Package org.apache.hadoop.hive.ql.plan

Examples of org.apache.hadoop.hive.ql.plan.ColStatistics

379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
   *          - column name
   * @return ColStatistics
   */
  public static ColStatistics getColStatistics(ColumnStatisticsObj cso, String tabName,
      String colName) {
    ColStatistics cs = new ColStatistics(tabName, colName, cso.getColType());
    String colType = cso.getColType();
    ColumnStatisticsData csd = cso.getStatsData();
    if (colType.equalsIgnoreCase(serdeConstants.TINYINT_TYPE_NAME)
        || colType.equalsIgnoreCase(serdeConstants.SMALLINT_TYPE_NAME)
        || colType.equalsIgnoreCase(serdeConstants.INT_TYPE_NAME)) {
      cs.setCountDistint(csd.getLongStats().getNumDVs());
      cs.setNumNulls(csd.getLongStats().getNumNulls());
      cs.setAvgColLen(JavaDataModel.get().primitive1());
    } else if (colType.equalsIgnoreCase(serdeConstants.BIGINT_TYPE_NAME)) {
      cs.setCountDistint(csd.getLongStats().getNumDVs());
      cs.setNumNulls(csd.getLongStats().getNumNulls());
      cs.setAvgColLen(JavaDataModel.get().primitive2());
    } else if (colType.equalsIgnoreCase(serdeConstants.FLOAT_TYPE_NAME)) {
      cs.setCountDistint(csd.getDoubleStats().getNumDVs());
      cs.setNumNulls(csd.getDoubleStats().getNumNulls());
      cs.setAvgColLen(JavaDataModel.get().primitive1());
    } else if (colType.equalsIgnoreCase(serdeConstants.DOUBLE_TYPE_NAME)) {
      cs.setCountDistint(csd.getDoubleStats().getNumDVs());
      cs.setNumNulls(csd.getDoubleStats().getNumNulls());
      cs.setAvgColLen(JavaDataModel.get().primitive2());
    } else if (colType.equalsIgnoreCase(serdeConstants.STRING_TYPE_NAME)
        || colType.startsWith(serdeConstants.CHAR_TYPE_NAME)
        || colType.startsWith(serdeConstants.VARCHAR_TYPE_NAME)) {
      cs.setCountDistint(csd.getStringStats().getNumDVs());
      cs.setNumNulls(csd.getStringStats().getNumNulls());
      cs.setAvgColLen(csd.getStringStats().getAvgColLen());
    } else if (colType.equalsIgnoreCase(serdeConstants.BOOLEAN_TYPE_NAME)) {
      if (csd.getBooleanStats().getNumFalses() > 0 && csd.getBooleanStats().getNumTrues() > 0) {
        cs.setCountDistint(2);
      } else {
        cs.setCountDistint(1);
      }
      cs.setNumTrues(csd.getBooleanStats().getNumTrues());
      cs.setNumFalses(csd.getBooleanStats().getNumFalses());
      cs.setNumNulls(csd.getBooleanStats().getNumNulls());
      cs.setAvgColLen(JavaDataModel.get().primitive1());
    } else if (colType.equalsIgnoreCase(serdeConstants.BINARY_TYPE_NAME)) {
      cs.setAvgColLen(csd.getBinaryStats().getAvgColLen());
      cs.setNumNulls(csd.getBinaryStats().getNumNulls());
    } else if (colType.equalsIgnoreCase(serdeConstants.TIMESTAMP_TYPE_NAME)) {
      cs.setAvgColLen(JavaDataModel.get().lengthOfTimestamp());
    } else if (colType.startsWith(serdeConstants.DECIMAL_TYPE_NAME)) {
      cs.setAvgColLen(JavaDataModel.get().lengthOfDecimal());
    } else if (colType.equalsIgnoreCase(serdeConstants.DATE_TYPE_NAME)) {
      cs.setAvgColLen(JavaDataModel.get().lengthOfDate());
    } else {
      // Columns statistics for complex datatypes are not supported yet
      return null;
    }
    return cs;
View Full Code Here
454
455
456
457
458
459
460
461
462
463
464
465
    try {
      List<ColumnStatisticsObj> colStat = Hive.get().getTableColumnStatistics(
          dbName, tabName, neededColsInTable);
      stats = new ArrayList<ColStatistics>(colStat.size());
      for (ColumnStatisticsObj statObj : colStat) {
        ColStatistics cs = getColStatistics(statObj, tabName, statObj.getColName());
        cs.setTableAlias(colToTabAlias.get(cs.getColumnName()));
        stats.add(cs);
      }
    } catch (HiveException e) {
      LOG.error("Failed to retrieve table statistics: ", e);
      stats = null;
View Full Code Here
489
490
491
492
493
494
495
496
497
498
499
500
          dbName, tabName, partNames, neededColsInTable);
      stats = new HashMap<String, List<ColStatistics>>(colStat.size());
      for (Map.Entry<String, List<ColumnStatisticsObj>> entry : colStat.entrySet()) {
        List<ColStatistics> partStat = new ArrayList<ColStatistics>(entry.getValue().size());
        for (ColumnStatisticsObj statObj : entry.getValue()) {
          ColStatistics cs = getColStatistics(statObj, tabName, statObj.getColName());
          cs.setTableAlias(colToTabAlias.get(cs.getColumnName()));
          partStat.add(cs);
        }
        stats.put(entry.getKey(), partStat);
      }
    } catch (HiveException e) {
View Full Code Here
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
        ExprNodeDesc end = colExprMap.get(outColName);
        if (end == null) {
          outColName = StatsUtils.stripPrefixFromColumnName(outColName);
          end = colExprMap.get(outColName);
        }
        ColStatistics colStat = getColStatisticsFromExpression(conf, parentStats, end);
        if (colStat != null) {
          outColName = StatsUtils.stripPrefixFromColumnName(outColName);
          colStat.setColumnName(outColName);
          colStat.setTableAlias(outTabAlias);
        }
        cs.add(colStat);
      }
    }
    return cs;
View Full Code Here
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
        countDistincts = numRows;
        oi = encd.getWritableObjectInspector();
      } else {

