diff --git a/db/TODO b/db/TODO
index 1cd6194d913d30e1a25ab6efe7b347dc11802f07..a40e07d047ac75c9b3414e2917c628789056a414 100644
--- a/db/TODO
+++ b/db/TODO
@@ -2,3 +2,11 @@ Enter new table INPUT_JOIN1 into db_doc.md
Also work on documentation of views.
And of data files, which were split from main create_db.sql
+
+New view BENCH_COST_MD.
+New table BENCH_COST.
+Columns TC_COST etc. removed from table INPUT_GRAPHS
+
+Probably, it would be possible new to merge TCFF_PLOT_READ and SGFF_PLOT_REAL
+and so on (the new table BENCH_COST is more general than previous distinct
+columns).
diff --git a/db/bench_cost.sql b/db/bench_cost.sql
new file mode 100644
index 0000000000000000000000000000000000000000..bc55e99cfe15c4d10b9e6a6f98ffe7b35be948cd
--- /dev/null
+++ b/db/bench_cost.sql
@@ -0,0 +1,165 @@
+-- ============================================================================
+-- Project: rbench - Logic Programming and Database Benchmarks
+-- Filename: db/bench_cost.sql
+-- Purpose: Database for Benchmark Runs: Cost Measure Data for Benchmarks
+-- Last Change: 04.10.2019
+-- Language: SQL (Tested with PostgreSQL)
+-- Author: Stefan Brass
+-- EMail: brass@informatik.uni-halle.de
+-- WWW: http://www.informatik.uni-halle.de/~brass/
+-- Address: Feldschloesschen 15, D-06120 Halle (Saale), GERMANY
+-- Copyright: (c) 2019 by Stefan Brass
+-- License: See file "LICENSE" for copying conditions.
+-- Note: There is no warranty at all - this code may contain bugs.
+-- ============================================================================
+
+
+-- ============================================================================
+-- Cost Measures for Benchmarks (for Analysis of Runtimes):
+-- ============================================================================
+
+-- CREATE TABLE BENCH_COST (
+-- BENCH VARCHAR(20) NOT NULL,
+-- FILE_ID VARCHAR(10) NOT NULL,
+-- SIZE NUMERIC(12) NOT NULL,
+-- ITER NUMERIC(12) NOT NULL,
+-- INST NUMERIC(12) NOT NULL,
+-- CONSTRAINT BENCH_COST_PK
+-- PRIMARY KEY(BENCH, FILE_ID),
+-- CONSTRAINT BENCH_COST_REF_BENCHMARK
+-- FOREIGN KEY (BENCH) REFERENCES BENCHMARK,
+-- CONSTRAINT BENCH_COST_REF_INPUT_FILE
+-- FOREIGN KEY(FILE_ID) REFERENCES INPUT_FILE,
+-- CONSTRAINT BENCH_COST_SIZE_NONNEGATIVE
+-- CHECK(SIZE >= 0),
+-- CONSTRAINT BENCH_COST_ITER_NONNEGATIVE
+-- CHECK(ITER >= 0),
+-- CONSTRAINT BENCH_COST_INST_NONNEGATIVE
+-- CHECK(INST >= 0));
+--
+-- SIZE is the number of derived tuples of the benchmark (result size).
+-- INST is the number of applicable rule instances in the benchmark program.
+-- This is also the number of rule instances that semi-naive evaluation would
+-- consider.
+-- ITER is the number of iterations that the T_P operator needs to compute
+-- the entire minimal model (counting only iterations that derive new tuples).
+-- The application of the non-recursive rules is the first iteration
+-- (if they derive at least one tuple, otherwise it does not count as an
+-- iteration).
+
+
+-- ============================================================================
+-- INSERT Statements for Input Graph Data:
+-- ============================================================================
+
+INSERT INTO BENCH_COST VALUES('tcff','k50',2500,1,127500);
+INSERT INTO BENCH_COST VALUES('tcff','k100',10000,1,1010000);
+INSERT INTO BENCH_COST VALUES('tcff','k500',250000,1,125250000);
+INSERT INTO BENCH_COST VALUES('tcff','k1k',1000000,1,1001000000);
+INSERT INTO BENCH_COST VALUES('tcff','k2k',4000000,1,8004000000);
+INSERT INTO BENCH_COST VALUES('tcff','t50',1225,1,20825);
+INSERT INTO BENCH_COST VALUES('tcff','t100',4950,1,166650);
+INSERT INTO BENCH_COST VALUES('tcff','t500',124750,1,20833250);
+INSERT INTO BENCH_COST VALUES('tcff','t1k',499500,1,166666500);
+INSERT INTO BENCH_COST VALUES('tcff','t2k',1999000,1,1333333000);
+INSERT INTO BENCH_COST VALUES('tcff','c1k',1000000,1000,1001000);
+INSERT INTO BENCH_COST VALUES('tcff','c2k',4000000,2000,4002000);
+INSERT INTO BENCH_COST VALUES('tcff','c3k',9000000,3000,9003000);
+INSERT INTO BENCH_COST VALUES('tcff','c4k',16000000,4000,16004000);
+INSERT INTO BENCH_COST VALUES('tcff','s1k_1',1000000,500,2002000);
+INSERT INTO BENCH_COST VALUES('tcff','s1k_2',1000000,201,3003000);
+INSERT INTO BENCH_COST VALUES('tcff','s1k_3',1000000,250,4004000);
+INSERT INTO BENCH_COST VALUES('tcff','s1k_4',1000000,200,5005000);
+INSERT INTO BENCH_COST VALUES('tcff','s1k_5',1000000,42,6006000);
+INSERT INTO BENCH_COST VALUES('tcff','s2k_1',4000000,1000,8004000);
+INSERT INTO BENCH_COST VALUES('tcff','s2k_2',4000000,288,12006000);
+INSERT INTO BENCH_COST VALUES('tcff','s2k_3',4000000,500,16008000);
+INSERT INTO BENCH_COST VALUES('tcff','s2k_4',4000000,400,20010000);
+INSERT INTO BENCH_COST VALUES('tcff','s2k_5',4000000,127,24012000);
+INSERT INTO BENCH_COST VALUES('tcff','p1k',499500,999,499500);
+INSERT INTO BENCH_COST VALUES('tcff','p2k',1999000,1999,1999000);
+INSERT INTO BENCH_COST VALUES('tcff','p3k',4498500,2999,4498500);
+INSERT INTO BENCH_COST VALUES('tcff','p4k',7998000,3999,7998000);
+INSERT INTO BENCH_COST VALUES('tcff','m4_2ki',12288,3,12288);
+INSERT INTO BENCH_COST VALUES('tcff','m16_512',61440,15,61440);
+INSERT INTO BENCH_COST VALUES('tcff','m64_128',258048,63,258048);
+INSERT INTO BENCH_COST VALUES('tcff','m256_32',1044480,255,1044480);
+INSERT INTO BENCH_COST VALUES('tcff','m1ki_8',4190208,1023,4190208);
+INSERT INTO BENCH_COST VALUES('tcff','m4ki_2',16773120,4095,16773120);
+INSERT INTO BENCH_COST VALUES('tcff','b17',1966082,16,1966082);
+INSERT INTO BENCH_COST VALUES('tcff','b18',4194306,17,4194306);
+INSERT INTO BENCH_COST VALUES('tcff','b19',8912898,18,8912898);
+INSERT INTO BENCH_COST VALUES('tcff','v10',8194,9,8194);
+INSERT INTO BENCH_COST VALUES('tcff','v11',18434,10,18434);
+INSERT INTO BENCH_COST VALUES('tcff','v12',40962,11,40962);
+INSERT INTO BENCH_COST VALUES('tcff','v17',1966082,16,1966082);
+INSERT INTO BENCH_COST VALUES('tcff','v18',4194306,17,4194306);
+INSERT INTO BENCH_COST VALUES('tcff','v19',8912898,18,8912898);
+INSERT INTO BENCH_COST VALUES('tcff','y500_4k',9998000,4000,9998000);
+INSERT INTO BENCH_COST VALUES('tcff','y500_8k',35996000,8000,35996000);
+INSERT INTO BENCH_COST VALUES('tcff','y1k_4k',11998000,4000,11998000);
+INSERT INTO BENCH_COST VALUES('tcff','y1k_8k',39996000,8000,39996000);
+INSERT INTO BENCH_COST VALUES('tcff','u1k_50k',1000000,3,50050000);
+INSERT INTO BENCH_COST VALUES('tcff','u1k_125k',1000000,2,125125000);
+INSERT INTO BENCH_COST VALUES('tcff','u1k_250k',1000000,2,250250000);
+INSERT INTO BENCH_COST VALUES('tcff','u2k_200k',4000000,3,400200000);
+INSERT INTO BENCH_COST VALUES('tcff','u2k_500k',4000000,2,1000500000);
+INSERT INTO BENCH_COST VALUES('tcff','u2k_1m',4000000,2,2001000000);
+INSERT INTO BENCH_COST VALUES('tcff','a1k_50k',471621,8,15269270);
+INSERT INTO BENCH_COST VALUES('tcff','a1k_125k',492680,5,40853746);
+INSERT INTO BENCH_COST VALUES('tcff','a1k_250k',497763,4,82941061);
+INSERT INTO BENCH_COST VALUES('tcff','a2k_200k',1944729,8,128420988);
+INSERT INTO BENCH_COST VALUES('tcff','a2k_500k',1985321,6,330666855);
+INSERT INTO BENCH_COST VALUES('tcff','a2k_1m',1995698,4,666033613);
+INSERT INTO BENCH_COST VALUES('tcff','w1k_1k',1000000,1,1000000);
+INSERT INTO BENCH_COST VALUES('tcff','x10k',100020000,2,100020000);
+
+INSERT INTO BENCH_COST VALUES('sgff','k50',2500,2,6252500);
+INSERT INTO BENCH_COST VALUES('sgff','k100',10000,2,100010000);
+INSERT INTO BENCH_COST VALUES('sgff','k500',250000,2,62500250000);
+INSERT INTO BENCH_COST VALUES('sgff','k1k',1000000,2,1000001000000);
