qContinuousRange &expr, ibis::bitvector &hits) const
	Compute the hits as a `bitvector`.

	fuzz (const ibis::column c=0, const char f=0)

	fuzz (const ibis::column c, ibis::fileManager::storage st, size_t start=8)
	Reconstruct an index from the content of a storage object. More...

virtual const char *	name () const
	Returns the name of the index, similar to the function `type`, but returns a string instead. More...

virtual void	print (std::ostream &out) const
	The printing function.

virtual int	read (const char *idxfile)
	Read an index from the specified location. More...

virtual int	read (ibis::fileManager::storage *st)
	Reconstruct an index from a storage object.

virtual INDEX_TYPE	type () const
	Returns an index type identifier.

virtual int	write (const char *dt) const
	Write the content of the index to the specified location. More...

Public Member Functions inherited from ibis::relic
long	append (const ibis::relic &tail)
	Append tail to this index. More...

long	append (const array_t< uint32_t > &ind)
	Append a list of integers. More...

virtual void	binBoundaries (std::vector< double > &b) const
	Return all distinct values as the bin boundaries.

virtual void	binWeights (std::vector< uint32_t > &b) const
	Return the exact count for each distinct value.

void	construct (const char *f=0)
	Construct a new index in memory. More...

template<typename E >
void	construct (const array_t< E > &arr)
	Construct an index from in-memory values. More...

virtual index *	dup () const
	Duplicate the content of an index object.

virtual void	estimate (const ibis::qContinuousRange &expr, ibis::bitvector &lower, ibis::bitvector &upper) const
	Computes an approximation of hits as a pair of lower and upper bounds. More...

virtual void	estimate (const ibis::qDiscreteRange &expr, ibis::bitvector &lower, ibis::bitvector &upper) const
	Estimate the hits for discrete ranges, i.e., those translated from 'a IN (x, y, ..)'. More...

virtual uint32_t	estimate (const ibis::qDiscreteRange &) const
	Compute the number of hits satisfying the discrete range expression.

virtual void	estimate (const ibis::relic &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const
	Estimate the pairs for the range join operator. More...

virtual void	estimate (const ibis::relic &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const

virtual int64_t	estimate (const ibis::relic &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask) const
	Estimate an upper bound for the number of pairs produced from marked records. More...

virtual int64_t	estimate (const ibis::relic &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2) const

virtual double	estimateCost (const ibis::qDiscreteRange &expr) const
	Estimate the cost of resolving the discrete range expression. More...

virtual long	evaluate (const ibis::qDiscreteRange &expr, ibis::bitvector &hits) const
	Resolve a discrete range condition. More...

virtual long	getCumulativeDistribution (std::vector< double > &bds, std::vector< uint32_t > &cts) const
	Compute a cumulative distribition.

virtual long	getDistribution (std::vector< double > &bds, std::vector< uint32_t > &cts) const
	Compute a histogram.

virtual double	getMax () const
	The maximum value recorded in the index.

virtual double	getMin () const
	The minimum value recorded in the index.

virtual double	getSum () const
	Compute the sum of all values of the column indexed.

array_t< uint32_t > *	keys (const ibis::bitvector &mask) const
	Convert the bitvector mask into bin numbers.

void	locate (const ibis::qContinuousRange &expr, uint32_t &hit0, uint32_t &hit1) const
	Locate the bitmaps covered by the range expression. More...

	relic (const relic &)
	Copy constructor.

	relic (const ibis::column c, const char f=0)
	Construct a basic bitmap index. More...

	relic (const ibis::column *c, uint32_t popu, uint32_t ntpl=0)
	Construct a dummy index. More...

	relic (const ibis::column *c, uint32_t card, array_t< uint32_t > &ints)
	Construct an index from an integer array. More...

	relic (const ibis::column c, ibis::fileManager::storage st, size_t start=8)
	Reconstruct from content of fileManager::storage. More...

	relic (const ibis::column c, uint32_t nb, double keys, int64_t *offs)
	Reconstruct index from keys and offsets.

	relic (const ibis::column c, uint32_t nb, double keys, int64_t offs, uint32_t bms)
	Reconstruct index from keys and offsets.

	relic (const ibis::column c, uint32_t nb, double keys, int64_t offs, void bms, FastBitReadBitmaps rd)
	Reconstruct index from keys and offsets.

virtual long	select (const ibis::qContinuousRange &, void *) const
	Evaluate the range condition and select values.