        // clone the column stats and return
        ColStatistics result = parentStats.getColumnStatisticsForColumn(tabAlias, colName);
        if (result != null) {
          try {
            return result.clone();
          } catch (CloneNotSupportedException e) {
            return null;
          }
        }
        return null;
      }
    } else if (end instanceof ExprNodeConstantDesc) {

      // constant projection
      ExprNodeConstantDesc encd = (ExprNodeConstantDesc) end;

      // null projection
      if (encd.getValue() == null) {
        colName = encd.getName();
        colType = "null";
        numNulls = numRows;
      } else {
        colName = encd.getName();
        colType = encd.getTypeString();
        countDistincts = 1;
        oi = encd.getWritableObjectInspector();
      }
    } else if (end instanceof ExprNodeGenericFuncDesc) {

      // udf projection
      ExprNodeGenericFuncDesc engfd = (ExprNodeGenericFuncDesc) end;
      colName = engfd.getName();
      colType = engfd.getTypeString();
      countDistincts = numRows;
      oi = engfd.getWritableObjectInspector();
    } else if (end instanceof ExprNodeNullDesc) {

      // null projection
      ExprNodeNullDesc ennd = (ExprNodeNullDesc) end;
      colName = ennd.getName();
      colType = "null";
      numNulls = numRows;
    }

    if (colType.equalsIgnoreCase(serdeConstants.STRING_TYPE_NAME)
        || colType.equalsIgnoreCase(serdeConstants.BINARY_TYPE_NAME)
        || colType.startsWith(serdeConstants.VARCHAR_TYPE_NAME)
        || colType.startsWith(serdeConstants.CHAR_TYPE_NAME)
        || colType.startsWith(serdeConstants.LIST_TYPE_NAME)
        || colType.startsWith(serdeConstants.MAP_TYPE_NAME)
        || colType.startsWith(serdeConstants.STRUCT_TYPE_NAME)
        || colType.startsWith(serdeConstants.UNION_TYPE_NAME)) {
      avgColSize = getAvgColLenOfVariableLengthTypes(conf, oi, colType);
    } else {
      avgColSize = getAvgColLenOfFixedLengthTypes(colType);
    }

    ColStatistics colStats = new ColStatistics(tabAlias, colName, colType);
    colStats.setAvgColLen(avgColSize);
    colStats.setCountDistint(countDistincts);
    colStats.setNumNulls(numNulls);

    return colStats;
  }
View Full Code Here
654
655
656
657
658
659
660
661
662
663
664
665
666
667
            if (!colExprMap.containsKey(ci.getInternalName())) {
              String colName = ci.getInternalName();
              colName = StatsUtils.stripPrefixFromColumnName(colName);
              String tabAlias = ci.getTabAlias();
              String colType = ci.getTypeName();
              ColStatistics cs = new ColStatistics(tabAlias, colName, colType);
              cs.setCountDistint(stats.getNumRows());
              cs.setNumNulls(0);
              cs.setAvgColLen(StatsUtils.getAvgColLenOfFixedLengthTypes(colType));
              aggColStats.add(cs);
            }
          }
          stats.addToColumnStats(aggColStats);
View Full Code Here
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
            List<Long> perAttrDVs = Lists.newArrayList();
            int numAttr = joinKeys.get(0).size();
            for (int idx = 0; idx < numAttr; idx++) {
              for (Integer i : joinKeys.keySet()) {
                String col = joinKeys.get(i).get(idx);
                ColStatistics cs = joinedColStats.get(col);
                if (cs != null) {
                  perAttrDVs.add(cs.getCountDistint());
                }
              }
              distinctVals.add(getDenominator(perAttrDVs));
              perAttrDVs.clear();
            }

