apriori-rails 0.2.4.0
Sign up to get free protection for your applications and to get access to all the features.
- data/History.txt +22 -0
- data/License.txt +20 -0
- data/Manifest.txt +121 -0
- data/README.txt +149 -0
- data/Rakefile +17 -0
- data/TODO.txt +60 -0
- data/attic/c_ext_test1/MyTest/MyTest.c +23 -0
- data/attic/c_ext_test1/MyTest/extconf.rb +11 -0
- data/attic/c_ext_test1/mytest.rb +10 -0
- data/attic/test.c +12 -0
- data/config/hoe.rb +88 -0
- data/config/requirements.rb +29 -0
- data/examples/01_simple_example.rb +39 -0
- data/examples/02_small_file_example.rb +17 -0
- data/examples/03_large_file_example.rb +22 -0
- data/examples/test_data/market_basket_basic_test.dat +9 -0
- data/ext/Apriori.c +149 -0
- data/ext/Makefile +149 -0
- data/ext/apriori/doc/apriori.html +1301 -0
- data/ext/apriori/doc/arem.gp +68 -0
- data/ext/apriori/doc/c_rev.gp +89 -0
- data/ext/apriori/doc/chi2.tex +156 -0
- data/ext/apriori/doc/copying +504 -0
- data/ext/apriori/doc/line.gif +0 -0
- data/ext/apriori/doc/uparrow.gif +0 -0
- data/ext/apriori/ex/flg2set +15 -0
- data/ext/apriori/ex/hdr2set +13 -0
- data/ext/apriori/ex/readme +71 -0
- data/ext/apriori/ex/row2set +7 -0
- data/ext/apriori/ex/rulesort +24 -0
- data/ext/apriori/ex/tab2set +9 -0
- data/ext/apriori/ex/test.app +2 -0
- data/ext/apriori/ex/test.rul +9 -0
- data/ext/apriori/ex/test1.rul +43 -0
- data/ext/apriori/ex/test1.tab +10 -0
- data/ext/apriori/ex/test2.tab +10 -0
- data/ext/apriori/ex/test3.tab +30 -0
- data/ext/apriori/ex/test4.tab +11 -0
- data/ext/apriori/ex/test5.tab +39 -0
- data/ext/apriori/ex/tid2set +23 -0
- data/ext/apriori/ex/xhdr2set +33 -0
- data/ext/apriori/src/apriori.c +750 -0
- data/ext/apriori/src/apriori.dsp +120 -0
- data/ext/apriori/src/apriori.dsw +29 -0
- data/ext/apriori/src/apriori.mak +99 -0
- data/ext/apriori/src/istree.c +1411 -0
- data/ext/apriori/src/istree.h +160 -0
- data/ext/apriori/src/makefile +105 -0
- data/ext/apriori/src/tract.c +870 -0
- data/ext/apriori/src/tract.h +261 -0
- data/ext/apriori_wrapper.c +757 -0
- data/ext/apriori_wrapper.h +10 -0
- data/ext/extconf.rb +32 -0
- data/ext/math/doc/copying +504 -0
- data/ext/math/src/chi2.c +151 -0
- data/ext/math/src/chi2.h +27 -0
- data/ext/math/src/choose.c +71 -0
- data/ext/math/src/choose.h +16 -0
- data/ext/math/src/gamma.c +446 -0
- data/ext/math/src/gamma.h +39 -0
- data/ext/math/src/intexp.c +35 -0
- data/ext/math/src/intexp.h +15 -0
- data/ext/math/src/makefile +164 -0
- data/ext/math/src/math.mak +48 -0
- data/ext/math/src/normal.c +387 -0
- data/ext/math/src/normal.h +44 -0
- data/ext/math/src/radfn.c +113 -0
- data/ext/math/src/radfn.h +34 -0
- data/ext/math/src/zeta.c +49 -0
- data/ext/math/src/zeta.h +15 -0
- data/ext/pre-clean.rb +8 -0
- data/ext/pre-setup.rb +9 -0
- data/ext/util/doc/copying +504 -0
- data/ext/util/src/listops.c +76 -0
- data/ext/util/src/listops.h +26 -0
- data/ext/util/src/makefile +103 -0
- data/ext/util/src/memsys.c +84 -0
- data/ext/util/src/memsys.h +42 -0
- data/ext/util/src/nstats.c +288 -0
- data/ext/util/src/nstats.h +69 -0
- data/ext/util/src/params.c +86 -0
- data/ext/util/src/params.h +19 -0
- data/ext/util/src/parse.c +133 -0
- data/ext/util/src/parse.h +81 -0
- data/ext/util/src/scan.c +767 -0
- data/ext/util/src/scan.h +111 -0
- data/ext/util/src/symtab.c +443 -0
- data/ext/util/src/symtab.h +121 -0
- data/ext/util/src/tabscan.c +279 -0
- data/ext/util/src/tabscan.h +99 -0
- data/ext/util/src/util.mak +91 -0
- data/ext/util/src/vecops.c +317 -0
- data/ext/util/src/vecops.h +42 -0
- data/lib/apriori.rb +133 -0
- data/lib/apriori/adapter.rb +13 -0
- data/lib/apriori/association_rule.rb +97 -0
- data/lib/apriori/version.rb +3 -0
- data/script/console +10 -0
- data/script/destroy +14 -0
- data/script/generate +14 -0
- data/script/txt2html +82 -0
- data/setup.rb +1585 -0
- data/tasks/apriori.rake +20 -0
- data/tasks/attic.rake +28 -0
- data/tasks/deployment.rake +34 -0
- data/tasks/environment.rake +7 -0
- data/tasks/install.rake +13 -0
- data/tasks/website.rake +17 -0
- data/test/apriori_test.rb +13 -0
- data/test/fixtures/market_basket_results_test.txt +5 -0
- data/test/fixtures/market_basket_string_test.txt +7 -0
- data/test/fixtures/results.txt +2 -0
- data/test/fixtures/sample.txt +7 -0
- data/test/test_helper.rb +5 -0
- data/test/unit/test_apriori.rb +68 -0
- data/test/unit/test_itemsets_and_parsing.rb +82 -0
- data/website/index.html +251 -0
- data/website/index.txt +154 -0
- data/website/javascripts/rounded_corners_lite.inc.js +285 -0
- data/website/stylesheets/screen.css +142 -0
- data/website/template.html.erb +49 -0
- metadata +267 -0
@@ -0,0 +1,160 @@
|
|
1
|
+
/*----------------------------------------------------------------------
|
2
|
+
File : istree.h
|
3
|
+
Contents: item set tree management
|
4
|
+
Author : Christian Borgelt
|
5
|
+
History : 1996.01.22 file created
|
6
|
+
1996.01.29 ISNODE.offset and ISNODE.id added
|
7
|
+
1996.02.08 ISTREE.tacnt, ISTREE.curr, ISTREE.index,
|
8
|
+
ISTREE.head and ISTREE.conf added
|
9
|
+
1996.03.28 support made relative to number of item sets
|
10
|
+
1996.11.23 ISTREE.levels (first nodes of each level) added
|
11
|
+
1996.11.24 ISTREE.arem (add. rule evaluation measure) added
|
12
|
+
1997.08.18 chi^2 evaluation measure added
|
13
|
+
parameter 'minlen' added to function ist_init()
|
14
|
+
1998.02.11 parameter 'minval' added to function ist_init()
|
15
|
+
1998.05.14 item set tree navigation functions added
|
16
|
+
1998.08.08 parameter 'apps' added to function ist_create()
|
17
|
+
1998.08.20 structure ISNODE redesigned
|
18
|
+
1998.09.07 function ist_hedge added
|
19
|
+
1998.12.08 function ist_gettac added,
|
20
|
+
float changed to double
|
21
|
+
1999.02.05 long int changed to int
|
22
|
+
1999.08.26 functions ist_first and ist_last added
|
23
|
+
1999.11.05 rule evaluation measure EM_AIMP added
|
24
|
+
1999.11.08 parameter 'aval' added to function ist_rule
|
25
|
+
2001.04.01 functions ist_set and ist_getcntx added
|
26
|
+
2001.12.28 sort function moved to module tract
|
27
|
+
2002.02.07 function ist_clear removed, ist_settac added
|
28
|
+
2002.02.11 optional use of identifier maps in nodes added
|
29
|
+
2002.02.12 ist_first and ist_last replaced by ist_next
|
30
|
+
2003.03.12 parameter lift added to function ist_rule
|
31
|
+
2003.07.17 functions ist_itemcnt and ist_check added
|
32
|
+
2003.07.18 function ist_maxfrq added (item set filter)
|
33
|
+
2003.08.11 item set filtering generalized (ist_filter)
|
34
|
+
2004.05.09 parameter 'aval' added to function ist_set
|
35
|
+
2008.03.24 creation based on ITEMSET structure
|
36
|
+
----------------------------------------------------------------------*/
|
37
|
+
#ifndef __ISTREE__
|
38
|
+
#define __ISTREE__
|
39
|
+
#include "tract.h"
|
40
|
+
|
41
|
+
/*----------------------------------------------------------------------
|
42
|
+
Preprocessor Definitions
|
43
|
+
----------------------------------------------------------------------*/
|
44
|
+
/* --- additional evaluation measures --- */
|
45
|
+
#define EM_NONE 0 /* no measure */
|
46
|
+
#define EM_DIFF 1 /* absolute conf. difference to prior */
|
47
|
+
#define EM_QUOT 2 /* difference of conf. quotient to 1 */
|
48
|
+
#define EM_AIMP 3 /* abs. diff. of improvement to 1 */
|
49
|
+
#define EM_INFO 4 /* information difference to prior */
|
50
|
+
#define EM_CHI2 5 /* normalized chi^2 measure */
|
51
|
+
#define EM_PVAL 6 /* p-value of chi^2 measure */
|
52
|
+
#define EM_UNKNOWN 7 /* unknown measure */
|
53
|
+
|
54
|
+
/* --- item appearances --- */
