gparai commented on a change in pull request #1744: Drill 7148 - Join order,
multi-col ndv and aggregate rowcount fixes for TPCH queries
URL: https://github.com/apache/drill/pull/1744#discussion_r275195359
##########
File path:
exec/java-exec/src/main/java/org/apache/drill/exec/planner/cost/DrillRelMdDistinctRowCount.java
##########
@@ -208,8 +244,64 @@ public Double getDistinctRowCount(DrillJoinRelBase
joinRel, RelMetadataQuery mq,
Double leftDistRowCount = null;
Double rightDistRowCount = null;
double distRowCount = 1;
+ int gbyCols = 0;
ImmutableBitSet lmb = leftMask.build();
ImmutableBitSet rmb = rightMask.build();
+ PlannerSettings plannerSettings =
PrelUtil.getPlannerSettings(joinRel.getCluster().getPlanner());
+ if (((int)plannerSettings.getStatisticsJoinCardinalityMode() & 4) == 4) {
+ /*
+ * The NDV for a multi-column GBY key past a join is determined as follows:
+ * GBY(s1, s2, s3) = CNDV(s1)*CNDV(s2)*CNDV(s3)
+ * where CNDV is determined as follows:
+ * A) If sX is present as a join column (sX = tX) CNDV(sX) = MIN(NDV(sX),
NDV(tX)) where X =1, 2, 3, etc
+ * B) Otherwise, based on independence assumption CNDV(sX) = NDV(sX)
+ */
+ Set<ImmutableBitSet> joinFiltersSet = new HashSet<>();
+ for (RexNode filter : joinFilters) {
+ final RelOptUtil.InputFinder inputFinder =
RelOptUtil.InputFinder.analyze(filter);
+ joinFiltersSet.add(inputFinder.inputBitSet.build());
+ }
+ for (int idx = 0; idx < groupKey.length(); idx++) {
+ if (groupKey.get(idx)) {
+ // GBY key is present in some filter - now try options A) and B) as
described above
+ double ndvSGby = 0;
+ ImmutableBitSet sGby = getSingleGbyKey(groupKey, idx);
+ if (sGby != null) {
+ for (ImmutableBitSet jFilter : joinFiltersSet) {
+ if (jFilter.contains(sGby)) {
+ // Found join condition containing this GBY key. Pick min NDV
across all columns in this join
+ for (int fidx : jFilter) {
+ if (fidx < left.getRowType().getFieldCount()) {
+ ndvSGby = Math.min(ndvSGby, mq.getDistinctRowCount(left,
ImmutableBitSet.of(fidx), leftPred));
+ } else {
+ ndvSGby = Math.min(ndvSGby, mq.getDistinctRowCount(right,
ImmutableBitSet.of(fidx-left.getRowType().getFieldCount()), rightPred));
+ }
+ }
+ break;
+ }
+ }
+ // Did not find it in any join condition(s)
+ if (ndvSGby == 0) {
+ for (int sidx : sGby) {
+ if (sidx < left.getRowType().getFieldCount()) {
+ ndvSGby = mq.getDistinctRowCount(left,
ImmutableBitSet.of(sidx), leftPred);
+ } else {
+ ndvSGby = mq.getDistinctRowCount(right,
ImmutableBitSet.of(sidx-left.getRowType().getFieldCount()), rightPred);
+ }
+ }
+ }
+ ++gbyCols;
+ // Multiply NDV(s) of different GBY cols to determine the overall
NDV
+ distRowCount *= ndvSGby *
plannerSettings.getStatisticsMultiColNdvAdjustmentFactor();
Review comment:
Done
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