+INSERT INTO BENCH_COST VALUES('sgff','k2k',4000000,2,16000004000000);
+INSERT INTO BENCH_COST VALUES('sgff','t50',2402,2,1386602);
+INSERT INTO BENCH_COST VALUES('sgff','t100',9802,2,23546952);
+INSERT INTO BENCH_COST VALUES('sgff','t500',249002,2,15438684752);
+INSERT INTO BENCH_COST VALUES('sgff','t1k',998002,2,248504744502);
+INSERT INTO BENCH_COST VALUES('sgff','t2k',3996002,2,3988018989002);
+INSERT INTO BENCH_COST VALUES('sgff','c1k',1000,1,2000);
+INSERT INTO BENCH_COST VALUES('sgff','c2k',2000,1,4000);
+INSERT INTO BENCH_COST VALUES('sgff','c3k',3000,1,6000);
+INSERT INTO BENCH_COST VALUES('sgff','c4k',4000,1,8000);
+INSERT INTO BENCH_COST VALUES('sgff','s1k_1',1000000,501,4002000);
+INSERT INTO BENCH_COST VALUES('sgff','s1k_4',1000000,101,25005000);
+INSERT INTO BENCH_COST VALUES('sgff','s2k_1',4000000,1001,16004000);
+INSERT INTO BENCH_COST VALUES('sgff','s2k_3',4000000,251,64008000);
+INSERT INTO BENCH_COST VALUES('sgff','s2k_4',4000000,201,100010000);
+INSERT INTO BENCH_COST VALUES('sgff','p1k',1000,2,1998);
+INSERT INTO BENCH_COST VALUES('sgff','p2k',2000,2,3998);
+INSERT INTO BENCH_COST VALUES('sgff','p3k',3000,2,5998);
+INSERT INTO BENCH_COST VALUES('sgff','p4k',4000,2,7998);
+INSERT INTO BENCH_COST VALUES('sgff','m4_2ki',8192,2,12288);
+INSERT INTO BENCH_COST VALUES('sgff','m16_512',8192,2,15360);
+INSERT INTO BENCH_COST VALUES('sgff','m64_128',8192,2,16128);
+INSERT INTO BENCH_COST VALUES('sgff','m256_32',8192,2,16320);
+INSERT INTO BENCH_COST VALUES('sgff','m1ki_8',8192,2,16368);
+INSERT INTO BENCH_COST VALUES('sgff','m4ki_2',8192,2,16380);
+INSERT INTO BENCH_COST VALUES('sgff','b17',131071,2,262140);
+INSERT INTO BENCH_COST VALUES('sgff','b18',262143,2,524284);
+INSERT INTO BENCH_COST VALUES('sgff','b19',524287,2,1048572);
+INSERT INTO BENCH_COST VALUES('sgff','v10',349525,9,350546);
+INSERT INTO BENCH_COST VALUES('sgff','v11',1398101,10,1400146);
+INSERT INTO BENCH_COST VALUES('sgff','v12',5592405,11,5596498);
+INSERT INTO BENCH_COST VALUES('sgff','v17',5726623061,16,5726754130);
+INSERT INTO BENCH_COST VALUES('sgff','v18',22906492245,17,22906754386);
+INSERT INTO BENCH_COST VALUES('sgff','v19',91625968981,18,91626493266);
+INSERT INTO BENCH_COST VALUES('sgff','y500_4k',254000,2,258498);
+INSERT INTO BENCH_COST VALUES('sgff','y500_8k',258000,2,266498);
+INSERT INTO BENCH_COST VALUES('sgff','y1k_4k',1004000,2,1008998);
+INSERT INTO BENCH_COST VALUES('sgff','y1k_8k',1008000,2,1016998);
+
+
+-- ============================================================================
+-- Automatically Added Benchmark Cost Data:
+-- ============================================================================
+
+-- DO NOT DELETE THIS MARK
diff --git a/db/create_db.sql b/db/create_db.sql
index c2b39d1ee8f1e6cbddcecbb820b2ca1eb0de5dd3..911e01a50014eadaa03833d2dfbae297d6749733 100644
--- a/db/create_db.sql
+++ b/db/create_db.sql
@@ -403,12 +403,6 @@ CREATE TABLE INPUT_GRAPH (
MIN_OUT_DEGREE NUMERIC(12) NOT NULL,
MAX_OUT_DEGREE NUMERIC(12) NOT NULL,
CYCLES CHAR(1) NOT NULL,
- TC_SIZE NUMERIC(12) NOT NULL,
- TC_ITER NUMERIC(12) NOT NULL,
- TC_COST NUMERIC(12) NOT NULL,
- SG_SIZE NUMERIC(12) NULL,
- SG_ITER NUMERIC(12) NULL,
- SG_COST NUMERIC(16) NULL,
CONSTRAINT INPUT_GRAPH_PK
PRIMARY KEY(FILE_ID),
CONSTRAINT INPUT_GRAPH_REF_INPUT_FILE
@@ -438,19 +432,7 @@ CREATE TABLE INPUT_GRAPH (
CONSTRAINT INPUT_GRAPH_MAX_OUT_DEGREE_NONNEGATIVE
CHECK(MAX_OUT_DEGREE >= 0),
CONSTRAINT INPUT_GRAPH_CYCLES_YES_NO
- CHECK(CYCLES IN ('Y','N')),
- CONSTRAINT INPUT_GRAPH_TC_SIZE_NONNEGATIVE
- CHECK(TC_SIZE >= NUM_EDGES),
- CONSTRAINT INPUT_GRAPH_TC_ITER_NONNEGATIVE
- CHECK(TC_ITER >= 0),
- CONSTRAINT INPUT_GRAPH_TC_COST_NONNEGATIVE
- CHECK(TC_COST >= 0),
- CONSTRAINT INPUT_GRAPH_SG_SIZE_NONNEGATIVE
- CHECK(SG_SIZE >= 0),
- CONSTRAINT INPUT_GRAPH_SG_ITER_NONNEGATIVE
- CHECK(SG_ITER >= 0),
- CONSTRAINT INPUT_GRAPH_SG_COST_NONNEGATIVE
- CHECK(SG_COST >= 0));
+ CHECK(CYCLES IN ('Y','N')));
-- GTYPE, PARAM_1, PARAM_2, and PARAM_3 are the parameters for the
@@ -467,31 +449,6 @@ CREATE TABLE INPUT_GRAPH (
-- LOOPS='Y' means that there are edges of the form (x,x).
--- TC_SIZE, TC_ITER, and TC_COST are for the "Transitive Closure" benchmark:
--- tc(X, Y) :- par(X, Y).
--- tc(X, Z) :- par(X, Y), tc(Y, Z).
--- This is the standard tail-recursive Prolog program for computing the
--- transitive closure of the par relation (all pairs of nodes that are
--- connected by a path of "par"-edges).
--- TC_SIZE is the number of tuples in the derived "tc" relation.
--- TC_COST is the number of applicable rule instances in the above program.
--- This is also the number of rule instances that semi-naive evaluation would
--- consider.
--- It can be computed as the sum of the number of edges
--- (i.e. the size of the par relation) and
--- the size of the join of par and tc.
-
--- SG_SIZE, SG_ITER, and SG_COST are for the "Same Generation Cousins"
--- benchmark:
--- sg(_,Y) :- par(_, Y).
--- sg(X1, Y1) :- par(X1, X), sg(X, Y), par(Y1, Y).
--- SG_SIZE is the size of the so computed sg relation.
--- SG_COST is the number of applicable rule instances in the above program.
--- SG_ITER is the number of iterations that the T_P operator needs to compute
--- the entire sg relation (counting only iterations that derive new tuples).
--- The application of the non-recursive rule is the first iteration
--- (if par is non-empty).
-
-- The data is in file input_graphs.sql
@@ -512,10 +469,10 @@ CREATE TABLE INPUT_JOIN1 (
B1_SIZE NUMERIC(12) NOT NULL,
B2_SIZE NUMERIC(12) NOT NULL,
C1_SIZE NUMERIC(12) NOT NULL,
- A_COST NUMERIC(12) NOT NULL,
- B1_COST NUMERIC(12) NOT NULL,
- B2_COST NUMERIC(12) NOT NULL,
- C1_COST NUMERIC(12) NOT NULL,
+ A_INST NUMERIC(12) NOT NULL,
+ B1_INST NUMERIC(12) NOT NULL,
+ B2_INST NUMERIC(12) NOT NULL,
+ C1_INST NUMERIC(12) NOT NULL,
CONSTRAINT INPUT_JOIN1_PK
PRIMARY KEY(FILE_ID),
CONSTRAINT INPUT_JOIN1_REF_INPUT_FILE
@@ -540,22 +497,22 @@ CREATE TABLE INPUT_JOIN1 (
CHECK(D1_SIZE >= 0),
CONSTRAINT INPUT_JOIN1_D2_SIZE_NONNEGATIVE
CHECK(D2_SIZE >= 0),
- CONSTRAINT INPUT_JOIN1_A_COST_NONNEGATIVE
- CHECK(A_COST >= 0),
- CONSTRAINT INPUT_JOIN1_B1_COST_NONNEGATIVE
- CHECK(B1_COST >= 0),
- CONSTRAINT INPUT_JOIN1_B2_COST_NONNEGATIVE
- CHECK(B2_COST >= 0),
- CONSTRAINT INPUT_JOIN1_C1_COST_NONNEGATIVE
- CHECK(C1_COST >= 0),
- CONSTRAINT INPUT_JOIN1_A_COST_GE_A_SIZE
- CHECK(A_COST >= A_SIZE),
- CONSTRAINT INPUT_JOIN1_B1_COST_GE_B1_SIZE
- CHECK(B1_COST >= B1_SIZE),
- CONSTRAINT INPUT_JOIN1_B2_COST_GE_B2_SIZE
- CHECK(B2_COST >= B2_SIZE),
- CONSTRAINT INPUT_JOIN1_C1_COST_GE_C1_SIZE
- CHECK(C1_COST >= C1_SIZE));
+ CONSTRAINT INPUT_JOIN1_A_INST_NONNEGATIVE
+ CHECK(A_INST >= 0),
+ CONSTRAINT INPUT_JOIN1_B1_INST_NONNEGATIVE
+ CHECK(B1_INST >= 0),
+ CONSTRAINT INPUT_JOIN1_B2_INST_NONNEGATIVE
+ CHECK(B2_INST >= 0),
+ CONSTRAINT INPUT_JOIN1_C1_INST_NONNEGATIVE
+ CHECK(C1_INST >= 0),
+ CONSTRAINT INPUT_JOIN1_A_INST_GE_A_SIZE
+ CHECK(A_INST >= A_SIZE),
+ CONSTRAINT INPUT_JOIN1_B1_INST_GE_B1_SIZE
+ CHECK(B1_INST >= B1_SIZE),
+ CONSTRAINT INPUT_JOIN1_B2_INST_GE_B2_SIZE
+ CHECK(B2_INST >= B2_SIZE),
+ CONSTRAINT INPUT_JOIN1_C1_INST_GE_C1_SIZE
+ CHECK(C1_INST >= C1_SIZE));
-- A_SIZE and so on are the number of facts for the predicate.
-- The input file contains the predicates D1, D2, C2, C3, C4.
@@ -565,10 +522,46 @@ CREATE TABLE INPUT_JOIN1 (
-- b2(X, Y) :- c3(X, Z), c4(Z, Y).
-- c1(X, Y) :- d1(X, Z), d2(Z, Y).
--- A_COST and so on are the number of instances of the rule that derives A.
+-- A_INST and so on are the number of instances of the rule that derives A.