|
55
|
+
#define IST_IGNORE 0 /* ignore item */
|
56
|
+
#define IST_BODY 1 /* item may appear in rule body */
|
57
|
+
#define IST_HEAD 2 /* item may appear in rule head */
|
58
|
+
#define IST_BOTH (IST_HEAD|IST_BODY)
|
59
|
+
|
60
|
+
/* --- search mode flags --- */
|
61
|
+
#define IST_MEMOPT 4 /* optimize memory usage */
|
62
|
+
|
63
|
+
/* --- item set filter modes --- */
|
64
|
+
#define IST_CLEAR 0 /* clear markers */
|
65
|
+
#define IST_CLOSED 1 /* closed item sets */
|
66
|
+
#define IST_MAXFRQ 2 /* maximal item sets */
|
67
|
+
|
68
|
+
/*----------------------------------------------------------------------
|
69
|
+
Type Definitions
|
70
|
+
----------------------------------------------------------------------*/
|
71
|
+
typedef struct _isnode { /* --- item set node --- */
|
72
|
+
struct _isnode *parent; /* parent node */
|
73
|
+
struct _isnode *succ; /* successor node on same level */
|
74
|
+
int id; /* identifier used in parent node */
|
75
|
+
int chcnt; /* number of child nodes */
|
76
|
+
int size; /* size of counter vector */
|
77
|
+
int offset; /* offset of counter vector */
|
78
|
+
int cnts[1]; /* counter vector */
|
79
|
+
} ISNODE; /* (item set node) */
|
80
|
+
|
81
|
+
typedef struct { /* --- item set tree --- */
|
82
|
+
ITEMSET *set; /* underlying item set */
|
83
|
+
int mode; /* search mode (e.g. support def.) */
|
84
|
+
int tacnt; /* number of transactions */
|
85
|
+
int vsz; /* size of level vector */
|
86
|
+
int height; /* tree height (number of levels) */
|
87
|
+
ISNODE **lvls; /* first node of each level */
|
88
|
+
int rule; /* minimal support of an assoc. rule */
|
89
|
+
int supp; /* minimal support of an item set */
|
90
|
+
double conf; /* minimal confidence of a rule */
|
91
|
+
int arem; /* additional rule evaluation measure */
|
92
|
+
double minval; /* minimal evaluation measure value */
|
93
|
+
ISNODE *curr; /* current node for traversal */
|
94
|
+
int size; /* size of item set/rule/hyperedge */
|
95
|
+
ISNODE *node; /* item set node for extraction */
|
96
|
+
int index; /* index in item set node */
|
97
|
+
ISNODE *head; /* head item node for extraction */
|
98
|
+
int item; /* head item of previous rule */
|
99
|
+
int *buf; /* buffer for paths (support check) */
|
100
|
+
int *path; /* current path / (partial) item set */
|
101
|
+
int plen; /* current path length */
|
102
|
+
int hdonly; /* head only item in current set */
|
103
|
+
int *map; /* to create identifier maps */
|
104
|
+
#ifdef BENCH /* if benchmark version */
|
105
|
+
int sccnt; /* number of support counters */
|
106
|
+
int scnec; /* number of necessary supp. counters */
|
107
|
+
int cpcnt; /* number of child pointers */
|
108
|
+
int cpnec; /* number of necessary child pointers */
|
109
|
+
int bytes; /* number of bytes used */
|
110
|
+
#endif
|
111
|
+
} ISTREE; /* (item set tree) */
|
112
|
+
|
113
|
+
/*----------------------------------------------------------------------
|
114
|
+
Functions
|
115
|
+
----------------------------------------------------------------------*/
|
116
|
+
extern ISTREE* ist_create (ITEMSET *set, int mode,
|
117
|
+
int supp, double conf);
|
118
|
+
extern void ist_delete (ISTREE *ist);
|
119
|
+
extern int ist_itemcnt (ISTREE *ist);
|
120
|
+
|
121
|
+
extern void ist_count (ISTREE *ist, int *set, int cnt);
|
122
|
+
extern void ist_countx (ISTREE *ist, TATREE *tat);
|
123
|
+
extern int ist_settac (ISTREE *ist, int cnt);
|
124
|
+
extern int ist_gettac (ISTREE *ist);
|
125
|
+
extern int ist_check (ISTREE *ist, char *marks);
|
126
|
+
extern int ist_addlvl (ISTREE *ist);
|
127
|
+
extern int ist_height (ISTREE *ist);
|
128
|
+
|
129
|
+
extern void ist_up (ISTREE *ist, int root);
|
130
|
+
extern int ist_down (ISTREE *ist, int item);
|
131
|
+
extern int ist_next (ISTREE *ist, int item);
|
132
|
+
extern void ist_setcnt (ISTREE *ist, int item, int cnt);
|
133
|
+
extern int ist_getcnt (ISTREE *ist, int item);
|
134
|
+
extern int ist_getcntx (ISTREE *ist, int *set, int cnt);
|
135
|
+
|
136
|
+
extern void ist_filter (ISTREE *ist, int mode);
|
137
|
+
extern void ist_init (ISTREE *ist, int minlen,
|
138
|
+
int arem, double minval);
|
139
|
+
extern int ist_set (ISTREE *ist, int *set, int *supp,
|
140
|
+
double *aval);
|
141
|
+
extern int ist_rule (ISTREE *ist, int *rule, int *supp,
|
142
|
+
double *conf, double *lift, double *aval);
|
143
|
+
extern int ist_hedge (ISTREE *ist, int *hedge, int *supp,
|
144
|
+
double *conf, double *aval);
|
145
|
+
extern int ist_group (ISTREE *ist, int *asmb, int *supp,
|
146
|
+
double *aval);
|
147
|
+
|
148
|
+
#ifndef NDEBUG
|
149
|
+
extern void ist_show (ISTREE *ist);
|
150
|
+
#endif
|
151
|
+
|
152
|
+
/*----------------------------------------------------------------------
|
153
|
+
Preprocessor Definitions
|
154
|
+
----------------------------------------------------------------------*/
|
155
|
+
#define ist_itemcnt(t) ((t)->levels[0]->size)
|
156
|
+
#define ist_settac(t,n) ((t)->tacnt = (n))
|
157
|
+
#define ist_gettac(t) ((t)->tacnt)
|
158
|
+
#define ist_height(t) ((t)->height)
|
159
|
+
|
160
|
+
#endif
|
@@ -0,0 +1,105 @@
|
|
1
|
+
#-----------------------------------------------------------------------
|
2
|
+
# File : makefile
|
3
|
+
# Contents: build apriori program
|
4
|
+
# Author : Christian Borgelt
|
5
|
+
# History : ??.??.1995 file created
|
6
|
+
# 1997.10.13 macro ADDFLAGS added
|
7
|
+
# 1997.12.07 minor improvements
|
8
|
+
# 1998.01.04 table scanner management added
|
9
|
+
# 1999.11.11 vector operations module added
|
10
|
+
# 2000.11.04 modules vecops, symtab, and tabscan made external
|
11
|
+
# 2001.11.18 module tract (transaction management) added
|
12
|
+
# 2003.12.12 preprocessor definition ARCH64 added
|
13
|
+
#-----------------------------------------------------------------------
|
14
|
+
CC = gcc
|
15
|
+
CFBASE = -ansi -Wall -pedantic -I$(UTILDIR) -I$(MATHDIR) $(ADDFLAGS)
|
16
|
+
CFLAGS = $(CFBASE) -DNDEBUG -O3
|
17
|
+
# CFLAGS = $(CFBASE) -DNDEBUG -O3 -DBENCH
|
18
|
+
# CFLAGS = $(CFBASE) -DNDEBUG -O3 -DARCH64
|
19
|
+
# CFLAGS = $(CFBASE) -g
|
20
|
+
# CFLAGS = $(CFBASE) -g -DARCH64
|
21
|
+
# CFLAGS = $(CFBASE) -g -DSTORAGE $(ADDINC)
|
22
|
+
LDFLAGS =
|
23
|
+
LIBS = -lm
|
24
|
+
# ADDINC = -I../../misc/src
|
25
|
+
# ADDOBJ = storage.o
|
26
|
+
|
27
|
+
UTILDIR = ../../util/src
|
28
|
+
MATHDIR = ../../math/src
|
29
|
+
HDRS = $(UTILDIR)/vecops.h $(UTILDIR)/symtab.h \
|
30
|
+
$(UTILDIR)/tabscan.h $(UTILDIR)/scan.h \
|
31
|
+
$(MATHDIR)/gamma.h $(MATHDIR)/chi2.h \
|
32
|
+
tract.h istree.h
|
33
|
+
OBJS = $(UTILDIR)/vecops.o $(UTILDIR)/nimap.o \
|
34
|
+
$(UTILDIR)/tabscan.o $(UTILDIR)/scform.o \
|
35
|
+
$(MATHDIR)/gamma.o $(MATHDIR)/chi2.o \
|
36
|
+
tract.o istree.o apriori.o $(ADDOBJ)
|
37
|
+
|
38
|
+
#-----------------------------------------------------------------------
|
39
|
+
# Build Program
|
40
|
+
#-----------------------------------------------------------------------
|
41
|
+
all: apriori
|
42
|
+
|
43
|
+
apriori: $(OBJS) makefile
|
44
|
+
$(CC) $(LDFLAGS) $(OBJS) $(LIBS) -o $@
|
45
|
+
|
46
|
+
#-----------------------------------------------------------------------
|
47
|
+
# Main Program
|
48
|
+
#-----------------------------------------------------------------------
|
49
|
+
apriori.o: tract.h istree.h $(UTILDIR)/symtab.h
|
50
|
+
apriori.o: apriori.c makefile
|
51
|
+
$(CC) $(CFLAGS) -c apriori.c -o $@
|
52
|
+
|
53
|
+
#-----------------------------------------------------------------------
|
54
|
+
# Item and Transaction Management
|
55
|