            for (Long l : distinctVals) {
              denom *= l;
            }
          } else {
            for (List<String> jkeys : joinKeys.values()) {
              for (String jk : jkeys) {
                ColStatistics cs = joinedColStats.get(jk);
                if (cs != null) {
                  distinctVals.add(cs.getCountDistint());
                }
              }
            }
            denom = getDenominator(distinctVals);
          }

          // column statistics from different sources are put together and rename
          // fully qualified column names based on output schema of join operator
          Map<String, ExprNodeDesc> colExprMap = jop.getColumnExprMap();
          RowSchema rs = jop.getSchema();
          List<ColStatistics> outColStats = Lists.newArrayList();
          for (ColumnInfo ci : rs.getSignature()) {
            String key = ci.getInternalName();
            ExprNodeDesc end = colExprMap.get(key);
            if (end instanceof ExprNodeColumnDesc) {
              String colName = ((ExprNodeColumnDesc) end).getColumn();
              colName = StatsUtils.stripPrefixFromColumnName(colName);
              String tabAlias = ((ExprNodeColumnDesc) end).getTabAlias();
              String fqColName = StatsUtils.getFullyQualifiedColumnName(tabAlias, colName);
              ColStatistics cs = joinedColStats.get(fqColName);
              String outColName = key;
              String outTabAlias = ci.getTabAlias();
              outColName = StatsUtils.stripPrefixFromColumnName(outColName);
              if (cs != null) {
                cs.setColumnName(outColName);
                cs.setTableAlias(outTabAlias);
              }
              outColStats.add(cs);
            }
          }

View Full Code Here
324
325
326
327
328
329
330
331
332
333
334
335
336
        ExprNodeColumnDesc encd = (ExprNodeColumnDesc) pred;
        String colName = encd.getColumn();
        String tabAlias = encd.getTabAlias();
        String colType = encd.getTypeString();
        if (colType.equalsIgnoreCase(serdeConstants.BOOLEAN_TYPE_NAME)) {
          ColStatistics cs = stats.getColumnStatisticsForColumn(tabAlias, colName);
          if (cs != null) {
            return cs.getNumTrues();
          }
        }

        // if not boolean column return half the number of rows
        return stats.getNumRows() / 2;
View Full Code Here
369
370
371
372
373
374
375
376
377
378
379
380
381
            ExprNodeColumnDesc encd = (ExprNodeColumnDesc) leaf;
            String colName = encd.getColumn();
            String tabAlias = encd.getTabAlias();
            String colType = encd.getTypeString();
            if (colType.equalsIgnoreCase(serdeConstants.BOOLEAN_TYPE_NAME)) {
              ColStatistics cs = stats.getColumnStatisticsForColumn(tabAlias, colName);
              if (cs != null) {
                return cs.getNumFalses();
              }
            }
            // if not boolean column return half the number of rows
            return numRows / 2;
          }
View Full Code Here
399
400
401
402
403
404
405
406
407
408
409
410
411
          if (leaf instanceof ExprNodeColumnDesc) {
            ExprNodeColumnDesc colDesc = (ExprNodeColumnDesc) leaf;
            String colName = colDesc.getColumn();
            String tabAlias = colDesc.getTabAlias();
            ColStatistics cs = stats.getColumnStatisticsForColumn(tabAlias, colName);
            if (cs != null) {
              long dvs = cs.getCountDistint();
              numRows = dvs == 0 ? numRows / 2 : numRows / dvs;
              return numRows;
            }
          }
        }
View Full Code Here
TOP

Related Classes of org.apache.hadoop.hive.ql.plan.ColStatistics

  • org.apache.hadoop.hive.ql.optimizer.optiq.RelOptHiveTable
  • org.apache.hadoop.hive.ql.optimizer.ReduceSinkMapJoinProc
  • org.apache.hadoop.hive.ql.optimizer.stats.annotation.StatsRulesProcFactory$FilterStatsRule
  • org.apache.hadoop.hive.ql.optimizer.stats.annotation.StatsRulesProcFactory$GroupByStatsRule
  • org.apache.hadoop.hive.ql.optimizer.stats.annotation.StatsRulesProcFactory$JoinStatsRule
  • org.apache.hadoop.hive.ql.optimizer.stats.annotation.StatsRulesProcFactory$ReduceSinkStatsRule
  • org.apache.hadoop.hive.ql.stats.StatsUtils
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.