-- The data is in file input_join1.sql
+-- ============================================================================
+-- Cost Measures for Benchmarks (for Analysis of Runtimes):
+-- ============================================================================
+
+CREATE TABLE BENCH_COST (
+ BENCH VARCHAR(20) NOT NULL,
+ FILE_ID VARCHAR(10) NOT NULL,
+ SIZE NUMERIC(12) NOT NULL,
+ ITER NUMERIC(12) NOT NULL,
+ INST NUMERIC(12) NOT NULL,
+ CONSTRAINT BENCH_COST_PK
+ PRIMARY KEY(BENCH, FILE_ID),
+ CONSTRAINT BENCH_COST_REF_BENCHMARK
+ FOREIGN KEY (BENCH) REFERENCES BENCHMARK,
+ CONSTRAINT BENCH_COST_REF_INPUT_FILE
+ FOREIGN KEY(FILE_ID) REFERENCES INPUT_FILE,
+ CONSTRAINT BENCH_COST_SIZE_NONNEGATIVE
+ CHECK(SIZE >= 0),
+ CONSTRAINT BENCH_COST_ITER_NONNEGATIVE
+ CHECK(ITER >= 0),
+ CONSTRAINT BENCH_COST_INST_NONNEGATIVE
+ CHECK(INST >= 0));
+
+-- SIZE is the number of derived tuples of the benchmark (result size).
+-- INST is the number of applicable rule instances in the benchmark program.
+-- This is also the number of rule instances that semi-naive evaluation would
+-- consider.
+-- ITER is the number of iterations that the T_P operator needs to compute
+-- the entire minimal model (counting only iterations that derive new tuples).
+-- The application of the non-recursive rules is the first iteration
+-- (if they derive at least one tuple, otherwise it does not count as an
+-- iteration).
+
+-- The data is in file bench_cost.sql
+
+
-- ============================================================================
-- Table for Results of Benchmark Runs (Test Runs):
-- ============================================================================
diff --git a/db/drop_db.sql b/db/drop_db.sql
index 4511220db3534848857655c2511c08f332e94375..66224b714b399f82493089af2b3facc246b2528c 100644
--- a/db/drop_db.sql
+++ b/db/drop_db.sql
@@ -2,7 +2,7 @@
-- Project: ddbench - Logic Programming and Database Benchmarks
-- Filename: db/drop_db.sql
-- Purpose: Database for Benchmark Runs: Drop all Tables and Views
--- Last Change: 28.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -39,6 +39,7 @@ DROP VIEW BENCH_FILE_LIST;
-- Views to Print Table Data in Markdown Format for Docs (File v_doc.sql):
-- ============================================================================
+DROP VIEW BENCH_COST_MD;
DROP VIEW INPUT_GRAPH_MD;
DROP VIEW INPUT_FILE_MD;
DROP VIEW BENCH_IMPL_MD;
@@ -190,6 +191,7 @@ DROP TABLE DUMMY_TAB;
DROP TABLE BENCH_RUN;
DROP TABLE BENCH_IMPL;
+DROP TABLE BENCH_COST;
DROP TABLE INPUT_JOIN1;
DROP TABLE INPUT_GRAPH;
DROP TABLE GRAPH_TYPE;
diff --git a/db/input_graphs.sql b/db/input_graphs.sql
index 9770a79aee6860e76f8571ca6245143f34882ab0..2c8fe1b60453a29c66e49d282ab201a020e4463d 100644
--- a/db/input_graphs.sql
+++ b/db/input_graphs.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/input_graphs.sql
-- Purpose: Database for Benchmark Runs: Data of Input Graphs
--- Last Change: 16.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -33,12 +33,6 @@
-- MIN_OUT_DEGREE NUMERIC(12) NOT NULL,
-- MAX_OUT_DEGREE NUMERIC(12) NOT NULL,
-- CYCLES CHAR(1) NOT NULL,
--- TC_SIZE NUMERIC(12) NOT NULL,
--- TC_ITER NUMERIC(12) NOT NULL,
--- TC_COST NUMERIC(12) NOT NULL,
--- SG_SIZE NUMERIC(12) NULL,
--- SG_ITER NUMERIC(12) NULL,
--- SG_COST NUMERIC(12) NULL,
-- CONSTRAINT INPUT_GRAPH_PK
-- PRIMARY KEY(FILE_ID),
-- CONSTRAINT INPUT_GRAPH_REF_INPUT_FILE
@@ -68,19 +62,7 @@
-- CONSTRAINT INPUT_GRAPH_MAX_OUT_DEGREE_NONNEGATIVE
-- CHECK(MAX_OUT_DEGREE >= 0),
-- CONSTRAINT INPUT_GRAPH_CYCLES_YES_NO
--- CHECK(CYCLES IN ('Y','N')),
--- CONSTRAINT INPUT_GRAPH_TC_SIZE_NONNEGATIVE
--- CHECK(TC_SIZE >= NUM_EDGES),
--- CONSTRAINT INPUT_GRAPH_TC_ITER_NONNEGATIVE
--- CHECK(TC_ITER >= 0),
--- CONSTRAINT INPUT_GRAPH_TC_COST_NONNEGATIVE
--- CHECK(TC_COST >= 0),
--- CONSTRAINT INPUT_GRAPH_SG_SIZE_NONNEGATIVE
--- CHECK(SG_SIZE >= 0),
--- CONSTRAINT INPUT_GRAPH_SG_ITER_NONNEGATIVE
--- CHECK(SG_ITER >= 0),
--- CONSTRAINT INPUT_GRAPH_SG_COST_NONNEGATIVE
--- CHECK(SG_COST >= 0));
+-- CHECK(CYCLES IN ('Y','N'));
-- GTYPE, PARAM_1, PARAM_2, and PARAM_3 are the parameters for the
@@ -97,224 +79,197 @@
-- LOOPS='Y' means that there are edges of the form (x,x).
--- TC_SIZE, TC_ITER, and TC_COST are for the "Transitive Closure" benchmark:
--- tc(X, Y) :- par(X, Y).
--- tc(X, Z) :- par(X, Y), tc(Y, Z).
--- This is the standard tail-recursive Prolog program for computing the
--- transitive closure of the par relation (all pairs of nodes that are
--- connected by a path of "par"-edges).
--- TC_SIZE is the number of tuples in the derived "tc" relation.
--- TC_COST is the number of applicable rule instances in the above program.
--- This is also the number of rule instances that semi-naive evaluation would
--- consider.
--- It can be computed as the sum of the number of edges
--- (i.e. the size of the par relation) and
--- the size of the join of par and tc.
-
--- SG_SIZE, SG_ITER, and SG_COST are for the "Same Generation Cousins"
--- benchmark:
--- sg(_,Y) :- par(_, Y).
--- sg(X1, Y1) :- par(X1, X), sg(X, Y), par(Y1, Y).
--- SG_SIZE is the size of the so computed sg relation.
--- SG_COST is the number of applicable rule instances in the above program.
--- SG_ITER is the number of iterations that the T_P operator needs to compute
--- the entire sg relation (counting only iterations that derive new tuples).
--- The application of the non-recursive rule is the first iteration
--- (if par is non-empty).
-
-
-- ============================================================================
-- INSERT Statements for Input Graph Data:
-- ============================================================================
INSERT INTO INPUT_GRAPH VALUES('b17','B',17,NULL, NULL,
-131071,131070,0,'N',0,1,0,2,'N',1966082,16,1966082,131071,2,262140);
+131071,131070,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('b18','B',18,NULL, NULL,
-262143,262142,0,'N',0,1,0,2,'N',4194306,17,4194306,262143,2,524284);
+262143,262142,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('b19','B',19,NULL, NULL,
-524287,524286,0,'N',0,1,0,2,'N',8912898,18,8912898,524287,2,1048572);
+524287,524286,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v10','B',10,NULL,'I',
-1023,1022,0,'N',0,1,0,2,'N',8194,9,8194,349525,9,350546);
+1023,1022,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v11','B',11,NULL,'I',
-2047,2046,0,'N',0,1,0,2,'N',18434,10,18434,1398101,10,1400146);
+2047,2046,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v12','B',12,NULL,'I',
-4095,4094,0,'N',0,1,0,2,'N',40962,11,40962,5592405,11,5596498);
+4095,4094,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v17','B',17,NULL,'I',
-131071,131070,0,'N',0,1,0,2,'N',1966082,16,1966082,5726623061,16,5726754130);
+131071,131070,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v18','B',18,NULL,'I',
-262143,262142,0,'N',0,1,0,2,'N',4194306,17,4194306,22906492245,17,22906754386);
+262143,262142,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('v19','B',19,NULL,'I',
-524287,524286,0,'N',0,1,0,2,'N',8912898,18,8912898,91625968981,18,91626493266);
+524287,524286,0,'N',0,1,0,2,'N');
INSERT INTO INPUT_GRAPH VALUES('c1k','C',1000,NULL, NULL,
-1000,1000,0,'N',1,1,1,1,'Y',1000000,1000,1001000,1000,1,2000);
+1000,1000,0,'N',1,1,1,1,'Y');
INSERT INTO INPUT_GRAPH VALUES('c2k','C',2000,NULL, NULL,