+
#-----------------------------------------------------------------------
|
56
|
+
tract.o: tract.h $(UTILDIR)/symtab.h
|
57
|
+
tract.o: tract.c makefile
|
58
|
+
$(CC) $(CFLAGS) -c tract.c -o $@
|
59
|
+
|
60
|
+
#-----------------------------------------------------------------------
|
61
|
+
# Frequent Item Set Tree Management
|
62
|
+
#-----------------------------------------------------------------------
|
63
|
+
istree.o: istree.h tract.h $(MATHDIR)/gamma.h
|
64
|
+
istree.o: istree.c makefile
|
65
|
+
$(CC) $(CFLAGS) -c istree.c -o $@
|
66
|
+
|
67
|
+
#-----------------------------------------------------------------------
|
68
|
+
# External Modules
|
69
|
+
#-----------------------------------------------------------------------
|
70
|
+
$(UTILDIR)/vecops.o:
|
71
|
+
cd $(UTILDIR); $(MAKE) vecops.o ADDFLAGS=$(ADDFLAGS)
|
72
|
+
$(UTILDIR)/nimap.o:
|
73
|
+
cd $(UTILDIR); $(MAKE) nimap.o ADDFLAGS=$(ADDFLAGS)
|
74
|
+
$(UTILDIR)/tabscan.o:
|
75
|
+
cd $(UTILDIR); $(MAKE) tabscan.o ADDFLAGS=$(ADDFLAGS)
|
76
|
+
$(UTILDIR)/scform.o:
|
77
|
+
cd $(UTILDIR); $(MAKE) scform.o ADDFLAGS=$(ADDFLAGS)
|
78
|
+
$(MATHDIR)/gamma.o:
|
79
|
+
cd $(MATHDIR); $(MAKE) gamma.o ADDFLAGS=$(ADDFLAGS)
|
80
|
+
$(MATHDIR)/chi2.o:
|
81
|
+
cd $(MATHDIR); $(MAKE) chi2.o ADDFLAGS=$(ADDFLAGS)
|
82
|
+
|
83
|
+
#-----------------------------------------------------------------------
|
84
|
+
# Storage Debugging
|
85
|
+
#-----------------------------------------------------------------------
|
86
|
+
storage.o: ../../misc/src/storage.h
|
87
|
+
storage.o: ../../misc/src/storage.c
|
88
|
+
$(CC) $(CFLAGS) -c ../../misc/src/storage.c -o $@
|
89
|
+
|
90
|
+
#-----------------------------------------------------------------------
|
91
|
+
# Install
|
92
|
+
#-----------------------------------------------------------------------
|
93
|
+
install:
|
94
|
+
cp apriori $(HOME)/bin
|
95
|
+
|
96
|
+
#-----------------------------------------------------------------------
|
97
|
+
# Clean up
|
98
|
+
#-----------------------------------------------------------------------
|
99
|
+
clean:
|
100
|
+
rm -f *.o *~ *.flc core apriori
|
101
|
+
cd $(UTILDIR); $(MAKE) clean
|
102
|
+
cd $(MATHDIR); $(MAKE) clean
|
103
|
+
|
104
|
+
localclean:
|
105
|
+
rm -f *.o *~ *.flc core apriori
|
@@ -0,0 +1,870 @@
|
|
1
|
+
/*----------------------------------------------------------------------
|
2
|
+
File : tract.c
|
3
|
+
Contents: item and transaction management
|
4
|
+
Author : Christian Borgelt
|
5
|
+
History : 1996.02.14 file created as apriori.c
|
6
|
+
1996.06.24 function _get_item optimized
|
7
|
+
1996.07.01 adapted to modified symtab module
|
8
|
+
1998.01.04 scan functions moved to module 'tabscan'
|
9
|
+
1998.06.09 vector enlargement modified
|
10
|
+
1998.06.20 adapted to changed st_create function
|
11
|
+
1998.08.07 bug in function _get_tract (is_read) fixed
|
12
|
+
1998.08.08 item appearances added
|
13
|
+
1998.08.17 item sorting and recoding added
|
14
|
+
1998.09.02 several assertions added
|
15
|
+
1999.02.05 long int changed to int
|
16
|
+
1999.10.22 bug in item appearances reading fixed
|
17
|
+
1999.11.11 adapted to name/identifier maps
|
18
|
+
1999.12.01 check of item appearance added to sort function
|
19
|
+
2000.03.15 removal of infrequent items added
|
20
|
+
2001.07.14 adapted to modified module tabscan
|
21
|
+
2001.12.27 item functions made a separate module
|
22
|
+
2001.11.18 transaction functions made a separate module
|
23
|
+
2001.12.28 first version of this module completed
|
24
|
+
2002.01.12 empty field at end of record reported as error
|
25
|
+
2002.02.06 item sorting reversed (ascending order)
|
26
|
+
2002.02.19 transaction tree functions added
|
27
|
+
2003.07.17 functions is_filter, ta_filter, tas_filter added
|
28
|
+
2003.08.15 bug in function tat_delete fixed
|
29
|
+
2003.08.21 parameter 'heap' added to tas_sort, tat_create
|
30
|
+
2003.09.20 empty transactions in input made possible
|
31
|
+
2003.12.18 padding for 64 bit architecture added
|
32
|
+
2004.02.26 item frequency counting moved to is_read
|
33
|
+
2004.11.20 function tat_mark added
|
34
|
+
2005.06.20 function _nocmp added for neutral sorting
|
35
|
+
2006.11.26 structures ISFMTR and ISEVAL added
|
36
|
+
2007.02.13 adapted to modified tabscan module
|
37
|
+
2008.01.25 bug in function ise_eval fixed (prefix)
|
38
|
+
2008.06.30 support argument to ise_eval changed to double
|
39
|
+
----------------------------------------------------------------------*/
|
40
|
+
#include <stdio.h>
|
41
|
+
#include <stdlib.h>
|
42
|
+
#include <string.h>
|
43
|
+
#include <limits.h>
|
44
|
+
#include <assert.h>
|
45
|
+
#include <math.h>
|
46
|
+
#include "tract.h"
|
47
|
+
#include "scan.h"
|
48
|
+
#ifdef STORAGE
|
49
|
+
#include "storage.h"
|
50
|
+
#endif
|
51
|
+
|
52
|
+
/*----------------------------------------------------------------------
|
53
|
+
Preprocessor Definitions
|
54
|
+
----------------------------------------------------------------------*/
|
55
|
+
#define BLKSIZE 256 /* block size for enlarging vectors */
|
56
|
+
|
57
|
+
#define LN_2 0.69314718055994530942 /* ln(2) */
|
58
|
+
|
59
|
+
/*----------------------------------------------------------------------
|
60
|
+
Constants
|
61
|
+
----------------------------------------------------------------------*/
|
62
|
+
/* --- item appearance indicators --- */
|
63
|
+
static const char *i_body[] = { /* item to appear in bodies only */
|
64
|
+
"i", "in", "a", "ante", "antecedent", "b", "body", NULL };
|
65
|
+
static const char *i_head[] = { /* item to appear in heads only */
|
66
|
+
"o", "out", "c", "cons", "consequent", "h", "head", NULL };
|
67
|
+
static const char *i_both[] = { /* item to appear in both */
|
68
|
+
"io", "inout", "ac", "bh", "both", NULL };
|
69
|
+
static const char *i_ignore[] ={/* item to ignore */
|
70
|
+
"n", "neither", "none", "ign", "ignore", "-", NULL };
|
71
|
+
|
72
|
+
/*----------------------------------------------------------------------
|
73
|
+
Auxiliary Functions
|
74
|
+
----------------------------------------------------------------------*/
|
75
|
+
|
76
|
+
static int _appcode (const char *s)
|
77
|
+
{ /* --- get appearance indicator code */
|
78
|
+
const char **p; /* to traverse indicator list */
|
79
|
+
|
80
|
+
assert(s); /* check the function argument */
|
81
|
+
for (p = i_body; *p; p++) /* check 'body' indicators */
|
82
|
+
if (strcmp(s, *p) == 0) return APP_BODY;
|
83
|
+
for (p = i_head; *p; p++) /* check 'head' indicators */
|
84
|
+
if (strcmp(s, *p) == 0) return APP_HEAD;
|
85
|
+
for (p = i_both; *p; p++) /* check 'both' indicators */
|
86
|
+
if (strcmp(s, *p) == 0) return APP_BOTH;
|
87
|
+
for (p = i_ignore; *p; p++) /* check 'ignore' indicators */
|
88
|
+
if (strcmp(s, *p) == 0) return APP_NONE;
|
89
|
+
return -1; /* if none found, return error code */
|
90
|
+
} /* _appcode() */
|
91
|
+
|
92
|
+
/*--------------------------------------------------------------------*/
|
93
|
+
|
94
|
+
static int _get_item (ITEMSET *iset, FILE *file)
|
95
|
+
{ /* --- read an item */
|
96
|
+
int d; /* delimiter type */
|
97
|
+
char *buf; /* read buffer */
|
98
|
+
ITEM *item; /* pointer to item */
|
99
|
+
int *vec; /* new item vector */
|
100
|
+
int size; /* new item vector size */
|
101
|
+
|
102
|
+
assert(iset && file); /* check the function arguments */
|
103
|
+
d = ts_next(iset->tscan, file, NULL, 0);
|
104
|
+
buf = ts_buf(iset->tscan); /* read the next field (item name) */
|
105
|
+
if ((d == TS_ERR) || (buf[0] == '\0')) return d;
