Search in sources :

Example 1 with ColumnStatisticsData

use of org.apache.hadoop.hive.metastore.api.ColumnStatisticsData in project hive by apache.

the class StatObjectConverter method getTableColumnStatisticsObj.

public static ColumnStatisticsObj getTableColumnStatisticsObj(MTableColumnStatistics mStatsObj) {
    ColumnStatisticsObj statsObj = new ColumnStatisticsObj();
    statsObj.setColType(mStatsObj.getColType());
    statsObj.setColName(mStatsObj.getColName());
    String colType = mStatsObj.getColType().toLowerCase();
    ColumnStatisticsData colStatsData = new ColumnStatisticsData();
    if (colType.equals("boolean")) {
        BooleanColumnStatsData boolStats = new BooleanColumnStatsData();
        boolStats.setNumFalses(mStatsObj.getNumFalses());
        boolStats.setNumTrues(mStatsObj.getNumTrues());
        boolStats.setNumNulls(mStatsObj.getNumNulls());
        colStatsData.setBooleanStats(boolStats);
    } else if (colType.equals("string") || colType.startsWith("varchar") || colType.startsWith("char")) {
        StringColumnStatsData stringStats = new StringColumnStatsData();
        stringStats.setNumNulls(mStatsObj.getNumNulls());
        stringStats.setAvgColLen(mStatsObj.getAvgColLen());
        stringStats.setMaxColLen(mStatsObj.getMaxColLen());
        stringStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setStringStats(stringStats);
    } else if (colType.equals("binary")) {
        BinaryColumnStatsData binaryStats = new BinaryColumnStatsData();
        binaryStats.setNumNulls(mStatsObj.getNumNulls());
        binaryStats.setAvgColLen(mStatsObj.getAvgColLen());
        binaryStats.setMaxColLen(mStatsObj.getMaxColLen());
        colStatsData.setBinaryStats(binaryStats);
    } else if (colType.equals("bigint") || colType.equals("int") || colType.equals("smallint") || colType.equals("tinyint") || colType.equals("timestamp")) {
        LongColumnStatsData longStats = new LongColumnStatsData();
        longStats.setNumNulls(mStatsObj.getNumNulls());
        Long longHighValue = mStatsObj.getLongHighValue();
        if (longHighValue != null) {
            longStats.setHighValue(longHighValue);
        }
        Long longLowValue = mStatsObj.getLongLowValue();
        if (longLowValue != null) {
            longStats.setLowValue(longLowValue);
        }
        longStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setLongStats(longStats);
    } else if (colType.equals("double") || colType.equals("float")) {
        DoubleColumnStatsData doubleStats = new DoubleColumnStatsData();
        doubleStats.setNumNulls(mStatsObj.getNumNulls());
        Double doubleHighValue = mStatsObj.getDoubleHighValue();
        if (doubleHighValue != null) {
            doubleStats.setHighValue(doubleHighValue);
        }
        Double doubleLowValue = mStatsObj.getDoubleLowValue();
        if (doubleLowValue != null) {
            doubleStats.setLowValue(doubleLowValue);
        }
        doubleStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDoubleStats(doubleStats);
    } else if (colType.startsWith("decimal")) {
        DecimalColumnStatsData decimalStats = new DecimalColumnStatsData();
        decimalStats.setNumNulls(mStatsObj.getNumNulls());
        String decimalHighValue = mStatsObj.getDecimalHighValue();
        if (decimalHighValue != null) {
            decimalStats.setHighValue(createThriftDecimal(decimalHighValue));
        }
        String decimalLowValue = mStatsObj.getDecimalLowValue();
        if (decimalLowValue != null) {
            decimalStats.setLowValue(createThriftDecimal(decimalLowValue));
        }
        decimalStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDecimalStats(decimalStats);
    } else if (colType.equals("date")) {
        DateColumnStatsData dateStats = new DateColumnStatsData();
        dateStats.setNumNulls(mStatsObj.getNumNulls());
        Long highValue = mStatsObj.getLongHighValue();
        if (highValue != null) {
            dateStats.setHighValue(new Date(highValue));
        }
        Long lowValue = mStatsObj.getLongLowValue();
        if (lowValue != null) {
            dateStats.setLowValue(new Date(lowValue));
        }
        dateStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDateStats(dateStats);
    }
    statsObj.setStatsData(colStatsData);
    return statsObj;
}
Also used : BooleanColumnStatsData(org.apache.hadoop.hive.metastore.api.BooleanColumnStatsData) DateColumnStatsData(org.apache.hadoop.hive.metastore.api.DateColumnStatsData) StringColumnStatsData(org.apache.hadoop.hive.metastore.api.StringColumnStatsData) LongColumnStatsData(org.apache.hadoop.hive.metastore.api.LongColumnStatsData) BinaryColumnStatsData(org.apache.hadoop.hive.metastore.api.BinaryColumnStatsData) Date(org.apache.hadoop.hive.metastore.api.Date) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) DoubleColumnStatsData(org.apache.hadoop.hive.metastore.api.DoubleColumnStatsData) DecimalColumnStatsData(org.apache.hadoop.hive.metastore.api.DecimalColumnStatsData) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData)

