/**
* Create a new TableStats object, that keeps track of statistics on each column of a table
*
* @param tableid The table over which to compute statistics
* @param ioCostPerPage The cost per page of IO. This doesn't differentiate between
* sequential-scan IO and disk seeks.
*/
public TableStats(int tableid, int ioCostPerPage) {
// For this function, we use the DbFile for the table in question,
// then scan through its tuples and calculate the values that you
// to build the histograms.
// TODO: Fill out the rest of the constructor.
// Feel free to change anything already written, it's only a guideline
this.ioCostPerPage = ioCostPerPage;
DbFile file = Database.getCatalog().getDbFile(tableid);
tupleDesc = file.getTupleDesc();
numPages = ((HeapFile) file).numPages();
numTuples = 0;
int numFields = tupleDesc.numFields();
// TODO: what goes here?
statistics = new ArrayList<Object>();
for (int i = 0; i < numFields; i++) {
if (Type.INT_TYPE.equals(tupleDesc.getFieldType(i))) {
statistics.add(new IntStatistics(NUM_HIST_BINS));
} else {
statistics.add(new StringHistogram(NUM_HIST_BINS));
}
}
final DbFileIterator iter = file.iterator(null);
try {
iter.open();
while (iter.hasNext()) {
Tuple t = iter.next();
numTuples++;
// TODO: and here?
for (int i = 0; i < numFields; i++) {
if (Type.INT_TYPE.equals(tupleDesc.getFieldType(i))) {
((IntStatistics) statistics.get(i)).addValue(((IntField) t.getField(i)).getValue());
} else {
((StringHistogram) statistics.get(i))
.addValue(((StringField) t.getField(i)).getValue());
}
}
}
iter.close();
} catch (DbException e) {
e.printStackTrace();
} catch (TransactionAbortedException e) {
e.printStackTrace();
}
}
/**
* Merge a new tuple into the aggregate, grouping as indicated in the constructor
*
* @param tup the Tuple containing an aggregate field and a group-by field
*/
public void mergeTupleIntoGroup(Tuple tup) {
if (isGrouping()) {
Field gbValue = tup.getField(this.gbField);
if (this.tupleStorage.containsKey(gbValue)) {
this.tupleStorage.get(gbValue).add((IntField) tup.getField(this.aField));
} else {
this.tupleStorage.put(gbValue, new ArrayList<IntField>());
this.tupleStorage.get(gbValue).add((IntField) tup.getField(this.aField));
}
} else {
this.tupleStorageNoGrouping.add((IntField) tup.getField(this.aField));
}
}
/** Unit test for Tuple.getField() and Tuple.setField() */
@Test
public void modifyFields() {
TupleDesc td = Utility.getTupleDesc(2);
Tuple tup = new Tuple(td);
tup.setField(0, new IntField(-1));
tup.setField(1, new IntField(0));
assertEquals(new IntField(-1), tup.getField(0));
assertEquals(new IntField(0), tup.getField(1));
tup.setField(0, new IntField(1));
tup.setField(1, new IntField(37));
assertEquals(new IntField(1), tup.getField(0));
assertEquals(new IntField(37), tup.getField(1));
}
/**
* Merge a new tuple into the aggregate, grouping as indicated in the constructor
*
* @param tup the Tuple containing an aggregate field and a group-by field
*/
public void merge(Tuple tup) {
if (_gb == NO_GROUPING) {
_count++;
} else {
Field f = tup.getField(_gb);
Integer c = _counts.get(f);
if (c == null) c = 1;
else c++;
_counts.put(f, c);
}
}