|
106
|
+
item = nim_byname(iset->nimap, buf);
|
107
|
+
if (!item) { /* look up the name in name/id map */
|
108
|
+
if (iset->app == APP_NONE) /* if new items are to be ignored, */
|
109
|
+
return d; /* do not register the item */
|
110
|
+
item = nim_add(iset->nimap, buf, sizeof(ITEM));
|
111
|
+
if (!item) return E_NOMEM; /* add the new item to the map, */
|
112
|
+
item->frq = item->xfq = 0; /* initialize the frequency counters */
|
113
|
+
item->app = iset->app; /* (occurrence and sum of t.a. sizes) */
|
114
|
+
} /* and set the appearance indicator */
|
115
|
+
size = iset->vsz; /* get the item vector size */
|
116
|
+
if (iset->cnt >= size) { /* if the item vector is full */
|
117
|
+
size += (size > BLKSIZE) ? (size >> 1) : BLKSIZE;
|
118
|
+
vec = (int*)realloc(iset->items, size *sizeof(int));
|
119
|
+
if (!vec) return E_NOMEM; /* enlarge the item vector */
|
120
|
+
iset->items = vec; iset->vsz = size;
|
121
|
+
} /* set the new vector and its size */
|
122
|
+
iset->items[iset->cnt++] = item->id;
|
123
|
+
return d; /* add the item to the transaction */
|
124
|
+
} /* _get_item() */ /* and return the delimiter type */
|
125
|
+
|
126
|
+
/*--------------------------------------------------------------------*/
|
127
|
+
|
128
|
+
static int _nocmp (const void *p1, const void *p2, void *data)
|
129
|
+
{ /* --- compare item frequencies */
|
130
|
+
if (((const ITEM*)p1)->app == APP_NONE)
|
131
|
+
return (((const ITEM*)p2)->app == APP_NONE) ? 0 : 1;
|
132
|
+
if (((const ITEM*)p2)->app == APP_NONE) return -1;
|
133
|
+
#ifdef ARCH64
|
134
|
+
if (((const ITEM*)p1)->frq < (long)data)
|
135
|
+
return (((const ITEM*)p2)->frq < (long)data) ? 0 : 1;
|
136
|
+
if (((const ITEM*)p2)->frq < (long)data) return -1;
|
137
|
+
#else
|
138
|
+
if (((const ITEM*)p1)->frq < (int)data)
|
139
|
+
return (((const ITEM*)p2)->frq < (int)data) ? 0 : 1;
|
140
|
+
if (((const ITEM*)p2)->frq < (int)data) return -1;
|
141
|
+
#endif
|
142
|
+
if (((const ITEM*)p1)->id > ((const ITEM*)p2)->id) return 1;
|
143
|
+
if (((const ITEM*)p1)->id < ((const ITEM*)p2)->id) return -1;
|
144
|
+
return 0; /* return sign of identifier diff. */
|
145
|
+
} /* _nocmp() */
|
146
|
+
|
147
|
+
/*--------------------------------------------------------------------*/
|
148
|
+
|
149
|
+
static int _asccmp (const void *p1, const void *p2, void *data)
|
150
|
+
{ /* --- compare item frequencies */
|
151
|
+
if (((const ITEM*)p1)->app == APP_NONE)
|
152
|
+
return (((const ITEM*)p2)->app == APP_NONE) ? 0 : 1;
|
153
|
+
if (((const ITEM*)p2)->app == APP_NONE) return -1;
|
154
|
+
#ifdef ARCH64
|
155
|
+
if (((const ITEM*)p1)->frq < (long)data)
|
156
|
+
return (((const ITEM*)p2)->frq < (long)data) ? 0 : 1;
|
157
|
+
if (((const ITEM*)p2)->frq < (long)data) return -1;
|
158
|
+
#else
|
159
|
+
if (((const ITEM*)p1)->frq < (int)data)
|
160
|
+
return (((const ITEM*)p2)->frq < (int)data) ? 0 : 1;
|
161
|
+
if (((const ITEM*)p2)->frq < (int)data) return -1;
|
162
|
+
#endif
|
163
|
+
if (((const ITEM*)p1)->frq > ((const ITEM*)p2)->frq) return 1;
|
164
|
+
if (((const ITEM*)p1)->frq < ((const ITEM*)p2)->frq) return -1;
|
165
|
+
return 0; /* return sign of frequency diff. */
|
166
|
+
} /* _asccmp() */
|
167
|
+
|
168
|
+
/*--------------------------------------------------------------------*/
|
169
|
+
|
170
|
+
static int _descmp (const void *p1, const void *p2, void *data)
|
171
|
+
{ /* --- compare item frequencies */
|
172
|
+
if (((const ITEM*)p1)->app == APP_NONE)
|
173
|
+
return (((const ITEM*)p2)->app == APP_NONE) ? 0 : 1;
|
174
|
+
if (((const ITEM*)p2)->app == APP_NONE) return -1;
|
175
|
+
if (((const ITEM*)p1)->frq > ((const ITEM*)p2)->frq) return -1;
|
176
|
+
if (((const ITEM*)p1)->frq < ((const ITEM*)p2)->frq) return 1;
|
177
|
+
return 0; /* return sign of frequency diff. */
|
178
|
+
} /* _descmp() */
|
179
|
+
|
180
|
+
/*--------------------------------------------------------------------*/
|
181
|
+
|
182
|
+
static int _asccmpx (const void *p1, const void *p2, void *data)
|
183
|
+
{ /* --- compare item frequencies */
|
184
|
+
if (((const ITEM*)p1)->app == APP_NONE)
|
185
|
+
return (((const ITEM*)p2)->app == APP_NONE) ? 0 : 1;
|
186
|
+
if (((const ITEM*)p2)->app == APP_NONE) return -1;
|
187
|
+
#ifdef ARCH64
|
188
|
+
if (((const ITEM*)p1)->frq < (long)data)
|
189
|
+
return (((const ITEM*)p2)->frq < (long)data) ? 0 : 1;
|
190
|
+
if (((const ITEM*)p2)->frq < (long)data) return -1;
|
191
|
+
#else
|
192
|
+
if (((const ITEM*)p1)->frq < (int)data)
|
193
|
+
return (((const ITEM*)p2)->frq < (int)data) ? 0 : 1;
|
194
|
+
if (((const ITEM*)p2)->frq < (int)data) return -1;
|
195
|
+
#endif
|
196
|
+
if (((const ITEM*)p1)->xfq > ((const ITEM*)p2)->xfq) return 1;
|
197
|
+
if (((const ITEM*)p1)->xfq < ((const ITEM*)p2)->xfq) return -1;
|
198
|
+
return 0; /* return sign of frequency diff. */
|
199
|
+
} /* _asccmpx() */
|
200
|
+
|
201
|
+
/*--------------------------------------------------------------------*/
|
202
|
+
|
203
|
+
static int _descmpx (const void *p1, const void *p2, void *data)
|
204
|
+
{ /* --- compare item frequencies */
|
205
|
+
if (((const ITEM*)p1)->app == APP_NONE)
|
206
|
+
return (((const ITEM*)p2)->app == APP_NONE) ? 0 : 1;
|
207
|
+
if (((const ITEM*)p2)->app == APP_NONE) return -1;
|
208
|
+
#ifdef ARCH64
|
209
|
+
if (((const ITEM*)p1)->frq < (long)data)
|
210
|
+
return (((const ITEM*)p2)->frq < (long)data) ? 0 : 1;
|
211
|
+
if (((const ITEM*)p2)->frq < (long)data) return -1;
|
212
|
+
#else
|
213
|
+
if (((const ITEM*)p1)->frq < (int)data)
|
214
|
+
return (((const ITEM*)p2)->frq < (int)data) ? 0 : 1;
|
215
|
+
if (((const ITEM*)p2)->frq < (int)data) return -1;
|
216
|
+
#endif
|
217
|
+
if (((const ITEM*)p1)->xfq > ((const ITEM*)p2)->xfq) return -1;
|
218
|
+
if (((const ITEM*)p1)->xfq < ((const ITEM*)p2)->xfq) return 1;
|
219
|
+
return 0; /* return sign of frequency diff. */
|
220
|
+
} /* _descmpx() */
|
221
|
+
|
222
|
+
/*----------------------------------------------------------------------
|
223
|
+
Item Set Functions
|
224
|
+
----------------------------------------------------------------------*/
|
225
|
+
|
226
|
+
ITEMSET* is_create (int cnt)
|
227
|
+
{ /* --- create an item set */
|
228
|
+
ITEMSET *iset; /* created item set */
|
229
|
+
|
230
|
+
if (cnt <= 0) cnt = BLKSIZE; /* check and adapt number of items */
|
231
|
+
iset = malloc(sizeof(ITEMSET));
|
232
|
+
if (!iset) return NULL; /* create an item set */
|
233
|
+
iset->tscan = ts_create(); /* and its components */
|
234
|
+
ts_chars(iset->tscan, TS_NULL, "");
|
235
|
+
iset->nimap = nim_create(0, 0, (HASHFN*)0, (SYMFN*)0);
|
236
|
+
iset->items = (int*)malloc(cnt *sizeof(int));
|
237
|
+
if (!iset->tscan || !iset->nimap || !iset->items) {
|
238
|
+
is_delete(iset); return NULL; }
|
239
|
+
iset->tac = iset->cnt = 0; /* initialize the other fields */
|
240
|
+
iset->app = APP_BOTH;
|
241
|
+
iset->vsz = cnt;
|
242
|
+
iset->chars[0] = ' '; iset->chars[1] = ' ';
|
243
|
+
iset->chars[2] = '\n'; iset->chars[3] = '\0';
|
244
|
+
return iset; /* return the created item set */
|
245
|
+
} /* is_create() */
|
246
|
+
|
247
|
+
/*--------------------------------------------------------------------*/
|
248
|
+
|
249
|
+
void is_delete (ITEMSET *iset)
|
250
|
+
{ /* --- delete an item set */
|
251
|
+
assert(iset); /* check the function argument */
|
252
|
+
if (iset->items) free(iset->items);
|
253
|
+
if (iset->nimap) nim_delete(iset->nimap);
|
254
|
+
if (iset->tscan) ts_delete(iset->tscan);
|
255
|
+