Example 2 with ColumnStatisticsData

use of org.apache.hadoop.hive.metastore.api.ColumnStatisticsData in project hive by apache.

the class StatObjectConverter method getPartitionColumnStatisticsObj.

public static ColumnStatisticsObj getPartitionColumnStatisticsObj(MPartitionColumnStatistics mStatsObj) {
    ColumnStatisticsObj statsObj = new ColumnStatisticsObj();
    statsObj.setColType(mStatsObj.getColType());
    statsObj.setColName(mStatsObj.getColName());
    String colType = mStatsObj.getColType().toLowerCase();
    ColumnStatisticsData colStatsData = new ColumnStatisticsData();
    if (colType.equals("boolean")) {
        BooleanColumnStatsData boolStats = new BooleanColumnStatsData();
        boolStats.setNumFalses(mStatsObj.getNumFalses());
        boolStats.setNumTrues(mStatsObj.getNumTrues());
        boolStats.setNumNulls(mStatsObj.getNumNulls());
        colStatsData.setBooleanStats(boolStats);
    } else if (colType.equals("string") || colType.startsWith("varchar") || colType.startsWith("char")) {
        StringColumnStatsData stringStats = new StringColumnStatsData();
        stringStats.setNumNulls(mStatsObj.getNumNulls());
        stringStats.setAvgColLen(mStatsObj.getAvgColLen());
        stringStats.setMaxColLen(mStatsObj.getMaxColLen());
        stringStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setStringStats(stringStats);
    } else if (colType.equals("binary")) {
        BinaryColumnStatsData binaryStats = new BinaryColumnStatsData();
        binaryStats.setNumNulls(mStatsObj.getNumNulls());
        binaryStats.setAvgColLen(mStatsObj.getAvgColLen());
        binaryStats.setMaxColLen(mStatsObj.getMaxColLen());
        colStatsData.setBinaryStats(binaryStats);
    } else if (colType.equals("tinyint") || colType.equals("smallint") || colType.equals("int") || colType.equals("bigint") || colType.equals("timestamp")) {
        LongColumnStatsData longStats = new LongColumnStatsData();
        longStats.setNumNulls(mStatsObj.getNumNulls());
        if (mStatsObj.getLongHighValue() != null) {
            longStats.setHighValue(mStatsObj.getLongHighValue());
        }
        if (mStatsObj.getLongLowValue() != null) {
            longStats.setLowValue(mStatsObj.getLongLowValue());
        }
        longStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setLongStats(longStats);
    } else if (colType.equals("double") || colType.equals("float")) {
        DoubleColumnStatsData doubleStats = new DoubleColumnStatsData();
        doubleStats.setNumNulls(mStatsObj.getNumNulls());
        if (mStatsObj.getDoubleHighValue() != null) {
            doubleStats.setHighValue(mStatsObj.getDoubleHighValue());
        }
        if (mStatsObj.getDoubleLowValue() != null) {
            doubleStats.setLowValue(mStatsObj.getDoubleLowValue());
        }
        doubleStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDoubleStats(doubleStats);
    } else if (colType.startsWith("decimal")) {
        DecimalColumnStatsData decimalStats = new DecimalColumnStatsData();
        decimalStats.setNumNulls(mStatsObj.getNumNulls());
        if (mStatsObj.getDecimalHighValue() != null) {
            decimalStats.setHighValue(createThriftDecimal(mStatsObj.getDecimalHighValue()));
        }
        if (mStatsObj.getDecimalLowValue() != null) {
            decimalStats.setLowValue(createThriftDecimal(mStatsObj.getDecimalLowValue()));
        }
        decimalStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDecimalStats(decimalStats);
    } else if (colType.equals("date")) {
        DateColumnStatsData dateStats = new DateColumnStatsData();
        dateStats.setNumNulls(mStatsObj.getNumNulls());
        dateStats.setHighValue(new Date(mStatsObj.getLongHighValue()));
        dateStats.setLowValue(new Date(mStatsObj.getLongLowValue()));
        dateStats.setNumDVs(mStatsObj.getNumDVs());
        colStatsData.setDateStats(dateStats);
    }
    statsObj.setStatsData(colStatsData);
    return statsObj;
}
Also used : BooleanColumnStatsData(org.apache.hadoop.hive.metastore.api.BooleanColumnStatsData) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) DoubleColumnStatsData(org.apache.hadoop.hive.metastore.api.DoubleColumnStatsData) DecimalColumnStatsData(org.apache.hadoop.hive.metastore.api.DecimalColumnStatsData) DateColumnStatsData(org.apache.hadoop.hive.metastore.api.DateColumnStatsData) StringColumnStatsData(org.apache.hadoop.hive.metastore.api.StringColumnStatsData) LongColumnStatsData(org.apache.hadoop.hive.metastore.api.LongColumnStatsData) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) BinaryColumnStatsData(org.apache.hadoop.hive.metastore.api.BinaryColumnStatsData) Date(org.apache.hadoop.hive.metastore.api.Date)