-2000,2000,0,'N',1,1,1,1,'Y',4000000,2000,4002000,2000,1,4000);
+2000,2000,0,'N',1,1,1,1,'Y');
INSERT INTO INPUT_GRAPH VALUES('c3k','C',3000,NULL, NULL,
-3000,3000,0,'N',1,1,1,1,'Y',9000000,3000,9003000,3000,1,6000);
+3000,3000,0,'N',1,1,1,1,'Y');
INSERT INTO INPUT_GRAPH VALUES('c4k','C',4000,NULL, NULL,
-4000,4000,0,'N',1,1,1,1,'Y',16000000,4000,16004000,4000,1,8000);
+4000,4000,0,'N',1,1,1,1,'Y');
INSERT INTO INPUT_GRAPH VALUES('k50','K',50,NULL, NULL,
-50,2500,0,'Y',50,50,50,50,'Y',2500,1,127500,2500,2,6252500);
+50,2500,0,'Y',50,50,50,50,'Y');
INSERT INTO INPUT_GRAPH VALUES('k100','K',100,NULL, NULL,
-100,10000,0,'Y',100,100,100,100,'Y',10000,1,1010000,10000,2,100010000);
+100,10000,0,'Y',100,100,100,100,'Y');
INSERT INTO INPUT_GRAPH VALUES('k500','K',500,NULL, NULL,
-500,250000,0,'Y',500,500,500,500,'Y',250000,1,125250000,250000,2,62500250000);
+500,250000,0,'Y',500,500,500,500,'Y');
INSERT INTO INPUT_GRAPH VALUES('k1k','K',1000,NULL, NULL,
-1000,1000000,0,'Y',1000,1000,1000,1000,'Y',1000000,1,1001000000,1000000,2,1000001000000);
+1000,1000000,0,'Y',1000,1000,1000,1000,'Y');
INSERT INTO INPUT_GRAPH VALUES('k2k','K',2000,NULL, NULL,
-2000,4000000,0,'Y',2000,2000,2000,2000,'Y',4000000,1,8004000000,4000000,2,16000004000000);
+2000,4000000,0,'Y',2000,2000,2000,2000,'Y');
INSERT INTO INPUT_GRAPH VALUES('t50','T',50,NULL, NULL,
-50,1225,0,'N',0,49,0,49,'N',1225,1,20825,2402,2,1386602);
+50,1225,0,'N',0,49,0,49,'N');
INSERT INTO INPUT_GRAPH VALUES('t100','T',100,NULL, NULL,
-100,4950,0,'N',0,99,0,99,'N',4950,1,166650,9802,2,23546952);
+100,4950,0,'N',0,99,0,99,'N');
INSERT INTO INPUT_GRAPH VALUES('t500','T',500,NULL, NULL,
-500,124750,0,'N',0,499,0,499,'N',124750,1,20833250,249002,2,15438684752);
+500,124750,0,'N',0,499,0,499,'N');
INSERT INTO INPUT_GRAPH VALUES('t1k','T',1000,NULL, NULL,
-1000,499500,0,'N',0,999,0,999,'N',499500,1,166666500,998002,2,248504744502);
+1000,499500,0,'N',0,999,0,999,'N');
INSERT INTO INPUT_GRAPH VALUES('t2k','T',2000,NULL, NULL,
-2000,1999000,0,'N',0,1999,0,1999,'N',1999000,1,1333333000,3996002,2,3988018989002);
+2000,1999000,0,'N',0,1999,0,1999,'N');
INSERT INTO INPUT_GRAPH VALUES('m4ki_2','M',4096,2, NULL,
-8192,8190,0,'N',0,1,0,1,'N',16773120,4095,16773120,8192,2,16380);
+8192,8190,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('m1ki_8','M',1024,8, NULL,
-8192,8184,0,'N',0,1,0,1,'N',4190208,1023,4190208,8192,2,16368);
+8192,8184,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('m256_32','M',256,32, NULL,
-8192,8160,0,'N',0,1,0,1,'N',1044480,255,1044480,8192,2,16320);
+8192,8160,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('m64_128','M',64,128, NULL,
-8192,8064,0,'N',0,1,0,1,'N',258048,63,258048,8192,2,16128);
+8192,8064,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('m16_512','M', 16, 512, NULL,
-8192,7680,0,'N',0,1,0,1,'N',61440,15,61440,8192,2,15360);
+8192,7680,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('m4_2ki','M', 4, 2048, NULL,
-8192,6144,0,'N',0,1,0,1,'N',12288,3,12288,8192,2,12288);
+8192,6144,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('p1k','P',1000,NULL, NULL,
-1000,999,0,'N',0,1,0,1,'N',499500,999,499500,1000,2,1998);
+1000,999,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('p2k','P',2000,NULL, NULL,
-2000,1999,0,'N',0,1,0,1,'N',1999000,1999,1999000,2000,2,3998);
+2000,1999,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('p3k','P',3000,NULL, NULL,
-3000,2999,0,'N',0,1,0,1,'N',4498500,2999,4498500,3000,2,5998);
+3000,2999,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('p4k','P',4000,NULL, NULL,
-4000,3999,0,'N',0,1,0,1,'N',7998000,3999,7998000,4000,2,7998);
+4000,3999,0,'N',0,1,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('s2k_1','S', 2000, 1, NULL,
-2000,4000,0,'N',2,2,2,2,'Y',4000000,1000,8004000,4000000,1001,16004000);
+2000,4000,0,'N',2,2,2,2,'Y');
INSERT INTO INPUT_GRAPH VALUES('s2k_2','S', 2000, 2, NULL,
-2000,6000,0,'N',3,3,3,3,'Y',4000000,288,12006000,4000000,NULL,36006000);
+2000,6000,0,'N',3,3,3,3,'Y');
INSERT INTO INPUT_GRAPH VALUES('s2k_3','S', 2000, 3, NULL,
-2000,8000,0,'N',4,4,4,4,'Y',4000000,500,16008000,4000000,251,64008000);
+2000,8000,0,'N',4,4,4,4,'Y');
INSERT INTO INPUT_GRAPH VALUES('s2k_4','S', 2000, 4, NULL,
-2000,10000,0,'N',5,5,5,5,'Y',4000000,400,20010000,4000000,201,100010000);
+2000,10000,0,'N',5,5,5,5,'Y');
INSERT INTO INPUT_GRAPH VALUES('s2k_5','S', 2000, 5, NULL,
-2000,12000,0,'N',6,6,6,6,'Y',4000000,127,24012000,4000000,NULL,144012000);
+2000,12000,0,'N',6,6,6,6,'Y');
INSERT INTO INPUT_GRAPH VALUES('s1k_1','S', 1000, 1, NULL,
-1000,2000,0,'N',2,2,2,2,'Y',1000000,500,2002000,1000000,501,4002000);
+1000,2000,0,'N',2,2,2,2,'Y');
INSERT INTO INPUT_GRAPH VALUES('s1k_2','S', 1000, 2, NULL,
-1000,3000,0,'N',3,3,3,3,'Y',1000000,201,3003000,1000000,NULL,9003000);
+1000,3000,0,'N',3,3,3,3,'Y');
INSERT INTO INPUT_GRAPH VALUES('s1k_3','S', 1000, 3, NULL,
-1000,4000,0,'N',4,4,4,4,'Y',1000000,250,4004000,1000000,NULL,16004000);
+1000,4000,0,'N',4,4,4,4,'Y');
INSERT INTO INPUT_GRAPH VALUES('s1k_4','S', 1000, 4, NULL,
-1000,5000,0,'N',5,5,5,5,'Y',1000000,200,5005000,1000000,101,25005000);
+1000,5000,0,'N',5,5,5,5,'Y');
INSERT INTO INPUT_GRAPH VALUES('s1k_5','S', 1000, 5, NULL,
-1000,6000,0,'N',6,6,6,6,'Y',1000000,42,6006000,1000000,NULL,36006000);
+1000,6000,0,'N',6,6,6,6,'Y');
INSERT INTO INPUT_GRAPH VALUES('u1k_50k','U', 1000, 50000, NULL,
-1000,50000,0,'N',46,51,46,51,'Y',1000000,3,50050000,1000000,NULL,NULL);
+1000,50000,0,'N',46,51,46,51,'Y');
INSERT INTO INPUT_GRAPH VALUES('u1k_125k','U', 1000, 125000, NULL,
-1000,125000,0,'N',122,127,122,127,'Y',1000000,2,125125000,NULL,NULL,NULL);
+1000,125000,0,'N',122,127,122,127,'Y');
INSERT INTO INPUT_GRAPH VALUES('u1k_250k','U', 1000, 250000, NULL,
-1000,250000,0,'N',248,251,248,251,'Y',1000000,2,250250000,NULL,NULL,NULL);
+1000,250000,0,'N',248,251,248,251,'Y');
INSERT INTO INPUT_GRAPH VALUES('u2k_200k','U', 2000, 200000, NULL,
-2000,200000,0,'N',98,101,98,101,'Y',4000000,3,400200000,NULL,NULL,NULL);
+2000,200000,0,'N',98,101,98,101,'Y');
INSERT INTO INPUT_GRAPH VALUES('u2k_500k','U', 2000, 500000, NULL,
-2000,500000,0,'N',248,251,248,251,'Y',4000000,2,1000500000,NULL,NULL,NULL);
+2000,500000,0,'N',248,251,248,251,'Y');
INSERT INTO INPUT_GRAPH VALUES('u2k_1m','U', 2000, 1000000, NULL,
-2000,1000000,0,'N',499,502,499,502,'Y',4000000,2,2001000000,NULL,NULL,NULL);
+2000,1000000,0,'N',499,502,499,502,'Y');
INSERT INTO INPUT_GRAPH VALUES('a1k_50k','A', 1000, 50000, NULL,
-1000,50000,0,'N',0,102,0,102,'N',471621,8,15269270,NULL,NULL,NULL);
+1000,50000,0,'N',0,102,0,102,'N');
INSERT INTO INPUT_GRAPH VALUES('a1k_125k','A', 1000, 125000, NULL,
-1000,125000,0,'N',0,250,0,250,'N',492680,5,40853746,NULL,NULL,NULL);
+1000,125000,0,'N',0,250,0,250,'N');
INSERT INTO INPUT_GRAPH VALUES('a1k_250k','A', 1000, 250000, NULL,
-1000,250000,0,'N',0,500,0,501,'N',497763,4,82941061,NULL,NULL,NULL);
+1000,250000,0,'N',0,500,0,501,'N');
INSERT INTO INPUT_GRAPH VALUES('a2k_200k','A', 2000, 200000, NULL,
-2000,200000,0,'N',0,201,0,201,'N',1944729,8,128420988,NULL,NULL,NULL);
+2000,200000,0,'N',0,201,0,201,'N');
INSERT INTO INPUT_GRAPH VALUES('a2k_500k','A', 2000, 500000, NULL,
-2000,500000,0,'N',0,500,0,500,'N',1985321,6,330666855,NULL,NULL,NULL);
+2000,500000,0,'N',0,500,0,500,'N');
INSERT INTO INPUT_GRAPH VALUES('a2k_1m','A', 2000, 1000000, NULL,