virtual long	select (const ibis::qContinuousRange &, void *, ibis::bitvector &) const
	Evaluate the range condition, select values, and record the positions.

virtual void	serialSizes (uint64_t &, uint64_t &, uint64_t &) const
	Compute the size of arrays that would be generated by the serializatioin function (write). More...

virtual void	speedTest (std::ostream &out) const
	Time some logical operations and print out their speed.

virtual float	undecidable (const ibis::qContinuousRange &, ibis::bitvector &iffy) const
	This class and its derived classes should produce exact answers, therefore no undecidable rows. More...

virtual float	undecidable (const ibis::qDiscreteRange &, ibis::bitvector &iffy) const

virtual int	write (ibis::array_t< double > &, ibis::array_t< int64_t > &, ibis::array_t< uint32_t > &) const
	Save index to three arrays. Serialize the index in memory.

Public Member Functions inherited from ibis::index
void	addBins (uint32_t ib, uint32_t ie, ibis::bitvector &res) const
	Add the sum of `bits`[ib] through `bits`[ie-1] to `res`. More...

void	addBins (uint32_t ib, uint32_t ie, ibis::bitvector &res, const ibis::bitvector &tot) const
	Compute the sum of bit vectors [`ib`, `ie`). More...

virtual int	contractRange (ibis::qContinuousRange &) const

bool	empty () const
	The index object is considered empty if there is no bitmap or getNRows returns 0. More...

virtual void	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const
	Estimate the pairs for the range join operator.

virtual void	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const
	Estimate the pairs for the range join operator. More...

virtual void	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const

virtual int64_t	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr) const
	Estimate an upper bound for the number of pairs.

virtual int64_t	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask) const
	Estimate an upper bound for the number of pairs produced from marked records. More...

virtual int64_t	estimate (const ibis::index &idx2, const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2) const

virtual void	estimate (const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2, ibis::bitvector64 &lower, ibis::bitvector64 &upper) const
	Evaluating a join condition with one (likely composite) index.

virtual int64_t	estimate (const ibis::deprecatedJoin &expr, const ibis::bitvector &mask, const ibis::qRange const range1, const ibis::qRange const range2) const

virtual int	expandRange (ibis::qContinuousRange &) const
	The functions expandRange and contractRange expands or contracts the boundaries of a range condition so that the new range will have exact answers using the function estimate. More...

virtual const ibis::bitvector *	getBitvector (uint32_t i) const
	Return a pointer to the ith bitvector used in the index (may be 0).

uint32_t	getNRows () const
	Return the number of rows represented by this object.

virtual uint32_t	numBitvectors () const
	Returns the number of bit vectors used by the index.

float	sizeInBytes () const
	Estiamte the size of this index object measured in bytes. More...

void	sumBins (uint32_t ib, uint32_t ie, ibis::bitvector &res) const
	Sum up `bits[ib:ie-1]` and place the result in res. More...

void	sumBins (uint32_t ib, uint32_t ie, ibis::bitvector &res, uint32_t ib0, uint32_t ie0) const
	Compute a new sum for bit vectors [ib, ie) by taking advantage of the old sum for bitvectors [ib0, ie0). More...

void	sumBins (uint32_t ib, uint32_t ie, ibis::bitvector &res, uint32_t *buf) const
	Sum up `bits[ib:ie-1]` and place the result in res. More...

void	sumBins (const ibis::array_t< uint32_t > &, ibis::bitvector &) const
	Sum up the bits in in the specified bins.

virtual	~index ()
	The destructor.

Protected Member Functions
virtual void	clear ()
	Clear the existing content. More...

virtual size_t	getSerialSize () const throw ()
	Estiamte the size of the index in a file.

Protected Member Functions inherited from ibis::relic
virtual double	computeSum () const
	Compute the sum of all values of the column indexed.

uint32_t	locate (const double &val) const
	Find the smallest i such that vals[i] > val.

long	mergeValues (uint32_t, uint32_t, void *) const
	Merge the values in different bitmaps into a single list. More...

int	write32 (int fdes) const
	Write the content to a file already opened.

int	write64 (int fdes) const
	Write the content to a file already opened.