free(iset); /* delete the components */
|
256
|
+
} /* is_delete() */ /* and the item set body */
|
257
|
+
|
258
|
+
/*--------------------------------------------------------------------*/
|
259
|
+
|
260
|
+
void is_chars (ITEMSET *iset, const char *blanks, const char *fldseps,
|
261
|
+
const char *recseps, const char *comment)
|
262
|
+
{ /* --- set special characters */
|
263
|
+
assert(iset); /* check the function argument */
|
264
|
+
if (blanks) /* set blank characters */
|
265
|
+
iset->chars[0] = ts_chars(iset->tscan, TS_BLANK, blanks);
|
266
|
+
if (fldseps) /* set field separators */
|
267
|
+
iset->chars[1] = ts_chars(iset->tscan, TS_FLDSEP, fldseps);
|
268
|
+
if (recseps) /* set record separators */
|
269
|
+
iset->chars[2] = ts_chars(iset->tscan, TS_RECSEP, recseps);
|
270
|
+
if (comment) /* set comment indicators */
|
271
|
+
ts_chars(iset->tscan, TS_COMMENT, comment);
|
272
|
+
} /* is_chars() */
|
273
|
+
|
274
|
+
/*--------------------------------------------------------------------*/
|
275
|
+
|
276
|
+
int is_item (ITEMSET *iset, const char *name)
|
277
|
+
{ /* --- get an item identifier */
|
278
|
+
ITEM *item = nim_byname(iset->nimap, name);
|
279
|
+
return (item) ? item->id :-1; /* look up the given name */
|
280
|
+
} /* is_item() */ /* in the name/identifier map */
|
281
|
+
|
282
|
+
/*--------------------------------------------------------------------*/
|
283
|
+
|
284
|
+
int is_readapp (ITEMSET *iset, FILE *file)
|
285
|
+
{ /* --- read appearance indicators */
|
286
|
+
int d; /* delimiter type */
|
287
|
+
char *buf; /* read buffer */
|
288
|
+
ITEM *item; /* to access the item data */
|
289
|
+
|
290
|
+
assert(iset && file); /* check the function arguments */
|
291
|
+
buf = ts_buf(iset->tscan); /* read the first record (one field) */
|
292
|
+
d = ts_next(iset->tscan, file, NULL, 0);
|
293
|
+
if (d == TS_ERR) return E_FREAD;
|
294
|
+
if (d != TS_REC) return E_FLDCNT;
|
295
|
+
iset->app = _appcode(buf); /* get default appearance code */
|
296
|
+
if (iset->app < 0) return E_UNKAPP;
|
297
|
+
while (1) { /* read item/indicator pairs */
|
298
|
+
d = ts_next(iset->tscan, file, NULL, 0);
|
299
|
+
if (d <= TS_EOF) /* read the next item */
|
300
|
+
return (d == TS_ERR) ? E_FREAD : 0;
|
301
|
+
if (buf[0] == '\0') /* check for end of file */
|
302
|
+
return E_ITEMEXP; /* and for a missing item */
|
303
|
+
item = nim_add(iset->nimap, buf, sizeof(ITEM));
|
304
|
+
if (item == EXISTS) return E_DUPITEM; /* add the new item */
|
305
|
+
if (item == NULL) return E_NOMEM; /* to the name/id map */
|
306
|
+
item->frq = 0; /* clear the frequency counters */
|
307
|
+
item->xfq = 0; /* (occurrence and sum of t.a. sizes) */
|
308
|
+
if (d != TS_FLD) return E_APPEXP;
|
309
|
+
d = ts_next(iset->tscan, file, NULL, 0);
|
310
|
+
if (d == TS_ERR) return E_FREAD;
|
311
|
+
if (d == TS_FLD) return E_FLDCNT;
|
312
|
+
item->app = _appcode(buf); /* get the appearance indicator */
|
313
|
+
if (item->app < 0) return E_UNKAPP;
|
314
|
+
}
|
315
|
+
return 0; /* return 'ok' */
|
316
|
+
} /* is_readapp() */
|
317
|
+
|
318
|
+
/*--------------------------------------------------------------------*/
|
319
|
+
|
320
|
+
int is_read (ITEMSET *iset, FILE *file)
|
321
|
+
{ /* --- read a transaction */
|
322
|
+
int i, d; /* loop variable, delimiter type */
|
323
|
+
char *buf; /* read buffer */
|
324
|
+
ITEM *item; /* pointer to item */
|
325
|
+
|
326
|
+
|
327
|
+
assert(iset && file); /* check the function arguments */
|
328
|
+
iset->cnt = 0; /* initialize the item counter */
|
329
|
+
d = _get_item(iset, file); /* read the first item and */
|
330
|
+
buf = ts_buf(iset->tscan); /* get the read buffer */
|
331
|
+
|
332
|
+
if ((d == TS_EOF) /* if at the end of the file */
|
333
|
+
&& (buf[0] == '\0')) /* and no item has been read, */
|
334
|
+
return 1; /* return 'end of file' */
|
335
|
+
while ((d == TS_FLD) /* read the other items */
|
336
|
+
&& (buf[0] != '\0')) /* of the transaction */
|
337
|
+
d = _get_item(iset, file); /* up to the end of the record */
|
338
|
+
if (d == TS_ERR) return d; /* check for a read error */
|
339
|
+
if ((buf[0] == '\0') && (d == TS_FLD) && (iset->cnt > 0))
|
340
|
+
return E_ITEMEXP; /* check for an empty field */
|
341
|
+
ta_sort(iset->items, iset->cnt); /* prepare the transaction */
|
342
|
+
iset->cnt = ta_unique(iset->items, iset->cnt);
|
343
|
+
for (i = iset->cnt; --i >= 0; ) {
|
344
|
+
item = nim_byid(iset->nimap, iset->items[i]);
|
345
|
+
item->frq += 1; /* count the item and */
|
346
|
+
item->xfq += iset->cnt; /* sum the transaction sizes */
|
347
|
+
} /* as an importance indicator */
|
348
|
+
iset->tac += 1; /* count the transaction */
|
349
|
+
return 0; /* return 'ok' */
|
350
|
+
} /* is_read() */
|
351
|
+
|
352
|
+
/*--------------------------------------------------------------------*/
|
353
|
+
|
354
|
+
int is_recode (ITEMSET *iset, int minfrq, int dir, int *map)
|
355
|
+
{ /* --- recode items w.r.t. frequency */
|
356
|
+
int i, k, n, t; /* loop variables, buffer */
|
357
|
+
ITEM *item; /* to traverse the items */
|
358
|
+
SYMCMPFN *cmp; /* comparison function */
|
359
|
+
|
360
|
+
assert(iset); /* check the function arguments */
|
361
|
+
if (dir > 1) cmp = _asccmpx; /* get the appropriate */
|
362
|
+
else if (dir > 0) cmp = _asccmp; /* comparison function */
|
363
|
+
else if (dir >= 0) cmp = _nocmp; /* (ascending/descending) */
|
364
|
+
else if (dir > -2) cmp = _descmp; /* and sort the items */
|
365
|
+
else cmp = _descmpx; /* w.r.t. their frequency */
|
366
|
+
nim_sort(iset->nimap, cmp, (void*)minfrq, map, 1);
|
367
|
+
for (n = nim_cnt(iset->nimap); --n >= 0; ) {
|
368
|
+
item = (ITEM*)nim_byid(iset->nimap, n);
|
369
|
+
if (item->frq < minfrq) /* determine frequent items and */
|
370
|
+
item->app = APP_NONE; /* set all others to 'ignore' */
|
371
|
+
else if (item->app != APP_NONE)
|
372
|
+
break; /* in addition, skip all items */
|
373
|
+
} /* that have been set to 'ignore' */
|
374
|
+
if (map) { /* if a map vector is provided */
|
375
|
+
for (i = k = 0; i < iset->cnt; i++) {
|
376
|
+
t = map[iset->items[i]]; /* traverse the current transaction */
|
377
|
+
if (t <= n) iset->items[k++] = t;
|
378
|
+
} /* recode all items and */
|
379
|
+
iset->cnt = k; /* delete all items to ignore */
|
380
|
+
ta_sort(iset->items, k); /* resort the items */
|
381
|
+
}
|
382
|
+
return n+1; /* return number of frequent items */
|
383
|
+
} /* is_recode() */
|
384
|
+
|
385
|
+
/*--------------------------------------------------------------------*/
|
386
|
+
|
387
|
+
int is_filter (ITEMSET *iset, const char *marks)
|
388
|
+
{ /* --- filter items in transaction */
|
389
|
+
return iset->cnt = ta_filter(iset->items, iset->cnt, marks);
|
390
|
+
} /* is_filter() */
|
391
|
+
|
392
|
+
/*----------------------------------------------------------------------
|
393
|
+
Item Set Evaluation Functions
|
394
|
+
----------------------------------------------------------------------*/
|
395
|
+
|
396
|
+
ISEVAL* ise_create (ITEMSET *iset, int tacnt)
|
397
|
+
{ /* --- create an item set evaluation */
|
398
|
+
int i; /* loop variable */
|
399
|
+
ISEVAL *eval; /* created item set evaluator */
|
400
|
+
|
401
|
+
i = is_cnt(iset); /* get the number of items */
|
402
|
+
eval = (ISEVAL*)malloc(sizeof(ISEVAL) +(i+i) *sizeof(double));
|
403
|
+
if (!eval) return NULL; /* create an evaluation object */
|
404
|
+
eval->logfs = eval->lsums +i +1; /* and organize the memory */
|
405
|
+
eval->logta = log(tacnt); /* store log of number of trans. */
|
406
|
+
while (--i >= 0) /* compute logarithms of item freqs. */
|
407
|
+
eval->logfs[i] = log(is_getfrq(iset, i));
|
408
|
+
eval->lsums[0] = 0; /* init. first sum of logarithms */
|
409
|
+
return eval; /* return created item set evaluator */
|
410
|
+
} /* ise_create() */
|
411
|
+
|
412
|
+
/*--------------------------------------------------------------------*/
|
413
|
+
|
414
|
+
double ise_eval (ISEVAL *eval, int *ids, int cnt, int pfx, double supp)
|
415
|
+
{ /* --- evaluate an item set */
|
416
|
+
double sum; /* sum of logarithms of frequencies */
|
417
|
+
|
418
|
+
sum = (pfx > 0) /* if there is a prefix, */
|
419
|
+
? eval->lsums[pfx-1] : 0; /* get already known logarithm sum */
|
420
|
+
for ( ; pfx < cnt; pfx++) /* compute and add remaining terms */
|
421
|
+
eval->lsums[pfx] = sum += eval->logfs[ids[pfx]];
|
422
|
+
return (log(supp) -sum +(cnt-1) *eval->logta) * (1.0/LN_2);
|
423
|
+
} /* ise_eval() */ /* compute logarithm of quotient */
|
424
|
+
|
425
|
+
/*----------------------------------------------------------------------
|
426
|
+
Item Set Formatting Functions
|
427
|
+
----------------------------------------------------------------------*/
|
428
|
+
|
429
|
+
ISFMTR* isf_create (ITEMSET *iset, int scan)
|
430
|
+
{ /* --- create an item set formatter */
|
431
|
+
int i, k, n; /* loop variable, buffers */
|
432
|
+
int len, sum; /* length of an item name and sum */
|
433
|
+
ISFMTR *fmt; /* created item set formatter */
|
434
|
+
char buf[4*TS_SIZE+4]; /* buffer for formatting */
|
435
|
+
const char *name; /* to traverse the item names */
|
436
|
+
char *copy; /* for copies of formatted names */
|
437
|
+
|
438
|
+
n = is_cnt(iset); /* get the number of items */
|
439
|
+
fmt = (ISFMTR*)malloc(sizeof(ISFMTR) + n *sizeof(int)
|
440
|
+
+(n-1) *sizeof(char*));
|
441
|
+
if (!fmt) return NULL; /* create the base structure */
|
442
|
+
fmt->buf = NULL; /* and organize the memory */
|
443
|
+
fmt->offs = (int*)(fmt->names +n);
|
444
|
+
for (i = sum = fmt->cnt = 0; i < n; i++) {
|
445
|
+
name = is_name(iset, i); /* traverse the item names */
|
446
|
+
len = strlen(name); /* and get their length */
|
447
|
+
sum += k = (scan) ? sc_format(buf, name, 0) : len;
|
448
|
+
if (k > len) { /* if formatting was needed */
|
449
|
+
copy = (char*)malloc((k+1) *sizeof(char));
|
450
|
+
if (!copy) { fmt->cnt = i-1; isf_delete(fmt); return NULL; }
|
451
|
+
name = strcpy(copy, buf); /* copy the formatted name */
|
452
|
+
} /* into a newly created string */
|
453
|
+
fmt->names[i] = name; /* store (formatted) item name */
|
454
|
+
} /* afterwards create output buffer */
|
455
|
+
if (scan) fmt->cnt = n; /* note the number of items */
|
456
|
+
fmt->buf = (char*)malloc((sum +n +1) *sizeof(char));
|
457
|
+
if (!fmt->buf) { isf_delete(fmt); return NULL; }
|
458
|
+
fmt->offs[0] = 0; /* init. the first prefix offset */
|
459
|
+
return fmt; /* return created item set formatter */
|
460
|
+
} /* isf_create() */
|
461
|
+
|
462
|
+
/*--------------------------------------------------------------------*/
|
463
|
+
|
464
|
+
void isf_delete (ISFMTR *fmt)
|
465
|
+
{ /* --- delete an item set formatter */
|
466
|
+
int i; /* loop variable */
|
467
|
+
for (i = fmt->cnt; --i >= 0; )
|
468
|
+
if ((fmt->names[i] != NULL)
|
469
|
+
&& (fmt->names[i][0] == '"'))
|
470
|
+
free((void*)fmt->names[i]);
|
471
|
+
if (fmt->buf) free(fmt->buf); /* delete reformatted item names, */
|
472
|
+
free(fmt); /* the output buffer and the base */
|
473
|
+
} /* isf_delete() */
|
474
|
+
|
475
|
+
/*--------------------------------------------------------------------*/
|
476
|
+
|
477
|
+
const char* isf_format (ISFMTR *fmt, int *ids, int cnt, int pre)
|
478
|
+
{ /* --- format an item set */
|
479
|
+
char *p; /* to traverse the output buffer */
|
480
|
+
const char *name; /* to traverse the item names */
|
481
|
+
|
482
|
+
p = fmt->buf +fmt->offs[pre]; /* get position for appending */
|
483
|
+
while (pre < cnt) { /* traverse the additional items */
|
484
|
+
name = fmt->names[ids[pre]];/* copy the item name to the output */
|
485
|
+
while (*name) *p++ = *name++;
|
486
|
+
*p++ = ' '; /* add an item separator */
|
487
|
+
fmt->offs[++pre] = (int)(p-fmt->buf);
|
488
|
+
} /* record the new offset */
|
489
|
+
*p = '\0'; /* terminate the formatted item set */
|
490
|
+
fmt->len = (int)(p-fmt->buf); /* note the length of the description */
|
491
|
+
return fmt->buf; /* return the output buffer */
|
492
|
+
} /* isf_format() */
|
493
|
+
|
494
|
+
/*----------------------------------------------------------------------
|
495
|
+
Transaction Functions
|
496
|
+
----------------------------------------------------------------------*/
|
497
|
+
|
498
|
+
int ta_unique (int *items, int n)
|
499
|
+
{ /* --- remove duplicate items */
|
500
|
+
int *s, *d; /* to traverse the item vector */
|
501
|
+
|
502
|
+
assert(items && (n >= 0)); /* check the function arguments */
|
503
|
+
if (n <= 1) return n; /* check for 0 or 1 item */
|
504
|
+
for (d = s = items; --n > 0;) /* traverse the sorted vector */
|
505
|
+
if (*++s != *d) *++d = *s; /* and remove duplicate items */
|
506
|
+
return (int)(++d -items); /* return the new number of items */
|
507
|
+
} /* ta_unique() */
|
508
|
+
|
509
|
+
/*--------------------------------------------------------------------*/
|
510
|
+
|
511
|
+
int ta_filter (int *items, int n, const char *marks)
|
512
|
+
{ /* --- filter items in a transaction */
|
513
|
+
int i, k; /* loop variables */
|
514
|
+
|
515
|
+
assert(items && (n >= 0)); /* check the function arguments */
|
516
|
+
for (i = k = 0; i < n; i++) /* remove all unmarked items */
|
517
|
+
if (marks[items[i]]) items[k++] = items[i];
|
518
|
+
return k; /* return the new number of items */
|
519
|
+
} /* ta_filter() */
|
520
|
+
|
521
|
+
/*--------------------------------------------------------------------*/
|
522
|
+
|
523
|
+
static int ta_cmp (const void *p1, const void *p2, void *data)
|
524
|
+
{ /* --- compare transactions */
|
525
|
+
int k, k1, k2; /* loop variable, counters */
|
526
|
+
const int *i1, *i2; /* to traverse the item identifiers */
|
527
|
+
|
528
|
+
assert(p1 && p2); /* check the function arguments */
|
529
|
+
i1 = ((const TRACT*)p1)->items;
|
530
|
+
i2 = ((const TRACT*)p2)->items;
|
531
|
+
k1 = ((const TRACT*)p1)->cnt; /* get the item vectors */
|
532
|
+
k2 = ((const TRACT*)p2)->cnt; /* and the numbers of items */
|
533
|
+
for (k = (k1 < k2) ? k1 : k2; --k >= 0; i1++, i2++) {
|
534
|
+
if (*i1 > *i2) return 1; /* compare corresponding items */
|
535
|
+
if (*i1 < *i2) return -1; /* and abort the comparison */
|
536
|
+
} /* if one of them is greater */
|
537
|
+
if (k1 > k2) return 1; /* if one of the transactions */
|
538
|
+
if (k1 < k2) return -1; /* is not empty, it is greater */
|
539
|
+
return 0; /* otherwise the two trans. are equal */
|
540
|
+
} /* ta_cmp() */
|
541
|
+
|
542
|
+
/*--------------------------------------------------------------------*/
|
543
|
+
|
544
|
+
static int ta_cmpx (const TRACT *ta, const int *items, int n)
|
545
|
+
{ /* --- compare transactions */
|
546
|
+
int k, m; /* loop variable, counter */
|
547
|
+
const int *p; /* to traverse the item identifiers */
|
548
|
+
|
549
|
+
assert(ta && items); /* check the function arguments */
|
550
|
+
p = ta->items; m = ta->cnt; /* traverse the item vector */
|
551
|
+
m = ta->cnt;
|
552
|
+
for (k = (n < m) ? n : m; --k >= 0; p++, items++) {
|
553
|
+
if (*p > *items) return 1; /* compare corresponding items */
|
554
|
+
