Global Functions

Global functions. More...

Data Structures
class	ResizableArray
	A data structure that provides a resizable array This is used within Spam to avoid reading input data files multiple times, without having to use any STL library classes. More...
class	StringMap
	A data structure that acts as a two-way one-to-one mapping from integers to strings Optimized for SPAM to quickly retrieve customer IDs, transaction IDs, and item IDs from their corresponding string names, and vice versa. More...
Functions
DatasetInfo *	ReadDataset (bool isBinaryFile, bool isStringFile, char *filename, double minSupPercent, StringMap *&custStrMap, StringMap *&transStrMap, StringMap *&itemStrMap)
	It reads the input file and finds the frequent-1 itemsets.
int	compare (const void *arg1, const void *arg2)
void	LogStdoutSequence (const int c)
void	LogFileSequence (const int c)
void	LogSequence (const int c)
void	CreateOrBitmap (SeqBitmap **f1, int *indexList, int indexLength, SeqBitmap *&orBitmap)
	OR's all of the frequent-1 itemset bitmaps together This is used to create the refBitmap for bitmap compression.
void	Compress (SeqBitmap *refBitmap, SeqBitmap *tempAndBitmap, SeqBitmap *&returnBitmap, SeqBitmap **f1, SeqBitmap **newF1, int *indexList, int indexLength)
	Perform compression on a sequence bitmap.
void	FindSequentialPatterns (TreeNode *curNode)
	A recursive call that goes down the search lattice to find sequential patterns.
void	StartMining (DatasetInfo *info)
	Start the mining algorithm by generating the initial TreeNode to start recursing from.
void	PrintError ()
int	main (int argc, char **argv)

---

Detailed Description

Global functions.

---

Function Documentation

int compare (  const void *  arg1, const void *  arg2 )

Definition at line 140 of file Spam.cpp. Referenced by FindSequentialPatterns(), and StartMining().

void Compress (  SeqBitmap *  refBitmap, SeqBitmap *  tempAndBitmap, SeqBitmap *&  returnBitmap, SeqBitmap **  f1, SeqBitmap **  newF1, int *  indexList, int  indexLength )

Perform compression on a sequence bitmap. Compression converts as many Bitmap64 bits as possible to Bitmap32 bits, Bitmap32 to Bitmap16, and so on, with the goal of speeding up support counting. We can only get rid of a given bit in a bitmap if it will never be used further on down the tree. The refBitmap parameter should contain a 0 for every bit that is never again used, and a 1 for a bit that is used again. Note that the frequent-1 itemset bitmaps are also compressed so that their bits are still aligned with the sequence bitmap after compression. Parameters: refBitmap  Bitmap specifying which bits are compressible tempAndBitmap  The sequence bitmap to compress returnBitmap  The compressed sequence bitmap f1  Frequent-1 itemset bitmaps prior to compression newF1  The compressed versions of the frequent-1 itemset bitmaps indexList  The list of item names for the 1 itemsets that are still frequent indexLength  The length of indexList Definition at line 481 of file Spam.cpp. Referenced by FindSequentialPatterns().

void CreateOrBitmap (  SeqBitmap **  f1, int *  indexList, int  indexLength, SeqBitmap *&  orBitmap )

OR's all of the frequent-1 itemset bitmaps together This is used to create the refBitmap for bitmap compression. Parameters: f1  The frequent-1 itemset bitmaps indexList  The list of item names for the 1 itemsets that are still frequent indexLength  The length of indexList orBitmap  [output]The OR'd bitmap. This tells us which bits can be used in operations further down this branch of the sequence tree Definition at line 449 of file Spam.cpp. Referenced by FindSequentialPatterns().

void FindSequentialPatterns (  TreeNode *  curNode  )

A recursive call that goes down the search lattice to find sequential patterns. Parameters: curNode  information about the current node Definition at line 681 of file Spam.cpp. Referenced by StartMining().

void LogFileSequence (  const int  c  )

Definition at line 297 of file Spam.cpp. Referenced by LogSequence().

void LogSequence (  const int  c  )

Definition at line 149 of file Spam.cpp. Referenced by FindSequentialPatterns(), and StartMining().

void LogStdoutSequence (  const int  c  )

Definition at line 158 of file Spam.cpp. Referenced by LogSequence().

int main (  int  argc, char **  argv )

Definition at line 1285 of file Spam.cpp.

void PrintError (   )

Definition at line 1246 of file Spam.cpp. Referenced by main().

DatasetInfo* ReadDataset (  bool  isBinaryFile, bool  isStringFile, char *  filename, double  minSupPercent, StringMap *&  custStrMap, StringMap *&  transStrMap, StringMap *&  itemStrMap )

It reads the input file and finds the frequent-1 itemsets. Parameters: isBinaryFile  whether the input file is a binary data file isStringFile  whether the input file contains integers or the string names filename  the filename of the data file minSupPercent  the minimum support percentage custStrMap  [output] Maps cust IDs to strings (only used when isStringFile == true) transStrMap  [output] Maps trans IDs to strings (only used when isStringFile == true) itemStrMap  [output] Maps item IDs to strings (only used when isStringFile == true) Returns: DatasetInfo - the information gathered from the dataset Definition at line 877 of file FileInput.cpp. Referenced by main().

void StartMining (  DatasetInfo *  info  )

Start the mining algorithm by generating the initial TreeNode to start recursing from. Parameters: info  information about the data set Definition at line 1163 of file Spam.cpp. Referenced by main().

---