Protected Member Functions inherited from ibis::index
virtual void	activate () const
	Regenerate all bitvectors from the underlying storage. More...

virtual void	activate (uint32_t i) const
	Regenerate the ith bitvector from the underlying storage.

virtual void	activate (uint32_t i, uint32_t j) const
	Regenerate bitvectors i (inclusive) through j (exclusive) from the underlying storage. More...

void	computeMinMax (const char *f, double &min, double &max) const

void	dataFileName (std::string &name, const char *f=0) const
	Generate data file name from "f". More...

	index (const ibis::column *c=0)
	Default constructor. More...

	index (const ibis::column c, ibis::fileManager::storage s)
	Constructor with a storage object. More...

	index (const index &)
	Copy constructor.

void	indexFileName (std::string &name, const char *f=0) const
	Generates index file name from "f". More...

void	initBitmaps (int fdes)
	Prepare the bitmaps using the given file descriptor. More...

void	initBitmaps (ibis::fileManager::storage *st)
	Prepare bitmaps from the given storage object. More...

void	initBitmaps (uint32_t *st)
	Prepare bitmaps from the given raw pointer. More...

void	initBitmaps (void *ctx, FastBitReadBitmaps rd)
	Prepare bitmaps from the user provided function pointer and context. More...

int	initOffsets (int64_t *, size_t)
	Initialize the offsets from the given data array. More...

int	initOffsets (int fdes, const char offsize, size_t start, uint32_t nobs)
	Read in the offset array. More...

int	initOffsets (ibis::fileManager::storage *st, size_t start, uint32_t nobs)
	Regenerate the offsets array from the given storage object. More...

void	mapValues (const char *f, VMap &bmap) const
	Map the positions of each individual value. More...

void	mapValues (const char *f, histogram &hist, uint32_t count=0) const
	Generate a histogram. More...

index &	operator= (const index &)
	Assignment operator.

void	optionalUnpack (array_t< ibis::bitvector * > &bits, const char *opt)
	A function to decide whether to uncompress the bitvectors. More...

Additional Inherited Members
Public Types inherited from ibis::index
typedef std::map< double, uint32_t >	histogram

enum	INDEX_TYPE { BINNING =0, RANGE, MESA, AMBIT, PALE, PACK, ZONE, RELIC, ROSTER, SKIVE, FADE, SBIAD, SAPID, EGALE, MOINS, ENTRE, BAK, BAK2, KEYWORDS, MESH, BAND, DIREKTE, GENERIC, BYLT, FUZZ, ZONA, FUGE, SLICE, EXTERN }
	The integer values of this enum type are used in the index files to differentiate the indexes. More...

typedef std::map< double, ibis::bitvector * >	VMap

Static Public Member Functions inherited from ibis::index
static void	addBits (const array_t< bitvector * > &bits, uint32_t ib, uint32_t ie, ibis::bitvector &res)
	Add the `pile`[ib:ie-1] to `res`. More...

static index *	create (const column c, const char name=0, const char *spec=0, int inEntirety=0)
	Index factory. More...

static void	divideCounts (array_t< uint32_t > &bounds, const array_t< uint32_t > &cnt)
	Determine how to split the array `cnt`, so that each group has roughly the same total value. More...

static bool	isIndex (const char *f, INDEX_TYPE t)
	Is the named file an index file? Read the header of the named file to determine if it contains an index of the specified type. More...

template<typename E >
static void	mapValues (const array_t< E > &val, VMap &bmap)

template<typename E >
static void	mapValues (const array_t< E > &val, histogram &hist, uint32_t count=0)

template<typename E >
static void	mapValues (const array_t< E > &val, array_t< E > &bounds, std::vector< uint32_t > &cnts)

template<typename E1 , typename E2 >
static void	mapValues (const array_t< E1 > &val1, const array_t< E2 > &val2, array_t< E1 > &bnd1, array_t< E2 > &bnd2, std::vector< uint32_t > &cnts)
	Compute a two-dimensional histogram. More...

static void	printHeader (std::ostream &, const char *)

static void	setBases (array_t< uint32_t > &bases, uint32_t card, uint32_t nbase=2)
	Fill the array bases with the values that cover the range [0, card). More...

static void	sumBits (const array_t< bitvector * > &bits, uint32_t ib, uint32_t ie, ibis::bitvector &res)
	Sum up `pile`[ib:ie-1] and place the result in `res`. More...

static void	sumBits (const array_t< bitvector * > &bits, const ibis::bitvector &tot, uint32_t ib, uint32_t ie, ibis::bitvector &res)
	Sum up `pile`[ib:ie-1] and add the result to `res`. More...