Example 3 with ColumnStatisticsData

use of org.apache.hadoop.hive.metastore.api.ColumnStatisticsData in project hive by apache.

the class TestHiveMetaStoreStatsMerge method testStatsMerge.

public void testStatsMerge() throws Exception {
    int listSize = 0;
    List<ListenerEvent> notifyList = DummyListener.notifyList;
    assertEquals(notifyList.size(), listSize);
    msc.createDatabase(db);
    listSize++;
    assertEquals(listSize, notifyList.size());
    CreateDatabaseEvent dbEvent = (CreateDatabaseEvent) (notifyList.get(listSize - 1));
    assert dbEvent.getStatus();
    msc.createTable(table);
    listSize++;
    assertEquals(notifyList.size(), listSize);
    CreateTableEvent tblEvent = (CreateTableEvent) (notifyList.get(listSize - 1));
    assert tblEvent.getStatus();
    table = msc.getTable(dbName, tblName);
    ColumnStatistics cs = new ColumnStatistics();
    ColumnStatisticsDesc desc = new ColumnStatisticsDesc(true, dbName, tblName);
    cs.setStatsDesc(desc);
    ColumnStatisticsObj obj = new ColumnStatisticsObj();
    obj.setColName("a");
    obj.setColType("string");
    ColumnStatisticsData data = new ColumnStatisticsData();
    StringColumnStatsData scsd = new StringColumnStatsData();
    scsd.setAvgColLen(10);
    scsd.setMaxColLen(20);
    scsd.setNumNulls(30);
    scsd.setNumDVs(123);
    scsd.setBitVectors("{0, 4, 5, 7}{0, 1}{0, 1, 2}{0, 1, 4}{0}{0, 2}{0, 3}{0, 2, 3, 4}{0, 1, 4}{0, 1}{0}{0, 1, 3, 8}{0, 2}{0, 2}{0, 9}{0, 1, 4}");
    data.setStringStats(scsd);
    obj.setStatsData(data);
    cs.addToStatsObj(obj);
    List<ColumnStatistics> colStats = new ArrayList<>();
    colStats.add(cs);
    SetPartitionsStatsRequest request = new SetPartitionsStatsRequest(colStats);
    msc.setPartitionColumnStatistics(request);
    List<String> colNames = new ArrayList<>();
    colNames.add("a");
    StringColumnStatsData getScsd = msc.getTableColumnStatistics(dbName, tblName, colNames).get(0).getStatsData().getStringStats();
    assertEquals(getScsd.getNumDVs(), 123);
    cs = new ColumnStatistics();
    scsd = new StringColumnStatsData();
    scsd.setAvgColLen(20);
    scsd.setMaxColLen(5);
    scsd.setNumNulls(70);
    scsd.setNumDVs(456);
    scsd.setBitVectors("{0, 1}{0, 1}{1, 2, 4}{0, 1, 2}{0, 1, 2}{0, 2}{0, 1, 3, 4}{0, 1}{0, 1}{3, 4, 6}{2}{0, 1}{0, 3}{0}{0, 1}{0, 1, 4}");
    data.setStringStats(scsd);
    obj.setStatsData(data);
    cs.addToStatsObj(obj);
    request = new SetPartitionsStatsRequest(colStats);
    request.setNeedMerge(true);
    msc.setPartitionColumnStatistics(request);
    getScsd = msc.getTableColumnStatistics(dbName, tblName, colNames).get(0).getStatsData().getStringStats();
    assertEquals(getScsd.getAvgColLen(), 20.0);
    assertEquals(getScsd.getMaxColLen(), 20);
    assertEquals(getScsd.getNumNulls(), 100);