if (*p < *items) return -1; /* and abort the comparison */
|
555
|
+
} /* if one of them is greater */
|
556
|
+
if (m > n) return 1; /* if one of the transactions */
|
557
|
+
if (m < n) return -1; /* is not empty, it is greater */
|
558
|
+
return 0; /* otherwise the two trans. are equal */
|
559
|
+
} /* ta_cmpx() */
|
560
|
+
|
561
|
+
/*----------------------------------------------------------------------
|
562
|
+
Transaction Set Functions
|
563
|
+
----------------------------------------------------------------------*/
|
564
|
+
|
565
|
+
TASET* tas_create (ITEMSET *itemset)
|
566
|
+
{ /* --- create a transaction set */
|
567
|
+
TASET *taset; /* created transaction set */
|
568
|
+
|
569
|
+
assert(itemset); /* check the function argument */
|
570
|
+
taset = malloc(sizeof(TASET));
|
571
|
+
if (!taset) return NULL; /* create a transaction set */
|
572
|
+
taset->itemset = itemset; /* and store the item set */
|
573
|
+
taset->cnt = taset->vsz = taset->max = taset->total = 0;
|
574
|
+
taset->tracts = NULL; /* initialize the other fields */
|
575
|
+
return taset; /* return the created t.a. set */
|
576
|
+
} /* tas_create() */
|
577
|
+
|
578
|
+
/*--------------------------------------------------------------------*/
|
579
|
+
|
580
|
+
void tas_delete (TASET *taset, int delis)
|
581
|
+
{ /* --- delete a transaction set */
|
582
|
+
assert(taset); /* check the function argument */
|
583
|
+
if (taset->tracts) { /* if there are loaded transactions */
|
584
|
+
while (--taset->cnt >= 0) /* traverse the transaction vector */
|
585
|
+
free(taset->tracts[taset->cnt]);
|
586
|
+
free(taset->tracts); /* delete all transactions */
|
587
|
+
} /* and the transaction vector */
|
588
|
+
if (delis && taset->itemset) is_delete(taset->itemset);
|
589
|
+
free(taset); /* delete the item set and */
|
590
|
+
} /* tas_delete() */ /* the transaction set body */
|
591
|
+
|
592
|
+
/*--------------------------------------------------------------------*/
|
593
|
+
|
594
|
+
int tas_add (TASET *taset, const int *items, int n)
|
595
|
+
{ /* --- add a transaction */
|
596
|
+
TRACT *ta; /* new transaction */
|
597
|
+
int *p; /* to traverse the transaction */
|
598
|
+
TRACT **vec; /* new transaction vector */
|
599
|
+
int size; /* new transaction vector size */
|
600
|
+
|
601
|
+
assert(taset); /* check the function arguments */
|
602
|
+
size = taset->vsz; /* get the transaction vector size */
|
603
|
+
if (taset->cnt >= size) { /* if the transaction vector is full */
|
604
|
+
size += (size > BLKSIZE) ? (size >> 1) : BLKSIZE;
|
605
|
+
vec = (TRACT**)realloc(taset->tracts, size *sizeof(TRACT*));
|
606
|
+
if (!vec) return -1; /* enlarge the transaction vector */
|
607
|
+
taset->tracts = vec; taset->vsz = size;
|
608
|
+
} /* set the new vector and its size */
|
609
|
+
if (!items) { /* if no transaction is given */
|
610
|
+
items = is_tract(taset->itemset);
|
611
|
+
n = is_tsize(taset->itemset);
|
612
|
+
} /* get it from the item set */
|
613
|
+
ta = (TRACT*)malloc(sizeof(TRACT) +(n-1) *sizeof(int));
|
614
|
+
if (!ta) return -1; /* create a new transaction */
|
615
|
+
taset->tracts[taset->cnt++] = ta;
|
616
|
+
if (n > taset->max) /* store the transaction and */
|
617
|
+
taset->max = n; /* update maximal transaction size */
|
618
|
+
taset->total += n; /* sum the number of items */
|
619
|
+
for (p = ta->items +(ta->cnt = n); --n >= 0; )
|
620
|
+
*--p = items[n]; /* copy the items of the t.a. */
|
621
|
+
return 0; /* return 'ok' */
|
622
|
+
} /* tas_add() */
|
623
|
+
|
624
|
+
/*--------------------------------------------------------------------*/
|
625
|
+
|
626
|
+
void tas_recode (TASET *taset, int *map, int cnt)
|
627
|
+
{ /* --- recode items */
|
628
|
+
int i, k, n, x; /* loop variables, buffer */
|
629
|
+
TRACT *t; /* to traverse the transactions */
|
630
|
+
int *p; /* to traverse the item identifiers */
|
631
|
+
|
632
|
+
assert(taset && map); /* check the function arguments */
|
633
|
+
taset->max = taset->total = 0;/* clear the maximal size and total */
|
634
|
+
for (n = taset->cnt; --n >= 0; ) {
|
635
|
+
t = taset->tracts[n]; /* traverse the transactions and */
|
636
|
+
p = t->items; /* the items of each transaction */
|
637
|
+
for (i = k = 0; i < t->cnt; i++) {
|
638
|
+
x = map[p[i]]; /* recode the items and */
|
639
|
+
if (x < cnt) p[k++] = x; /* remove superfluous items */
|
640
|
+
} /* from the transaction */
|
641
|
+
if (k > taset->max) /* update the max. transaction size */
|
642
|
+
taset->max = k; /* with the new size of the t.a. */
|
643
|
+
taset->total += k; /* sum the number of items */
|
644
|
+
ta_sort(t->items, t->cnt = k);
|
645
|
+
} /* resort the item identifiers */
|
646
|
+
} /* tas_recode() */
|
647
|
+
|
648
|
+
/*--------------------------------------------------------------------*/
|
649
|
+
|
650
|
+
int tas_filter (TASET *taset, const char *marks)
|
651
|
+
{ /* --- filter items in a trans. set */
|
652
|
+
int i, max = 0; /* loop variable, max. num. of items */
|
653
|
+
TRACT *t; /* to traverse the transactions */
|
654
|
+
|
655
|
+
assert(taset && marks); /* check the function arguments */
|
656
|
+
taset->total = 0; /* clear the total number of items */
|
657
|
+
for (i = taset->cnt; --i >= 0; ) {
|
658
|
+
t = taset->tracts[i]; /* traverse the transactions */
|
659
|
+
t->cnt = ta_filter(t->items, t->cnt, marks);
|
660
|
+
if (t->cnt > max) max = t->cnt;
|
661
|
+
taset->total += t->cnt; /* filter each transaction and */
|
662
|
+
} /* update maximal size and total */
|
663
|
+
return max; /* return maximum number of items */
|
664
|
+
} /* tas_filter() */
|
665
|
+
|
666
|
+
/*--------------------------------------------------------------------*/
|
667
|
+
|
668
|
+
void tas_sort (TASET *taset, int heap)
|
669
|
+
{ /* --- sort a transaction set */
|
670
|
+
assert(taset); /* check the function argument */
|
671
|
+
if (heap) v_heapsort(taset->tracts, taset->cnt, ta_cmp, NULL);
|
672
|
+
else v_sort (taset->tracts, taset->cnt, ta_cmp, NULL);
|
673
|
+
} /* tas_sort() */
|
674
|
+
|
675
|
+
/*--------------------------------------------------------------------*/
|
676
|
+
|
677
|
+
int tas_occur (TASET *taset, const int *items, int n)
|
678
|
+
{ /* --- count transaction occurrences */
|
679
|
+
int l, r, m, k = taset->cnt; /* index variables */
|
680
|
+
|
681
|
+
assert(taset && items); /* check the function arguments */
|
682
|
+
for (r = m = 0; r < k; ) { /* find right boundary */
|
683
|
+
m = (r + k) >> 1; /* by a binary search */
|
684
|
+
if (ta_cmpx(taset->tracts[m], items, n) > 0) k = m;
|
685
|
+
else r = m+1;
|
686
|
+
}
|
687
|
+
for (l = m = 0; l < k; ) { /* find left boundary */
|
688
|
+
m = (l + k) >> 1; /* by a binary search */
|
689
|
+
if (ta_cmpx(taset->tracts[m], items, n) < 0) l = m+1;
|
690
|
+
else k = m;
|
691
|
+
}
|
692
|
+
return r -l; /* compute the number of occurrences */
|
693
|
+
} /* tas_occur() */
|
694
|
+
|
695
|
+
/*--------------------------------------------------------------------*/
|
696
|
+
#ifndef NDEBUG
|
697
|
+
|
698
|
+
void tas_show (TASET *taset)
|
699
|
+
{ /* --- show a transaction set */
|
700
|
+
int i, k; /* loop variables */
|
701
|
+
TRACT *t; /* to traverse the transactions */
|
702
|
+
|
703
|
+
assert(taset); /* check the function argument */
|
704
|
+
for (i = 0; i < taset->cnt; i++) {
|
705
|
+
t = taset->tracts[i]; /* traverse the transactions */
|
706
|
+
for (k = 0; k < t->cnt; k++) { /* traverse the items */
|
707
|
+
if (k > 0) putc(' ', stdout); /* print a separator */
|
708
|
+