Static Protected Member Functions inherited from ibis::relic
template<typename T >
static long	mergeValuesT (const array_t< T > &vs, const array_t< const bitvector * > &ps, array_t< T > &res)
	A template function to merge list of values and a list of positions. More...

Static Protected Member Functions inherited from ibis::index
static void	indexFileName (std::string &name, const ibis::column col1, const ibis::column col2, const char *f=0)
	Generate the index file name for the composite index fromed on two columns. More...

Protected Attributes inherited from ibis::relic
array_t< double >	vals

Protected Attributes inherited from ibis::index
array_t< ibis::bitvector * >	bits
	A list of bitvectors.

bitmapReader *	breader
	The functor to read serialized bitmaps from a more complex source.

const ibis::column *	col
	Pointer to the column this index is for.

const char *	fname
	The name of the file containing the index.

uint32_t	nrows
	The number of rows represented by the index. More...

array_t< int32_t >	offset32
	Starting positions of the bitvectors.

array_t< int64_t >	offset64
	Starting positions of the bitvectors. More...

ibis::fileManager::storage *	str
	The underlying storage. More...

Detailed Description

The precise version of two-level interval-equality index.

Note: In fuzzy classification / clustering, many interval equality conditions are used. Hence the crazy name.

Constructor & Destructor Documentation

ibis::fuzz::fuzz	(	const ibis::column *	c,
		ibis::fileManager::storage *	st,
		size_t	start = `8`
	)

Reconstruct an index from the content of a storage object.

The leading portion of the index file is the same as ibis::relic, which allows the constructor of the base class to work properly. The content following the last bitvector in ibis::relic is as follows,

See also: ibis::fuzz::writeCoarse.
nc (uint32_t) -- number of coarse bins.

cbounds (uint32_t[nc+1]) -- boundaries of the coarse bins.

coffsets([nc-ceil(nc/2)+2]) -- starting positions.

cbits (bitvector[nc-ceil(nc/2)+1]) -- bitvectors.

References ibis::fileManager::storage::begin(), ibis::index::bits, ibis::array_t< T >::clear(), clear(), ibis::index::col, ibis::part::name(), ibis::column::name(), ibis::index::nrows, ibis::index::offset32, ibis::index::offset64, print(), ibis::array_t< T >::size(), ibis::fileManager::storage::size(), and ibis::array_t< T >::swap().

Member Function Documentation

long ibis::fuzz::append	(	const char *	dt,
		const char *	df,
		uint32_t	nnew
	)

virtual

Create an index based on data in df and append the result to the index in dt.

Reimplemented from ibis::relic.

References ibis::relic::append(), and ibis::util::coarsen().

void ibis::fuzz::clear ( )

protectedvirtual

Clear the existing content.

Free the bitmap objectes common to all index objects.

Reimplemented from ibis::relic.

References ibis::relic::clear().

Referenced by fuzz().

uint32_t ibis::fuzz::estimate ( const ibis::qContinuousRange & expr ) const

virtual

Return the number of hits satisfying the given continuous range expression.

Reimplemented from ibis::relic.

double ibis::fuzz::estimateCost ( const ibis::qContinuousRange & expr ) const

virtual

Estimate the cost of resolving the continuous range expression.

The answer is in the number of bytes needed from this index.

Reimplemented from ibis::relic.

virtual const char* ibis::fuzz::name ( ) const

inlinevirtual

Returns the name of the index, similar to the function type, but returns a string instead.

Reimplemented from ibis::relic.

int ibis::fuzz::read ( const char * f )

virtual

Read an index from the specified location.

The incoming argument can be the name of a directory or a file. The actual index file name is determined by the function indexFileName.

Reimplemented from ibis::relic.

References ibis::util::clear(), ibis::fileManager::instance(), ibis::fileManager::recordPages(), ibis::util::strnewdup(), and UnixOpen.

int ibis::fuzz::write ( const char * dt ) const

virtual

Write the content of the index to the specified location.

The incoming argument can be the name of a directory or a file. The actual index file name is determined by the function indexFileName.

Reimplemented from ibis::relic.

References ibis::fileManager::flushFile(), ibis::index::FUZZ, ibis::fileManager::instance(), ibis::util::flock::isLocked(), ibis::index::RELIC, UnixOpen, ibis::relic::write32(), and ibis::relic::write64().

The documentation for this class was generated from the following files:

irelic.h
ixfuzz.cpp



	Contact us Disclaimers FastBit source code FastBit mailing list archive

Public Member Functions

Protected Member Functions

Additional Inherited Members

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation