/*****************************************************************************
We run the program by calling something along the lines of:
guessMushroom -train mushrooms.data -test moreMushrooms.data -out mushrooms.out
******************************************************************************/
int main(int argc, char* argv[]) {
MushroomP trainingShroomList;
trainingShroomList = NULL;
MushroomP testingShroomList;
testingShroomList = NULL;
ValueP availableValues;
DecisionP decisionTree;
availableValues = NULL;
decisionTree = NULL;
char infilenameTraining[STRLEN]; /* input file name for training data */
char infilenameTesting[STRLEN]; /* input file name for testing data */
char outfilename[STRLEN]; /* output file name for printing the results */
printf( "This program processes data about the attributes mushrooms.\n" );
/* initialize filenames to dummy strings */
strcpy( infilenameTraining, "none" );
strcpy( infilenameTesting, "none" );
strcpy( outfilename, "none" );
/* TAKEN FROM p5.c EXAMLPE
process command-line arguments */
while( *++argv ) { /* while there are still command-line args */
if ( !strcmp( *argv, "-train" ) ) /* set input file name */
strcpy( infilenameTraining, *++argv );
else if( !strcmp( *argv,"-test" ) ) /* set input file name */
strcpy( infilenameTesting, *++argv );
else if( !strcmp( *argv,"-out" ) ) /* set output file name */
strcpy( outfilename,*++argv );
else { /* error checking */
printf( "\nERROR: option %s not recognized\n", *argv );
printf( "...exiting program\n\n" );
exit( 0 );
}
} /* end, while there are still command-line args */
/* this is the training data */
trainingShroomList = readMushrooms( infilenameTraining );
/* this is the testing data */
testingShroomList = readMushrooms( infilenameTesting );
availableValues = findAvailableValues( 'e', 0, trainingShroomList );
decisionTree = createDecisionTree( 'e', 0, trainingShroomList, availableValues );
printDecisionTree( decisionTree, outfilename );
printTestResults( 'e', 0, decisionTree, testingShroomList, outfilename );
/* free the list of test mushrooms */
freeMushrooms( trainingShroomList );
freeMushrooms( testingShroomList );
/* note that Decision structs are freed in printDecisionTree */
/* free value list */
freeValues( availableValues );
return (0);
} /* end main */
SelectionTreeElement::~SelectionTreeElement()
{
/* Free the children.
* Must be done before freeing other data, because the children may hold
* references to data in this element. */
child.reset();
freeValues();
freeExpressionData();
freeCompilerData();
}
void
testPrintTree(void)
{
RID insert[] = {
{1,1},
{2,3},
{1,2},
{3,5},
{4,4},
{3,2},
};
int numInserts = 6;
Value **keys;
char *stringKeys[] = {
"i1",
"i11",
"i13",
"i17",
"i23",
"i52"
};
testName = "test b-tree print";
int i, testint;
BTreeHandle *tree = NULL;
keys = createValues(stringKeys, numInserts);
// init
TEST_CHECK(initIndexManager(NULL));
TEST_CHECK(createBtree("testidx", DT_INT, 2));
TEST_CHECK(openBtree(&tree, "testidx"));
printf("0"); // remove it later
// insert keys
for(i = 0; i < numInserts; i++)
TEST_CHECK(insertKey(tree, keys[i], insert[i]));
char * expected = "(0)[1,13,2,23,3]\n(1)[1.1,1,2.3,11,2]\n(2)[1.2,13,3.5,17,3]\n(3)[4.4,23,3.2,52]\n";
//test printTree function
ASSERT_EQUALS_STRING(expected,printTree(tree),"checking b-tree shape"); \
// cleanup
TEST_CHECK(closeBtree(tree));
TEST_CHECK(deleteBtree("testidx"));
TEST_CHECK(shutdownIndexManager());
freeValues(keys, numInserts);
TEST_DONE();
}
// ************************************************************
void
testInsertAndFind (void)
{
RID insert[] = {
{1,1},
{2,3},
{1,2},
{3,5},
{4,4},
{3,2},
};
int numInserts = 6;
Value **keys;
char *stringKeys[] = {
"i1",
"i11",
"i13",
"i17",
"i23",
"i52"
};
testName = "test b-tree inserting and search";
int i, testint;
BTreeHandle *tree = NULL;
keys = createValues(stringKeys, numInserts);
// init
TEST_CHECK(initIndexManager(NULL));
printf("\n Init");
TEST_CHECK(createBtree("testidx", DT_INT, 2));
printf("\n Created");
TEST_CHECK(openBtree(&tree, "testidx"));
// insert keys
for(i = 0; i < numInserts; i++)
TEST_CHECK(insertKey(tree, keys[i], insert[i]));
// check index stats
TEST_CHECK(getNumNodes(tree, &testint));
ASSERT_EQUALS_INT(testint,4, "number of nodes in btree");
TEST_CHECK(getNumEntries(tree, &testint));
ASSERT_EQUALS_INT(testint, numInserts, "number of entries in btree");
// search for keys
for(i = 0; i < 1000; i++)
{
int pos = rand() % numInserts;
RID rid;
Value *key = keys[pos];
TEST_CHECK(findKey(tree, key, &rid));
ASSERT_EQUALS_RID(insert[pos], rid, "did we find the correct RID?");
}
// cleanup
TEST_CHECK(closeBtree(tree));
TEST_CHECK(deleteBtree("testidx"));
TEST_CHECK(shutdownIndexManager());
freeValues(keys, numInserts);
TEST_DONE();
}
// ************************************************************
void
testIndexScan (void)
{
RID insert[] = {
{1,1},
{2,3},
{1,2},
{3,5},
{4,4},
{3,2},
};
int numInserts = 6;
Value **keys;
char *stringKeys[] = {
"i1",
"i11",
"i13",
"i17",
"i23",
"i52"
};
testName = "random insertion order and scan";
int i, testint, iter, rc;
BTreeHandle *tree = NULL;
BT_ScanHandle *sc = NULL;
RID rid;
keys = createValues(stringKeys, numInserts);
// init
TEST_CHECK(initIndexManager(NULL));
for(iter = 0; iter < 50; iter++)
{
int *permute;
// create permutation
permute = createPermutation(numInserts);
// create B-tree
TEST_CHECK(createBtree("testidx", DT_INT, 2));
TEST_CHECK(openBtree(&tree, "testidx"));
// insert keys
for(i = 0; i < numInserts; i++)
TEST_CHECK(insertKey(tree, keys[permute[i]], insert[permute[i]]));
// check index stats
TEST_CHECK(getNumEntries(tree, &testint));
ASSERT_EQUALS_INT(testint, numInserts, "number of entries in btree");
// execute scan, we should see tuples in sort order
openTreeScan(tree, &sc);
i = 0;
while((rc = nextEntry(sc, &rid)) == RC_OK)
{
RID expRid = insert[i++];
ASSERT_EQUALS_RID(expRid, rid, "did we find the correct RID?");
}
ASSERT_EQUALS_INT(RC_IM_NO_MORE_ENTRIES, rc, "no error returned by scan");
ASSERT_EQUALS_INT(numInserts, i, "have seen all entries");
closeTreeScan(sc);
// cleanup
TEST_CHECK(closeBtree(tree));
TEST_CHECK(deleteBtree("testidx"));
free(permute);
}
TEST_CHECK(shutdownIndexManager());
freeValues(keys, numInserts);
TEST_DONE();
}
// ************************************************************
void
testDelete (void)
{
RID insert[] = {
{1,1},
{2,3},
{1,2},
{3,5},
{4,4},
{3,2},
};
int numInserts = 6;
Value **keys;
char *stringKeys[] = {
"i1",
"i11",
"i13",
"i17",
"i23",
"i52"
};
testName = "test b-tree inserting and search";
int i, iter;
BTreeHandle *tree = NULL;
int numDeletes = 3;
bool *deletes = (bool *) malloc(numInserts * sizeof(bool));
keys = createValues(stringKeys, numInserts);
// init
TEST_CHECK(initIndexManager(NULL));
// create test b-tree and randomly remove entries
for(iter = 0; iter < 50; iter++)
{
// randomly select entries for deletion (may select the same on twice)
for(i = 0; i < numInserts; i++)
deletes[i] = FALSE;
for(i = 0; i < numDeletes; i++)
deletes[rand() % numInserts] = TRUE;
// init B-tree
TEST_CHECK(createBtree("testidx", DT_INT, 2));
TEST_CHECK(openBtree(&tree, "testidx"));
// insert keys
for(i = 0; i < numInserts; i++)
TEST_CHECK(insertKey(tree, keys[i], insert[i]));
// delete entries
for(i = 0; i < numInserts; i++)
{
if (deletes[i])
TEST_CHECK(deleteKey(tree, keys[i]));
}
// search for keys
for(i = 0; i < 1000; i++)
{
int pos = rand() % numInserts;
RID rid;
Value *key = keys[pos];
if (deletes[pos])
{
int rc = findKey(tree, key, &rid);
ASSERT_TRUE((rc == RC_IM_KEY_NOT_FOUND), "entry was deleted, should not find it");
}
else
{
TEST_CHECK(findKey(tree, key, &rid));
ASSERT_EQUALS_RID(insert[pos], rid, "did we find the correct RID?");
}
}
// cleanup
TEST_CHECK(closeBtree(tree));
TEST_CHECK(deleteBtree("testidx"));
}
TEST_CHECK(shutdownIndexManager());
freeValues(keys, numInserts);
free(deletes);
TEST_DONE();
}