-2000,1000000,0,'N',0,1000,0,1000,'N',1995698,4,666033613,NULL,NULL,NULL);
+2000,1000000,0,'N',0,1000,0,1000,'N');
INSERT INTO INPUT_GRAPH VALUES('y1k_4k','Y', 1000, 4000, NULL,
-5000,4999,0,'N',0,1000,0,1,'N',11998000,4000,11998000,1004000,2,1008998);
+5000,4999,0,'N',0,1000,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('y1k_8k','Y', 1000, 8000, NULL,
-9000,8999,0,'N',0,1000,0,1,'N',39996000,8000,39996000,1008000,2,1016998);
+9000,8999,0,'N',0,1000,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('y500_4k','Y', 500, 4000, NULL,
-4500,4499,0,'N',0,500,0,1,'N',9998000,4000,9998000,254000,2,258498);
+4500,4499,0,'N',0,500,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('y500_8k','Y', 500, 8000, NULL,
-8500,8499,0,'N',0,500,0,1,'N',35996000,8000,35996000,258000,2,266498);
+8500,8499,0,'N',0,500,0,1,'N');
INSERT INTO INPUT_GRAPH VALUES('w1k_1k','W', 1000, 1000, NULL,
-2000,1000000,0,'N',0,1000,0,1000,'N',1000000,1,1000000,NULL,NULL,NULL);
+2000,1000000,0,'N',0,1000,0,1000,'N');
INSERT INTO INPUT_GRAPH VALUES('x10k','X', 10000, NULL, NULL,
-20001,20000,0,'N',0,10000,0,10000,'N',100020000,2,100020000,NULL,NULL,NULL);
+20001,20000,0,'N',0,10000,0,10000,'N');
-- INSERT INTO INPUT_GRAPH VALUES(FILE_ID, GTYPE, PARAM_1, PARAM_2, PARAM_3,
-- NUM_NODES, NUM_EDGES, DUP_EDGES, LOOPS, MIN_IN_DEGREE, MAX_IN_DEGREE,
--- MIN_OUT_DEGREE, MAX_OUT_DEGREE, CYCLES, TC_SIZE, TC_ITER, TC_COST,
--- SG_SIZE, SG_ITER, SG_COST);
+-- MIN_OUT_DEGREE, MAX_OUT_DEGREE, CYCLES);
-- ============================================================================
-- Automatically Added Input Graph Data:
diff --git a/db/v_doc.sql b/db/v_doc.sql
index 74c65469c03c27f5285b48788d151a7fca473292..3b7936089622421b1f4658da36d248a7bac2f11c 100644
--- a/db/v_doc.sql
+++ b/db/v_doc.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_doc.sql
-- Purpose: Benchmark DB: Views to Produce Table Data in Markdown Format
--- Last Change: 04.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -66,7 +66,7 @@ SELECT 0 AS SECTION, 0 AS SEQ_NO,
'DUP_EDGES|LOOPS|' ||
'MIN_IN_DEGREE|MAX_IN_DEGREE|' ||
'MIN_OUT_DEGREE|MAX_OUT_DEGREE|' ||
- 'CYCLES|TC_SIZE|TC_ITER|TC_COST|' AS MD
+ 'CYCLES|' AS MD
UNION ALL
SELECT 1 AS SECTION, 0 AS SEQ_NO,
'|-------|' ||
@@ -75,7 +75,7 @@ SELECT 1 AS SECTION, 0 AS SEQ_NO,
'--------:|:---:|' ||
'------------:|------------:|' ||
'-------------:|-------------:|' ||
- ':----:|------:|-------:|---:|' AS MD
+ ':----:|' AS MD
UNION ALL
SELECT 2 AS SECTION, F.SEQ_NO,
'|' || G.FILE_ID || '|' ||
@@ -88,9 +88,31 @@ SELECT 2 AS SECTION, F.SEQ_NO,
DUP_EDGES || '|' || LOOPS || '|' ||
MIN_IN_DEGREE || '|' || MAX_IN_DEGREE || '|' ||
MIN_OUT_DEGREE || '|' || MAX_OUT_DEGREE || '|' ||
- CYCLES || '|' || TC_SIZE || '|' || TC_ITER || '|' || TC_COST || '|'
+ CYCLES || '|' || '|'
AS MD
FROM INPUT_GRAPH G, INPUT_FILE F
WHERE G.FILE_ID = F.FILE_ID
ORDER BY SECTION, SEQ_NO;
+
+-- ============================================================================
+-- BENCH_COST_MD:
+-- ============================================================================
+
+CREATE VIEW BENCH_COST_MD AS
+SELECT 0 AS SECTION, 0 AS SEQ_NO,
+ '|BENCH|FILE_ID|' ||
+ 'SIZE|ITER|INST|' AS MD
+UNION ALL
+SELECT 1 AS SECTION, 0 AS SEQ_NO,
+ '|-------|--------|' ||
+ '------------:|------:|------------:|' AS MD
+UNION ALL
+SELECT 2 AS SECTION, F.SEQ_NO,
+ '|' || B.BENCH || '|' || B.FILE_ID || '|' ||
+ B.SIZE || '|' || B.ITER || '|' || B.INST || '|'
+ AS MD
+FROM BENCH_COST B, INPUT_FILE F
+WHERE B.FILE_ID = F.FILE_ID
+ORDER BY SECTION, SEQ_NO;
+
diff --git a/db/v_estimate.sql b/db/v_estimate.sql
index b26e0857019c50968620efda57398833d889f600..9c1a50d90f03721f9675a70215ae7a747a0b09e0 100644
--- a/db/v_estimate.sql
+++ b/db/v_estimate.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_estimate.sql
-- Purpose: Benchmark DB: Views to Support Research on Runtime Estimation
--- Last Change: 16.04.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -75,12 +75,13 @@ SELECT R.SYS, E.STARTUP, E.LOAD_IDX, E.RULE_APP, E.RESULT,
ROUND(CAST(
E.STARTUP +
(G.NUM_EDGES * (LOG(G.NUM_EDGES)/LOG(2)) * E.LOAD_IDX / 1000000000) +
- (G.TC_COST * E.RULE_APP / 1000000000) +
- (G.TC_SIZE * E.RESULT / 1000000000) AS NUMERIC),
+ (C.INST * E.RULE_APP / 1000000000) +
+ (C.SIZE * E.RESULT / 1000000000) AS NUMERIC),
3) AS ESTIMATE,
R.REAL_T
-FROM RESULT R, INPUT_GRAPH G, ESTIMATE_PAR E
+FROM RESULT R, INPUT_GRAPH G, BENCH_COST C, ESTIMATE_PAR E
WHERE R.SYS = E.SYS AND E.BENCH = 'tcff' AND R.BENCH = 'tcff'
+AND C.BENCH = 'tcff' AND C.FILE_ID = G.FILE_ID
AND R.FILE_ID = G.FILE_ID
ORDER BY R.SYS, E.STARTUP, E.LOAD_IDX, E.RULE_APP, E.RESULT, ESTIMATE;
@@ -117,12 +118,13 @@ SELECT R.SYS, R.BENCH, R.IMPL,
'data_c("' || G.FILE_ID || '", ' ||
G.NUM_NODES || ', ' ||
G.NUM_EDGES || ', ' ||
- 'INT64_C(' || G.TC_SIZE || '), ' ||
- 'INT64_C(' || G.TC_COST || '), ' ||
- G.TC_ITER || ', ' ||
+ 'INT64_C(' || C.SIZE || '), ' ||
+ 'INT64_C(' || C.INST || '), ' ||
+ C.ITER || ', ' ||
ROUND(R.REAL_T*1000)|| '),' AS OBJ
-FROM INPUT_GRAPH G, INPUT_FILE F, BEST_RESULT R
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C, BEST_RESULT R
WHERE G.FILE_ID = F.FILE_ID AND R.FILE_ID = F.FILE_ID
+AND C.FILE_ID = F.FILE_ID AND C.BENCH = 'tcff'
ORDER BY R.SYS, R.BENCH, F.SEQ_NO;
-- ============================================================================
@@ -137,12 +139,13 @@ SELECT R.SYS, R.BENCH, R.IMPL, F.FILE_ID, F.SEQ_NO,
G.FILE_ID || '","' ||
G.NUM_NODES || '","' ||
G.NUM_EDGES || '","' ||
- G.TC_SIZE || '","' ||
- G.TC_COST || '","' ||
- G.TC_ITER || '","' ||
+ C.SIZE || '","' ||
+ C.INST || '","' ||
+ C.ITER || '","' ||
ROUND(R.REAL_T*1000) || '"' AS CSV
-FROM INPUT_GRAPH G, INPUT_FILE F, BEST_RESULT R
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C, BEST_RESULT R
WHERE R.BENCH = 'tcff' AND G.FILE_ID = F.FILE_ID AND R.FILE_ID = F.FILE_ID
+AND C.BENCH = 'tcff' AND C.FILE_ID = F.FILE_ID
UNION ALL
SELECT R.SYS, R.BENCH, R.IMPL, F.FILE_ID, F.SEQ_NO,
'"' || R.SYS || '","' ||
@@ -151,12 +154,14 @@ SELECT R.SYS, R.BENCH, R.IMPL, F.FILE_ID, F.SEQ_NO,
G.FILE_ID || '","' ||
G.NUM_NODES || '","' ||
G.NUM_EDGES || '","' ||
- G.SG_SIZE || '","' ||
- G.SG_COST || '","' ||
- G.SG_ITER || '","' ||
+ C.SIZE || '","' ||
+ C.INST || '","' ||
+ C.ITER || '","' ||
ROUND(R.REAL_T*1000) || '"' AS CSV
-FROM INPUT_GRAPH G, INPUT_FILE F, BEST_RESULT R
-WHERE R.BENCH = 'sgff' AND G.FILE_ID = F.FILE_ID AND R.FILE_ID = F.FILE_ID
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C, BEST_RESULT R
+WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'sgff' AND C.FILE_ID = F.FILE_ID
UNION ALL
SELECT '' AS SYS, '' AS BENCH, '' AS IMPL, '' AS FILE_ID, 0 AS SEQ_NO,
'"' || 'SYS' || '","' ||
@@ -191,13 +196,14 @@ ORDER BY R.SYS, R.BENCH, F.SEQ_NO;
CREATE VIEW PAR_RANGE AS
SELECT R.SYS, G.FILE_ID,
ROUND(R.REAL_T * 1000 * 1000 / G.NUM_EDGES, 3) AS PAR_LOAD,
- ROUND(R.REAL_T * 1000 * 1000 / G.TC_COST, 3) AS PAR_RULE,
- ROUND(R.REAL_T * 1000 * 1000 / G.TC_SIZE, 3) AS PAR_RESULT,
+ ROUND(R.REAL_T * 1000 * 1000 / C.INST, 3) AS PAR_RULE,
+ ROUND(R.REAL_T * 1000 * 1000 / C.SIZE, 3) AS PAR_RESULT,
ROUND(R.REAL_T * 1000 * 1000 / G.NUM_NODES, 3) AS PAR_DATA,
- ROUND(R.REAL_T * 1000 * 1000 / G.TC_ITER, 3) AS PAR_ITER
-FROM INPUT_GRAPH G, INPUT_FILE F, BEST_RESULT R
-WHERE G.FILE_ID = F.FILE_ID AND R.FILE_ID = F.FILE_ID
-AND R.BENCH = 'tcff'
+ ROUND(R.REAL_T * 1000 * 1000 / C.ITER, 3) AS PAR_ITER