    // since metastore is ObjectStore, we use the max function to merge.
    assertEquals(getScsd.getNumDVs(), 456);
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) CreateDatabaseEvent(org.apache.hadoop.hive.metastore.events.CreateDatabaseEvent) ArrayList(java.util.ArrayList) StringColumnStatsData(org.apache.hadoop.hive.metastore.api.StringColumnStatsData) SetPartitionsStatsRequest(org.apache.hadoop.hive.metastore.api.SetPartitionsStatsRequest) ListenerEvent(org.apache.hadoop.hive.metastore.events.ListenerEvent) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) CreateTableEvent(org.apache.hadoop.hive.metastore.events.CreateTableEvent) ColumnStatisticsDesc(org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData)

Example 4 with ColumnStatisticsData

use of org.apache.hadoop.hive.metastore.api.ColumnStatisticsData in project hive by apache.

the class BinaryColumnStatsAggregator method aggregate.

@Override
public ColumnStatisticsObj aggregate(String colName, List<String> partNames, List<ColumnStatistics> css) throws MetaException {
    ColumnStatisticsObj statsObj = null;
    BinaryColumnStatsData aggregateData = null;
    String colType = null;
    for (ColumnStatistics cs : css) {
        if (cs.getStatsObjSize() != 1) {
            throw new MetaException("The number of columns should be exactly one in aggrStats, but found " + cs.getStatsObjSize());
        }
        ColumnStatisticsObj cso = cs.getStatsObjIterator().next();
        if (statsObj == null) {
            colType = cso.getColType();
            statsObj = ColumnStatsAggregatorFactory.newColumnStaticsObj(colName, colType, cso.getStatsData().getSetField());
        }
        BinaryColumnStatsData newData = cso.getStatsData().getBinaryStats();
        if (aggregateData == null) {
            aggregateData = newData.deepCopy();
        } else {
            aggregateData.setMaxColLen(Math.max(aggregateData.getMaxColLen(), newData.getMaxColLen()));
            aggregateData.setAvgColLen(Math.max(aggregateData.getAvgColLen(), newData.getAvgColLen()));
            aggregateData.setNumNulls(aggregateData.getNumNulls() + newData.getNumNulls());
        }
    }
    ColumnStatisticsData columnStatisticsData = new ColumnStatisticsData();
    columnStatisticsData.setBinaryStats(aggregateData);
    statsObj.setStatsData(columnStatisticsData);
    return statsObj;
}
Also used : ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) BinaryColumnStatsData(org.apache.hadoop.hive.metastore.api.BinaryColumnStatsData) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) MetaException(org.apache.hadoop.hive.metastore.api.MetaException)

Example 5 with ColumnStatisticsData

use of org.apache.hadoop.hive.metastore.api.ColumnStatisticsData in project hive by apache.

the class BooleanColumnStatsAggregator method aggregate.