printf(is_name(taset->itemset, t->items[k]));
|
709
|
+
} /* print the next item */
|
710
|
+
putc('\n', stdout); /* terminate the transaction */
|
711
|
+
} /* finally print the number of t.a. */
|
712
|
+
printf("%d transaction(s)\n", taset->cnt);
|
713
|
+
} /* tas_show() */
|
714
|
+
|
715
|
+
#endif
|
716
|
+
/*----------------------------------------------------------------------
|
717
|
+
Transaction Tree Functions
|
718
|
+
----------------------------------------------------------------------*/
|
719
|
+
|
720
|
+
TATREE* _create (TRACT **tracts, int cnt, int index)
|
721
|
+
{ /* --- recursive part of tat_create() */
|
722
|
+
int i, k, t; /* loop variables, buffer */
|
723
|
+
int item, n; /* item and item counter */
|
724
|
+
TATREE *tat; /* created transaction tree */
|
725
|
+
TATREE **vec; /* vector of child pointers */
|
726
|
+
|
727
|
+
assert(tracts /* check the function arguments */
|
728
|
+
&& (cnt >= 0) && (index >= 0));
|
729
|
+
if (cnt <= 1) { /* if only one transaction left */
|
730
|
+
n = (cnt > 0) ? (*tracts)->cnt -index : 0;
|
731
|
+
tat = (TATREE*)malloc(sizeof(TATREE) +(n-1) *sizeof(int));
|
732
|
+
if (!tat) return NULL; /* create a transaction tree node */
|
733
|
+
tat->cnt = cnt; /* and initialize its fields */
|
734
|
+
tat->size = -n;
|
735
|
+
tat->max = n;
|
736
|
+
while (--n >= 0) tat->items[n] = (*tracts)->items[index +n];
|
737
|
+
return tat;
|
738
|
+
}
|
739
|
+
for (k = cnt; (--k >= 0) && ((*tracts)->cnt <= index); )
|
740
|
+
tracts++; /* skip t.a. that are too short */
|
741
|
+
n = 0; item = -1; /* init. item and item counter */
|
742
|
+
for (tracts += i = ++k; --i >= 0; ) {
|
743
|
+
t = (*--tracts)->items[index]; /* traverse the transactions */
|
744
|
+
if (t != item) { item = t; n++; }
|
745
|
+
} /* count the different items */
|
746
|
+
#ifdef ARCH64 /* adapt to even item number */
|
747
|
+
i = (n & 1) ? n : (n+1); /* so that pointer addresses are */
|
748
|
+
#else /* multiples of 8 on 64 bit systems */
|
749
|
+
i = n; /* on 32 bit systems, however, */
|
750
|
+
#endif /* use the exact number of items */
|
751
|
+
tat = (TATREE*)malloc(sizeof(TATREE) + (i-1) *sizeof(int)
|
752
|
+
+ n *sizeof(TATREE*));
|
753
|
+
if (!tat) return NULL; /* create a transaction tree node */
|
754
|
+
tat->cnt = cnt; /* and initialize its fields */
|
755
|
+
tat->size = n;
|
756
|
+
tat->max = 0;
|
757
|
+
if (n <= 0) return tat; /* if t.a. are fully captured, abort */
|
758
|
+
vec = (TATREE**)(tat->items +i);
|
759
|
+
item = tracts[--k]->items[index];
|
760
|
+
for (tracts += i = k; --i >= 0; ) {
|
761
|
+
t = (*--tracts)->items[index]; /* traverse the transactions, */
|
762
|
+
if (t == item) continue; /* but skip those with the same item */
|
763
|
+
tat->items[--n] = item; item = t;
|
764
|
+
vec[n] = _create(tracts+1, k-i, index+1);
|
765
|
+
if (!vec[n]) break; /* note the item identifier */
|
766
|
+
t = vec[n]->max +1; if (t > tat->max) tat->max = t;
|
767
|
+
k = i; /* recursively create subtrees */
|
768
|
+
} /* and adapt the section end index */
|
769
|
+
if (i < 0) { /* if child creation was successful */
|
770
|
+
tat->items[--n] = item; /* note the last item identifier */
|
771
|
+
vec[n] = _create(tracts, k+1, index+1);
|
772
|
+
if (vec[n]) { /* create the last child */
|
773
|
+
t = vec[n]->max +1; if (t > tat->max) tat->max = t;
|
774
|
+
return tat; /* return the created */
|
775
|
+
} /* transaction tree */
|
776
|
+
}
|
777
|
+
for (i = tat->size; --i > n; ) tat_delete(vec[i]);
|
778
|
+
free(tat); /* on error delete created subtrees */
|
779
|
+
return NULL; /* and the transaction tree node */
|
780
|
+
} /* _create() */
|
781
|
+
|
782
|
+
/*--------------------------------------------------------------------*/
|
783
|
+
|
784
|
+
TATREE* tat_create (TASET *taset, int heap)
|
785
|
+
{ /* --- create a transactions tree */
|
786
|
+
assert(taset); /* check the function argument */
|
787
|
+
if (heap) v_heapsort(taset->tracts, taset->cnt, ta_cmp, NULL);
|
788
|
+
else v_sort (taset->tracts, taset->cnt, ta_cmp, NULL);
|
789
|
+
return _create(taset->tracts, taset->cnt, 0);
|
790
|
+
} /* tat_create() */
|
791
|
+
|
792
|
+
/*--------------------------------------------------------------------*/
|
793
|
+
|
794
|
+
void tat_delete (TATREE *tat)
|
795
|
+
{ /* --- delete a transaction tree */
|
796
|
+
int i; /* loop variable */
|
797
|
+
TATREE **vec; /* vector of child nodes */
|
798
|
+
|
799
|
+
assert(tat); /* check the function argument */
|
800
|
+
#ifdef ARCH64 /* if 64 bit architecture */
|
801
|
+
i = (tat->size & 1) ? tat->size : (tat->size+1);
|
802
|
+
#else /* address must be a multiple of 8 */
|
803
|
+
i = tat->size; /* on 32 bit systems, however, */
|
804
|
+
#endif /* use the number of items directly */
|
805
|
+
vec = (TATREE**)(tat->items +i);
|
806
|
+
for (i = tat->size; --i >= 0; )
|
807
|
+
tat_delete(vec[i]); /* recursively delete the subtrees */
|
808
|
+
free(tat); /* and the tree node itself */
|
809
|
+
} /* tat_delete() */
|
810
|
+
|
811
|
+
/*--------------------------------------------------------------------*/
|
812
|
+
#ifdef ARCH64
|
813
|
+
|
814
|
+
TATREE* tat_child (TATREE *tat, int index)
|
815
|
+
{ /* --- go to a child node */
|
816
|
+
int s; /* padded size of the node */
|
817
|
+
|
818
|
+
assert(tat /* check the function arguments */
|
819
|
+
&& (index >= 0) && (index < tat->size));
|
820
|
+
s = (tat->size & 1) ? tat->size : (tat->size +1);
|
821
|
+
return ((TATREE**)(tat->items +s))[index];
|
822
|
+
} /* tat_child */ /* return the child node/subtree */
|
823
|
+
|
824
|
+
#endif
|
825
|
+
/*--------------------------------------------------------------------*/
|
826
|
+
|
827
|
+
void tat_mark (TATREE *tat)
|
828
|
+
{ /* --- mark end of transactions */
|
829
|
+
int i; /* loop variable */
|
830
|
+
|
831
|
+
assert(tat); /* check the function argument */
|
832
|
+
if (tat->size < 0) /* if there is a transaction, */
|
833
|
+
tat->items[tat->max-1] |= INT_MIN; /* mark end of trans. */
|
834
|
+
else { /* if there are subtrees */
|
835
|
+
for (i = tat->size; --i >= 0; )
|
836
|
+
tat_mark(tat_child(tat, i));
|
837
|
+
} /* recursively mark the subtrees */
|
838
|
+
} /* tat_mark() */
|
839
|
+
|
840
|
+
/*--------------------------------------------------------------------*/
|
841
|
+
#ifndef NDEBUG
|
842
|
+
|
843
|
+
void _show (TATREE *tat, int ind)
|
844
|
+
{ /* --- rekursive part of tat_show() */
|
845
|
+
int i, k; /* loop variables */
|
846
|
+
TATREE **vec; /* vector of child nodes */
|
847
|
+
|
848
|
+
assert(tat && (ind >= 0)); /* check the function arguments */
|
849
|
+
if (tat->size <= 0) { /* if this is a leaf node */
|
850
|
+
for (i = 0; i < tat->max; i++)
|
851
|
+
printf("%d ", tat->items[i] & ~INT_MIN);
|
852
|
+
printf("\n"); return; /* print the items in the */
|
853
|
+
} /* (rest of) the transaction */
|
854
|
+
vec = (TATREE**)(tat->items +tat->size);
|
855
|
+
for (i = 0; i < tat->size; i++) {
|
856
|
+
if (i > 0) for (k = ind; --k >= 0; ) printf(" ");
|
857
|
+
printf("%d ", tat->items[i]);
|
858
|
+
_show(vec[i], ind+1); /* traverse the items, print them, */
|
859
|
+
} /* and show the children recursively */
|
860
|
+
} /* _show() */
|
861
|
+
|
862
|
+
/*--------------------------------------------------------------------*/
|
863
|
+
|
864
|
+
void tat_show (TATREE *tat)
|
865
|
+
{ /* --- show a transaction tree */
|
866
|
+
assert(tat); /* check the function argument */
|
867
|
+
_show(tat, 0); /* just call the recursive function */
|
868
|
+
} /* tat_show() */
|
869
|
+
|
870
|
+
#endif
|