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C, BEST_RESULT R
+WHERE G.FILE_ID = F.FILE_ID
+AND R.FILE_ID = F.FILE_ID AND R.BENCH = 'tcff'
+AND C.FILE_ID = F.FILE_ID AND C.BENCH = 'tcff'
ORDER BY R.SYS, R.BENCH, F.SEQ_NO;
diff --git a/db/v_graph.sql b/db/v_graph.sql
index b816121ea25455df5c52aa27af7f5d2af7f5a60c..547e32c5116ad93482d087f7d8d2574490ba9d01 100644
--- a/db/v_graph.sql
+++ b/db/v_graph.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_graph.sql
-- Purpose: Benchmark DB: Views to Produce Graph Data in Various Formats
--- Last Change: 04.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -38,12 +38,12 @@ SELECT G.FILE_ID,
(CASE WHEN CYCLES='Y' THEN 'yes'
WHEN CYCLES='N' THEN 'no'
ELSE '?' END) || '&' ||
- G.TC_SIZE || '&' ||
- G.TC_ITER || '&' ||
- G.TC_COST || '\\'
+ C.SIZE || '&' ||
+ C.ITER || '&' ||
+ C.INST || '\\'
AS LATEX
-FROM INPUT_GRAPH G, INPUT_FILE F
-WHERE G.FILE_ID = F.FILE_ID
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C
+WHERE G.FILE_ID = F.FILE_ID AND C.FILE_ID = F.FILE_ID AND C.BENCH = 'tcff'
ORDER BY F.SEQ_NO;
-- ============================================================================
@@ -91,13 +91,13 @@ SELECT F.SEQ_NO, F.FILE_ID,
(CASE WHEN CYCLES='Y' THEN 'yes'
WHEN CYCLES='N' THEN 'no'
ELSE '?' END) || '</td>' || '
- ' || '<td class="colTCSize" >' || TC_SIZE || '</td>' || '
- ' || '<td class="colIter" >' || TC_ITER || '</td>' || '
- ' || '<td class="colRuleInst">' || TC_COST || '</td>' || '
+ ' || '<td class="colTCSize" >' || C.SIZE || '</td>' || '
+ ' || '<td class="colIter" >' || C.ITER || '</td>' || '
+ ' || '<td class="colRuleInst">' || C.INST || '</td>' || '
' || '</tr>'
AS HTML
-FROM INPUT_GRAPH G, INPUT_FILE F
-WHERE G.FILE_ID = F.FILE_ID
+FROM INPUT_GRAPH G, INPUT_FILE F, BENCH_COST C
+WHERE G.FILE_ID = F.FILE_ID AND C.FILE_ID = F.FILE_ID AND C.BENCH = 'tcff'
UNION ALL
SELECT 1000000 AS SEQ_NO, NULL AS FILE_ID,
'</table>' AS HTML
diff --git a/db/v_j1axy.sql b/db/v_j1axy.sql
index b15253b6a58962af6c8978599e735a292d6930fe..06d1babee1c6450f14bd82bce4093c0de5812f2d 100644
--- a/db/v_j1axy.sql
+++ b/db/v_j1axy.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_j1axy.sql
-- Purpose: Benchmark DB: Views to analyze Join1 a(X,Y) Benchmark Results
--- Last Change: 28.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -547,7 +547,7 @@ SELECT F.SEQ_NO, F.FILE_ID,
' || '<td class="colB1Size" >' || B1_SIZE || '</td>' || '
' || '<td class="colB2Size" >' || B2_SIZE || '</td>' || '
' || '<td class="colC1Size" >' || C1_SIZE || '</td>' || '
- ' || '<td class="colRuleInst">' || A_COST+B1_COST+B2_COST+C1_COST
+ ' || '<td class="colRuleInst">' || A_INST+B1_INST+B2_INST+C1_INST
|| '</td>' || '
' || '</tr>'
AS HTML
@@ -576,14 +576,14 @@ CREATE VIEW J1AXY_PLOT_REAL AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- (C.A_COST+C.B1_COST+C.B2_COST+C.C1_COST) AS COST,
+ (C.A_INST+C.B1_INST+C.B2_INST+C.C1_INST) AS COST,
R.REAL_T,
'"' || R.FILE_ID || '","' ||
- (C.A_COST+C.B1_COST+C.B2_COST+C.C1_COST) || '","'
+ (C.A_INST+C.B1_INST+C.B2_INST+C.C1_INST) || '","'
|| R.REAL_T || '"'
AS CSV,
R.FILE_ID || ' ' ||
- (C.A_COST+C.B1_COST+C.B2_COST+C.C1_COST) || ' ' || R.REAL_T
+ (C.A_INST+C.B1_INST+C.B2_INST+C.C1_INST) || ' ' || R.REAL_T
AS TSV
FROM RESULT R, INPUT_FILE F, INPUT_JOIN1 C, BENCH_IMPL I
WHERE R.BENCH = 'j1axy' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = C.FILE_ID
diff --git a/db/v_sgff.sql b/db/v_sgff.sql
index 1b7f1b3838f351b232eb34178f09765d609e5d36..471f1bbe5a63036cfe2564edd94a4ee873e2ee80 100644
--- a/db/v_sgff.sql
+++ b/db/v_sgff.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_sgff.sql
-- Purpose: Benchmark DB: Views for Analysis of SGFF Benchmark Results
--- Last Change: 18.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -30,15 +30,17 @@ CREATE VIEW SGFF_PLOT_REAL AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.SG_COST,
+ C.INST AS SG_COST,
R.REAL_T,
- '"' || R.FILE_ID || '","' || G.SG_COST || '","' || R.REAL_T || '"'
+ '"' || R.FILE_ID || '","' || C.INST || '","' || R.REAL_T || '"'
AS CSV,
- R.FILE_ID || ' ' || G.SG_COST || ' ' || R.REAL_T
+ R.FILE_ID || ' ' || C.INST || ' ' || R.REAL_T
AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'sgff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'sgff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS SG_COST, 0 AS REAL_T,
@@ -60,15 +62,17 @@ CREATE VIEW SGFF_PLOT_CPU AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.SG_COST,
+ C.INST AS SG_COST,
R.CPU_T,
- '"' || R.FILE_ID || '","' || G.SG_COST || '","' || R.CPU_T || '"'
+ '"' || R.FILE_ID || '","' || C.INST || '","' || R.CPU_T || '"'
AS CSV,
- R.FILE_ID || ' ' || G.SG_COST || ' ' || R.CPU_T
+ R.FILE_ID || ' ' || C.INST || ' ' || R.CPU_T
AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'sgff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'sgff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS SG_COST, 0 AS CPU_T,
@@ -90,15 +94,17 @@ CREATE VIEW SGFF_PLOT_MEM AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.NUM_EDGES + G.SG_SIZE AS FACTS,
+ G.NUM_EDGES + C.SIZE AS FACTS,
R.MEM_MB,
- '"' || R.FILE_ID || '","' || (G.NUM_EDGES+G.SG_SIZE) || '","' ||
+ '"' || R.FILE_ID || '","' || (G.NUM_EDGES+C.SIZE) || '","' ||
R.MEM_MB || '"' AS CSV,
- R.FILE_ID || ' ' || (G.NUM_EDGES+G.SG_SIZE) || ' ' ||
+ R.FILE_ID || ' ' || (G.NUM_EDGES+C.SIZE) || ' ' ||
R.MEM_MB AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'sgff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'sgff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'sgff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS FACTS, 0 AS CPU_T,
diff --git a/db/v_tcff.sql b/db/v_tcff.sql
index 71cc026c19a79695de533616ca6823b101dcdd19..51abb217f0347b9692a4b1990158c9e5f85b1ca3 100644
--- a/db/v_tcff.sql
+++ b/db/v_tcff.sql
@@ -2,7 +2,7 @@
-- Project: rbench - Logic Programming and Database Benchmarks
-- Filename: db/v_tcff.sql
-- Purpose: Benchmark DB: Views for Data Analysis of TCFF Benchmark Results
--- Last Change: 18.03.2019
+-- Last Change: 04.10.2019
-- Language: SQL (Tested with PostgreSQL)
-- Author: Stefan Brass
-- EMail: brass@informatik.uni-halle.de
@@ -30,15 +30,17 @@ CREATE VIEW TCFF_PLOT_REAL AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.TC_COST,
+ C.INST AS TC_COST,
R.REAL_T,
- '"' || R.FILE_ID || '","' || G.TC_COST || '","' || R.REAL_T || '"'
+ '"' || R.FILE_ID || '","' || C.INST || '","' || R.REAL_T || '"'
AS CSV,
- R.FILE_ID || ' ' || G.TC_COST || ' ' || R.REAL_T
+ R.FILE_ID || ' ' || C.INST || ' ' || R.REAL_T
AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'tcff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'tcff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS TC_COST, 0 AS REAL_T,
@@ -60,15 +62,17 @@ CREATE VIEW TCFF_PLOT_CPU AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.TC_COST,
+ C.INST AS TC_COST,
R.CPU_T,
- '"' || R.FILE_ID || '","' || G.TC_COST || '","' || R.CPU_T || '"'
+ '"' || R.FILE_ID || '","' || C.INST || '","' || R.CPU_T || '"'
AS CSV,
- R.FILE_ID || ' ' || G.TC_COST || ' ' || R.CPU_T
+ R.FILE_ID || ' ' || C.INST || ' ' || R.CPU_T
AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'tcff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'tcff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS TC_COST, 0 AS CPU_T,
@@ -90,15 +94,17 @@ CREATE VIEW TCFF_PLOT_MEM AS