@Override
public ColumnStatisticsObj aggregate(String colName, List<String> partNames, List<ColumnStatistics> css) throws MetaException {
    ColumnStatisticsObj statsObj = null;
    BooleanColumnStatsData aggregateData = null;
    String colType = null;
    for (ColumnStatistics cs : css) {
        if (cs.getStatsObjSize() != 1) {
            throw new MetaException("The number of columns should be exactly one in aggrStats, but found " + cs.getStatsObjSize());
        }
        ColumnStatisticsObj cso = cs.getStatsObjIterator().next();
        if (statsObj == null) {
            colType = cso.getColType();
            statsObj = ColumnStatsAggregatorFactory.newColumnStaticsObj(colName, colType, cso.getStatsData().getSetField());
        }
        BooleanColumnStatsData newData = cso.getStatsData().getBooleanStats();
        if (aggregateData == null) {
            aggregateData = newData.deepCopy();
        } else {
            aggregateData.setNumTrues(aggregateData.getNumTrues() + newData.getNumTrues());
            aggregateData.setNumFalses(aggregateData.getNumFalses() + newData.getNumFalses());
            aggregateData.setNumNulls(aggregateData.getNumNulls() + newData.getNumNulls());
        }
    }
    ColumnStatisticsData columnStatisticsData = new ColumnStatisticsData();
    columnStatisticsData.setBooleanStats(aggregateData);
    statsObj.setStatsData(columnStatisticsData);
    return statsObj;
}
Also used : BooleanColumnStatsData(org.apache.hadoop.hive.metastore.api.BooleanColumnStatsData) ColumnStatistics(org.apache.hadoop.hive.metastore.api.ColumnStatistics) ColumnStatisticsObj(org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj) ColumnStatisticsData(org.apache.hadoop.hive.metastore.api.ColumnStatisticsData) MetaException(org.apache.hadoop.hive.metastore.api.MetaException)

Aggregations

ColumnStatisticsData (org.apache.hadoop.hive.metastore.api.ColumnStatisticsData)127 ColumnStatisticsObj (org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj)104 ColumnStatistics (org.apache.hadoop.hive.metastore.api.ColumnStatistics)61 Test (org.junit.Test)57 ColumnStatisticsDesc (org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc)54 ArrayList (java.util.ArrayList)49 LongColumnStatsData (org.apache.hadoop.hive.metastore.api.LongColumnStatsData)39 BooleanColumnStatsData (org.apache.hadoop.hive.metastore.api.BooleanColumnStatsData)34 FieldSchema (org.apache.hadoop.hive.metastore.api.FieldSchema)34 Table (org.apache.hadoop.hive.metastore.api.Table)31 DoubleColumnStatsData (org.apache.hadoop.hive.metastore.api.DoubleColumnStatsData)30 StorageDescriptor (org.apache.hadoop.hive.metastore.api.StorageDescriptor)29 Partition (org.apache.hadoop.hive.metastore.api.Partition)28 SerDeInfo (org.apache.hadoop.hive.metastore.api.SerDeInfo)28 StringColumnStatsData (org.apache.hadoop.hive.metastore.api.StringColumnStatsData)28 AggrStats (org.apache.hadoop.hive.metastore.api.AggrStats)26 BinaryColumnStatsData (org.apache.hadoop.hive.metastore.api.BinaryColumnStatsData)26 DecimalColumnStatsData (org.apache.hadoop.hive.metastore.api.DecimalColumnStatsData)26 HashMap (java.util.HashMap)20 List (java.util.List)19