SELECT R.SYS,
R.FILE_ID,
F.SEQ_NO AS NO,
- G.NUM_EDGES + G.TC_SIZE AS FACTS,
+ G.NUM_EDGES + C.SIZE AS FACTS,
R.MEM_MB,
- '"' || R.FILE_ID || '","' || (G.NUM_EDGES+G.TC_SIZE) || '","' ||
+ '"' || R.FILE_ID || '","' || (G.NUM_EDGES+C.SIZE) || '","' ||
R.MEM_MB || '"' AS CSV,
- R.FILE_ID || ' ' || (G.NUM_EDGES+G.TC_SIZE) || ' ' ||
+ R.FILE_ID || ' ' || (G.NUM_EDGES+C.SIZE) || ' ' ||
R.MEM_MB AS TSV
-FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_IMPL I
-WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID AND R.FILE_ID = G.FILE_ID
-AND I.BENCH = R.BENCH AND I.IMPL = R.IMPL AND I.SYS = R.SYS
+FROM RESULT R, INPUT_FILE F, INPUT_GRAPH G, BENCH_COST C, BENCH_IMPL I
+WHERE R.BENCH = 'tcff' AND R.FILE_ID = F.FILE_ID
+AND G.FILE_ID = F.FILE_ID
+AND C.BENCH = 'tcff' AND C.FILE_ID = F.FILE_ID
+AND I.BENCH = 'tcff' AND I.IMPL = R.IMPL AND I.SYS = R.SYS
AND I.STATUS = 'B'
UNION
SELECT DISTINCT P.SYS, '', -1, 0 AS FACTS, 0 AS CPU_T,
diff --git a/graph/gen_graphs b/graph/gen_graphs
new file mode 100755
index 0000000000000000000000000000000000000000..57982db3b21e41cfd6604be15ca901f493c9fc85
--- /dev/null
+++ b/graph/gen_graphs
@@ -0,0 +1,69 @@
+#!/bin/bash
+P=../../data_p
+T=../../data_tsv
+C=../../data_csv
+S=../../data_sql
+./graph a1k_50k $P/a1k_50k.P $T/a1k_50k.tsv $C/a1k_50k.csv $S/a1k_50k.sql
+./graph a1k_125k $P/a1k_125k.P $T/a1k_125k.tsv $C/a1k_125k.csv $S/a1k_125k.sql
+./graph a1k_250k $P/a1k_250k.P $T/a1k_250k.tsv $C/a1k_250k.csv $S/a1k_250k.sql
+./graph a2k_200k $P/a2k_200k.P $T/a2k_200k.tsv $C/a2k_200k.csv $S/a2k_200k.sql
+./graph a2k_500k $P/a2k_500k.P $T/a2k_500k.tsv $C/a2k_500k.csv $S/a2k_500k.sql
+./graph a2k_1m $P/a2k_1m.P $T/a2k_1m.tsv $C/a2k_1m.csv $S/a2k_1m.sql
+./graph b17 $P/b17.P $T/b17.tsv $C/b17.csv $S/b17.sql
+./graph b18 $P/b18.P $T/b18.tsv $C/b18.csv $S/b18.sql
+./graph b19 $P/b19.P $T/b19.tsv $C/b19.csv $S/b19.sql
+./graph c1k $P/c1k.P $T/c1k.tsv $C/c1k.csv $S/c1k.sql
+./graph c2k $P/c2k.P $T/c2k.tsv $C/c2k.csv $S/c2k.sql
+./graph c3k $P/c3k.P $T/c3k.tsv $C/c3k.csv $S/c3k.sql
+./graph c4k $P/c4k.P $T/c4k.tsv $C/c4k.csv $S/c4k.sql
+./graph k100 $P/k100.P $T/k100.tsv $C/k100.csv $S/k100.sql
+./graph k1k $P/k1k.P $T/k1k.tsv $C/k1k.csv $S/k1k.sql
+./graph k1k5 $P/k1k5.P $T/k1k5.tsv $C/k1k5.csv $S/k1k5.sql
+./graph k2k $P/k2k.P $T/k2k.tsv $C/k2k.csv $S/k2k.sql
+./graph k50 $P/k50.P $T/k50.tsv $C/k50.csv $S/k50.sql
+./graph k500 $P/k500.P $T/k500.tsv $C/k500.csv $S/k500.sql
+./graph m4_2ki $P/m4_2ki.P $T/m4_2ki.tsv $C/m4_2ki.csv $S/m4_2ki.sql
+./graph m16_512 $P/m16_512.P $T/m16_512.tsv $C/m16_512.csv $S/m16_512.sql
+./graph m64_128 $P/m64_128.P $T/m64_128.tsv $C/m64_128.csv $S/m64_128.sql
+./graph m256_32 $P/m256_32.P $T/m256_32.tsv $C/m256_32.csv $S/m256_32.sql
+./graph m1ki_8 $P/m1ki_8.P $T/m1ki_8.tsv $C/m1ki_8.csv $S/m1ki_8.sql
+./graph m4ki_2 $P/m4ki_2.P $T/m4ki_2.tsv $C/m4ki_2.csv $S/m4ki_2.sql
+./graph p1k $P/p1k.P $T/p1k.tsv $C/p1k.csv $S/p1k.sql
+./graph p2k $P/p2k.P $T/p2k.tsv $C/p2k.csv $S/p2k.sql
+./graph p3k $P/p3k.P $T/p3k.tsv $C/p3k.csv $S/p3k.sql
+./graph p4k $P/p4k.P $T/p4k.tsv $C/p4k.csv $S/p4k.sql
+./graph s1k_1 $P/s1k_1.P $T/s1k_1.tsv $C/s1k_1.csv $S/s1k_1.sql
+./graph s1k_2 $P/s1k_2.P $T/s1k_2.tsv $C/s1k_2.csv $S/s1k_2.sql
+./graph s1k_3 $P/s1k_3.P $T/s1k_3.tsv $C/s1k_3.csv $S/s1k_3.sql
+./graph s1k_4 $P/s1k_4.P $T/s1k_4.tsv $C/s1k_4.csv $S/s1k_4.sql
+./graph s1k_5 $P/s1k_5.P $T/s1k_5.tsv $C/s1k_5.csv $S/s1k_5.sql
+./graph s2k_1 $P/s2k_1.P $T/s2k_1.tsv $C/s2k_1.csv $S/s2k_1.sql
+./graph s2k_2 $P/s2k_2.P $T/s2k_2.tsv $C/s2k_2.csv $S/s2k_2.sql
+./graph s2k_3 $P/s2k_3.P $T/s2k_3.tsv $C/s2k_3.csv $S/s2k_3.sql
+./graph s2k_4 $P/s2k_4.P $T/s2k_4.tsv $C/s2k_4.csv $S/s2k_4.sql
+./graph s2k_5 $P/s2k_5.P $T/s2k_5.tsv $C/s2k_5.csv $S/s2k_5.sql
+./graph t50 $P/t50.P $T/t50.tsv $C/t50.csv $S/t50.sql
+./graph t100 $P/t100.P $T/t100.tsv $C/t100.csv $S/t100.sql
+./graph t500 $P/t500.P $T/t500.tsv $C/t500.csv $S/t500.sql
+./graph t1k $P/t1k.P $T/t1k.tsv $C/t1k.csv $S/t1k.sql
+./graph t1k5 $P/t1k5.P $T/t1k5.tsv $C/t1k5.csv $S/t1k5.sql
+./graph t2k $P/t2k.P $T/t2k.tsv $C/t2k.csv $S/t2k.sql
+./graph u1k_50k $P/u1k_50k.P $T/u1k_50k.tsv $C/u1k_50k.csv $S/u1k_50k.sql
+./graph u1k_125k $P/u1k_125k.P $T/u1k_125k.tsv $C/u1k_125k.csv $S/u1k_125k.sql
+./graph u1k_250k $P/u1k_250k.P $T/u1k_250k.tsv $C/u1k_250k.csv $S/u1k_250k.sql
+./graph u2k_200k $P/u2k_200k.P $T/u2k_200k.tsv $C/u2k_200k.csv $S/u2k_200k.sql
+./graph u2k_500k $P/u2k_500k.P $T/u2k_500k.tsv $C/u2k_500k.csv $S/u2k_500k.sql
+./graph u2k_1m $P/u2k_1m.P $T/u2k_1m.tsv $C/u2k_1m.csv $S/u2k_1m.sql
+./graph v10 $P/v10.P $T/v10.tsv $C/v10.csv $S/v10.sql
+./graph v11 $P/v11.P $T/v11.tsv $C/v11.csv $S/v11.sql
+./graph v12 $P/v12.P $T/v12.tsv $C/v12.csv $S/v12.sql
+./graph v17 $P/v17.P $T/v17.tsv $C/v17.csv $S/v17.sql
+./graph v18 $P/v18.P $T/v18.tsv $C/v18.csv $S/v18.sql
+./graph v19 $P/v19.P $T/v19.tsv $C/v19.csv $S/v19.sql
+./graph w1k $P/w1k.P $T/w1k.tsv $C/w1k.csv $S/w1k.sql
+#./graph wintest $P/wintest.P $T/wintest.tsv $C/wintest.csv $S/wintest.sql
+./graph x10k $P/x10k.P $T/x10k.tsv $C/x10k.csv $S/x10k.sql
+./graph y1k_4k $P/y1k_4k.P $T/y1k_4k.tsv $C/y1k_4k.csv $S/y1k_4k.sql
+./graph y1k_8k $P/y1k_8k.P $T/y1k_8k.tsv $C/y1k_8k.csv $S/y1k_8k.sql
+./graph y500_4k $P/y500_4k.P $T/y500_4k.tsv $C/y500_4k.csv $S/y500_4k.sql
+./graph y500_8k $P/y500_8k.P $T/y500_8k.tsv $C/y500_8k.csv $S/y500_8k.sql
diff --git a/graph/graph.cpp b/graph/graph.cpp
index fe72c01bfb60128db9393ce6df703e9d35657485..efd6a0bc8e3bc4ba8ac4784331ff36eeb1849fd1 100644
--- a/graph/graph.cpp
+++ b/graph/graph.cpp
@@ -2,7 +2,7 @@
// Project: rbench - Logic Programming and Database Benchmarks
// Filename: graph/graph.cpp
// Purpose: Alternative Graph Generator
-// Last Change: 03.10.2019
+// Last Change: 04.10.2019
// Language: C++
// EMail: brass@informatik.uni-halle.de
// WWW: http://www.informatik.uni-halle.de/~brass/
@@ -102,6 +102,9 @@
// {(i, 1 + ((i + (n*j)/(k+1)) mod n) | i = 1,...,n; j=1,...,k }
// Tn: Total order graph (maximum acyclic graph)
+// n vertices, n * (n-1) / 2 edges
+// V = {1, ..., n}
+// E = {(i, j) | i = 1, ..., n-1; j = i+1, ..., n }
// Un_k: Random Graph, Uniform Distribution of Node Degrees
// n nodes, k edges
@@ -114,14 +117,20 @@
// Wn_k: Single level of edges, mesh, 2n nodes, degree k):
+// Wn: Abbreviation for Wn_n (n nodes connected to all n other nodes)
+// 2*n nodes, n*k edges
// V = {1, ..., 2n}
// E = { (i, (i+j-1 mod n)+n+1) | 1 <= i <= n and 0 <= j < k }
// Xn_k: X-graph (n edges to central node, k from there)
+// Xn: Abbreviation for Xn_n
+// n+k+1 nodes, n+k edges
+// V = {1, ..., n+k+1}
+// E = {(i, n+1) | i = 1, ..., n} u {(n+1, n+1+j) | j = 1, ..., k}
// Yn_k: Y-graph (n nodes point to central node, then path of length k)
-// n+k nodes: V= {1,...,n+k}
-// n+k-1 edges:
+// n+k nodes, n+k-1 edges:
+// V = {1,...,n+k}
// E = {(i, n+1)} | i = 1,...,n} u {(n+i, n+i+1) | i = 1,...,k-1}
// The program code for graph g is in the procedure gen_graph_g.
@@ -138,6 +147,19 @@
#define MAX_OUT_FILES 4
+//=============================================================================
+// Name of the predicate or relation for the edges:
+//=============================================================================
+
+#define PRED_NAME "par"
+
+//=============================================================================
+// Should Test Output be Written?
+//=============================================================================
+
+// 0 means no test output, 1 means test output:
+#define TEST_OUTPUT 0
+
//=============================================================================
// Include Files:
//=============================================================================
@@ -307,7 +329,7 @@ long str_int(str_t digits) {
int prev_result = result;
result = result * 10 + digit_val(*p);
if(result/10 != prev_result) {
- std::cout << "Overflow!\n";
+ std::cerr << "Overflow!\n";
exit(2);
}
}
@@ -340,7 +362,7 @@ class Node {
void store_connection(int to) {
// Check nodes:
if(to <= 0 || to > num_nodes_) {
- std::cout << "Invalid to node: " << to << "\n";
+ std::cerr << "Invalid to node: " << to << "\n";
exit(3);
}
@@ -357,7 +379,7 @@ class Node {
bool connection_exists(int to) const {
// Check node:
if(to <= 0 || to > num_nodes_) {
- std::cout << "Invalid to node: " << to << "\n";
+ std::cerr << "Invalid to node: " << to << "\n";
exit(4);
}
@@ -373,7 +395,7 @@ class Node {
// Copy other node into this node:
void copy(Node *node) {
if(node->num_nodes_ != num_nodes_) {
- std::cout << "Can copy only node of same size!\n";
+ std::cerr << "Can copy only node of same size!\n";
exit(5);
}
int bytes = (num_nodes_ + BYTE_SIZE - 1) / BYTE_SIZE;
@@ -422,11 +444,11 @@ class Graph {
void store_edge(int from, int to) {
// Check nodes:
if(from <= 0 || from > num_nodes_) {
- std::cout << "Invalid from node: " << from << "\n";
+ std::cerr << "Invalid from node: " << from << "\n";
exit(6);
}
if(to <= 0 || to > num_nodes_) {
- std::cout << "Invalid to node: " << to << "\n";
+ std::cerr << "Invalid to node: " << to << "\n";
exit(7);
}
@@ -439,11 +461,11 @@ class Graph {
bool edge_exists(int from, int to) const {
// Check nodes:
if(from <= 0 || from > num_nodes_) {
- std::cout << "Invalid from node: " << from << "\n";
+ std::cerr << "Invalid from node: " << from << "\n";
exit(8);
}
if(to <= 0 || to > num_nodes_) {
- std::cout << "Invalid to node: " << to << "\n";
+ std::cerr << "Invalid to node: " << to << "\n";
exit(9);
}
@@ -456,7 +478,7 @@ class Graph {
void copy(Graph *g)
{
if(g->num_nodes_ != num_nodes_) {
- std::cout << "Can copy only graph of same size!\n";
+ std::cerr << "Can copy only graph of same size!\n";
exit(10);
}
for(int i = 0; i < num_nodes_; i++)
@@ -886,7 +908,7 @@ typedef TC *tc_t;
// Data Type for File Formats:
//=============================================================================
-typedef enum { FORMAT_P, FORMAT_TSV } format_t;
+typedef enum { FORMAT_P, FORMAT_TSV, FORMAT_CSV, FORMAT_SQL } format_t;
str_t format_name(format_t f) {
switch(f) {
@@ -894,6 +916,10 @@ str_t format_name(format_t f) {
return ".P";
case FORMAT_TSV:
return ".tsv";
+ case FORMAT_CSV:
+ return ".csv";
+ case FORMAT_SQL:
+ return ".sql";
default:
return "UNKNOWN FORMAT";
}
@@ -924,7 +950,7 @@ class Output {
void add_file(str_t filename) {
// Check that there is still space in the array:
if(num_files_ >= MAX_OUT_FILES) {
- std::cout << "Too many output files!\n";
+ std::cerr << "Too many output files!\n";
exit(11);
}
@@ -936,8 +962,16 @@ class Output {
filename[n-3] == 't' && filename[n-2] == 's' &&
filename[n-1] == 'v')
format_[num_files_] = FORMAT_TSV;
+ else if(n > 4 && filename[n-4] == '.' &&
+ filename[n-3] == 'c' && filename[n-2] == 's' &&
+ filename[n-1] == 'v')
+ format_[num_files_] = FORMAT_CSV;
+ else if(n > 4 && filename[n-4] == '.' &&
+ filename[n-3] == 's' && filename[n-2] == 'q' &&
+ filename[n-1] == 'l')
+ format_[num_files_] = FORMAT_SQL;
else {
- std::cout << "Output file '" << filename <<
+ std::cerr << "Output file '" << filename <<
"' has unknown extension.\n";
exit(12);
}
@@ -946,7 +980,7 @@ class Output {
std::ofstream *output = files_ + num_files_;
output->open(filename);
if(output->fail()) {
- std::cout << "Opening output file '" << filename <<
+ std::cerr << "Opening output file '" << filename <<
"' failed.\n";
exit(13);
}
@@ -958,7 +992,7 @@ class Output {
// The graph data structure depends on the number of nodes:
if(num_nodes_ != 0) {
- std::cout << "Number of nodes in output can be set " <<
+ std::cerr << "Number of nodes in output can be set " <<
"only once!\n";
exit(14);
}
@@ -975,23 +1009,23 @@ class Output {
// The number of nodes must be set:
if(num_nodes_ <= 0) {
- std::cout << "The number of nodes must be set first!\n";
+ std::cerr << "The number of nodes must be set first!\n";
exit(15);
}
// Check validity of parameter values:
if(from <= 0 || from > num_nodes_) {
- std::cout << "from node: " << from << " invalid.\n";
+ std::cerr << "from node: " << from << " invalid.\n";
exit(16);
}
if(to <= 0 || to > num_nodes_) {
- std::cout << "to node: " << to << " invalid.\n";
+ std::cerr << "to node: " << to << " invalid.\n";
exit(17);
}
// Check that there is at least one file open:
if(num_files_ == 0) {
- std::cout << "No file open.\n";
+ std::cerr << "No file open.\n";
exit(18);
}
@@ -1000,12 +1034,21 @@ class Output {
std::ofstream &file = files_[i];
switch(format_[i]) {
case FORMAT_P:
- file << "par(" << from << ", " << to <<
+ file << PRED_NAME << "(" <<
+ from << ", " << to <<
")." << "\n";
break;
case FORMAT_TSV:
file << from << "\t" << to << "\n";
break;
+ case FORMAT_CSV:
+ file << from << "," << to << "\n";
+ break;
+ case FORMAT_SQL:
+ file << "INSERT INTO " << PRED_NAME <<
+ " VALUES (" <<
+ from << "," << to << ");\n";
+ break;
}
}
@@ -1021,7 +1064,7 @@ class Output {
for(int i = 0; i < num_files_; i++) {
std::ofstream &file = files_[i];
if(file.bad()) {
- std::cout << "Problem with output file " <<
+ std::cerr << "Problem with output file " <<
"(Format " <<
format_name(format_[i]) << ").\n";
exit(19);
@@ -1058,7 +1101,7 @@ typedef Output *output_t;
//-----------------------------------------------------------------------------
void no_second_par(str_t graph_name) {
- std::cout << "Graph '" << graph_name << "' has no second parameter.\n";
+ std::cerr << "Graph '" << graph_name << "' has no second parameter.\n";
exit(20);
}
@@ -1067,7 +1110,7 @@ void no_second_par(str_t graph_name) {
//-----------------------------------------------------------------------------
void second_par_missing(str_t graph_name) {
- std::cout << "Graph '" << graph_name << "' needs a second parameter.\n";
+ std::cerr << "Graph '" << graph_name << "' needs a second parameter.\n";
exit(21);
}
@@ -1076,7 +1119,7 @@ void second_par_missing(str_t graph_name) {
//-----------------------------------------------------------------------------
void par_too_small(str_t graph_name, str_t par_name, int min) {
- std::cout << "Value of parameter '" << par_name << "' " <<
+ std::cerr << "Value of parameter '" << par_name << "' " <<
"for Graph '" << graph_name << "' too small: " <<
"Minimum " << min << ".\n";
exit(22);
@@ -1123,7 +1166,7 @@ int next_prime(int p) {
while(!is_prime(p)) {
p++;
if(p < 0) {
- std::cout <<
+ std::cerr <<
"Overflow in computation of prime number\n";
exit(23);
}
@@ -1196,8 +1239,10 @@ int rand_until(int n) {
std::cerr << "Error in random number generator!\n";
exit(24);
}
+#if TEST_OUTPUT
std::cout << "rand_until(" << n << ") = " << result <<
"\t[rand = " << r << "]\n";
+#endif
return result;
}
@@ -1213,11 +1258,13 @@ void rand_shuffle(int arr[], int len) {
arr[j] = swap;
}
+#if TEST_OUTPUT
std::cout << "\n";
std::cout << "Result of Shuffle (Array Length: " << len << ")\n";
for(int i = 0; i < len; i++)
std::cout << "\t arr[" << i << "] = " << arr[i] << "\n";
std::cout << "\n";
+#endif
}
@@ -1417,13 +1464,13 @@ void gen_graph_s(int n, int k, output_t out, gsize_t gsize) {
// Check divisibility condition:
if(n % (k+1) != 0) {
- std::cout << "S[n,k] requires that n is divisible by k+1.\n";
+ std::cerr << "S[n,k] requires that n is divisible by k+1.\n";
exit(25);
}
// Check that k is not too large:
if(n < (k+1)) {
- std::cout << "S[n,k] requires that n > k.\n";
+ std::cerr << "S[n,k] requires that n > k.\n";
exit(26);
}
@@ -1650,7 +1697,7 @@ void gen_graph_w(int n, int k, output_t out, gsize_t gsize) {
// Default for second parameter:
if(k < 1)
- k = 1;
+ k = n;
// Set number of nodes:
out->set_nodes(2 * n);
@@ -1746,7 +1793,7 @@ void gen_graph_u_a(int n, int k, output_t out, gsize_t gsize, bool acyclic) {
// Ensure termination:
if(i >= p1 * p2) {
- std::cout <<
+ std::cerr <<
"Number of edges too large in random graph.\n";
exit(27);
}
@@ -1825,7 +1872,7 @@ void gen_graph_u(int n, int k, output_t out, gsize_t gsize) {
// Check relation between n and k:
if(k > n * n) {
- std::cout <<
+ std::cerr <<
"Impossible number of edges (too large) in U-Graph";
exit(28);
}
@@ -1867,7 +1914,7 @@ void gen_graph_a(int n, int k, output_t out, gsize_t gsize) {
// Check relation between n and k:
if(k > n * (n-1) / 2) {
- std::cout <<
+ std::cerr <<
"Impossible number of edges (too large) in A-Graph";
exit(29);
}
@@ -1890,7 +1937,7 @@ int main(int argc, str_t argv[])
// The program should be called with the graph and the output files:
if(argc < 3) {
- std::cout << "Usage: ./graph GraphID OutputFile1 ...\n";
+ std::cerr << "Usage: ./graph GraphID OutputFile1 ...\n";
exit(30);
}
@@ -1899,7 +1946,7 @@ int main(int argc, str_t argv[])
str_t p = graph_id;
char graph_code = *p++;
if(graph_code == '\0') {
- std::cout << "Impossible empty Graph ID.\n";
+ std::cerr << "Impossible empty Graph ID.\n";
exit(31);
}
@@ -1910,7 +1957,7 @@ int main(int argc, str_t argv[])
par_chars[i] = *p++;
par_chars[i] = 0;
if(i == 0) {
- std::cout << "First parameter in graph ID missing.\n";
+ std::cerr << "First parameter in graph ID missing.\n";
exit(32);
}
int par1 = str_int(par_chars);
@@ -1945,7 +1992,7 @@ int main(int argc, str_t argv[])
par_chars[i] = *p++;
par_chars[i] = 0;
if(i == 0) {
- std::cout << "Second parameter in graph ID missing.\n";
+ std::cerr << "Second parameter in graph ID missing.\n";
exit(33);
}
par2 = str_int(par_chars);
@@ -1975,7 +2022,7 @@ int main(int argc, str_t argv[])
// Check that we have successfully parsed the entire graph ID:
if(*p != '\0') {
- std::cout << "Unexpected characters at the end of graph ID.\n";
+ std::cerr << "Unexpected characters at the end of graph ID.\n";
exit(34);
}
@@ -2060,7 +2107,7 @@ int main(int argc, str_t argv[])
gen_graph_a(par1, par2, &output, &formula);
break;
default:
- std::cout << "Unknown graph type '" << graph_code <<
+ std::cerr << "Unknown graph type '" << graph_code <<
"'.\n";
exit(35);
}
@@ -2078,7 +2125,9 @@ int main(int argc, str_t argv[])
std::cout << "\n";
// Test output:
- //tc.print();
+#if TEST_OUTPUT
+ tc.print();
+#endif
// Compute same generation data:
std::cout << "Compute same generation data ...\n";
@@ -2087,7 +2136,9 @@ int main(int argc, str_t argv[])
std::cout << "\n";
// Test output:
+#if TEST_OUTPUT
sg.print();
+#endif
// Create object for graph/benchmark size computed from graph:
GSize gsize;