oracle.pg.rdbms

Class SQLAnalyticsGraph

java.lang.Object

oracle.pg.common.OraclePropertyGraphBase

oracle.pg.rdbms.SQLAnalyticsGraph

All Implemented Interfaces:

java.io.Closeable, java.lang.AutoCloseable, oracle.pg.common.messages.MesgConsts, OracleCSVDatatypeConstants, OracleGraphAttributeUpdater, OracleIndexableGraph, OracleMessageConstants, OracleMetadataGraph, OraclePropertyGraphConstants, OraclePropertyGraphDatatypeConstants, OraclePropertyGraphAnalytics, org.apache.tinkerpop.gremlin.structure.Graph, org.apache.tinkerpop.gremlin.structure.util.Host

Direct Known Subclasses:

OraclePropertyGraph

public abstract class SQLAnalyticsGraph
extends OraclePropertyGraphBase
implements OraclePropertyGraphAnalytics

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	SQLAnalyticsGraph.Sort This enumeration denotes the sorting order for function getTopK

Nested classes/interfaces inherited from class oracle.pg.common.OraclePropertyGraphBase

OraclePropertyGraphBase.CacheStatus, OraclePropertyGraphBase.OptimizationFlag

Nested classes/interfaces inherited from interface org.apache.tinkerpop.gremlin.structure.Graph

org.apache.tinkerpop.gremlin.structure.Graph.Exceptions, org.apache.tinkerpop.gremlin.structure.Graph.Features, org.apache.tinkerpop.gremlin.structure.Graph.Hidden, org.apache.tinkerpop.gremlin.structure.Graph.OptIn, org.apache.tinkerpop.gremlin.structure.Graph.OptIns, org.apache.tinkerpop.gremlin.structure.Graph.OptOut, org.apache.tinkerpop.gremlin.structure.Graph.OptOuts, org.apache.tinkerpop.gremlin.structure.Graph.Variables

Nested classes/interfaces inherited from interface oracle.pg.common.OraclePropertyGraphConstants

OraclePropertyGraphConstants.QueueAction

Field Summary

Fields inherited from interface org.apache.tinkerpop.gremlin.structure.Graph

GRAPH

Fields inherited from interface oracle.pg.common.OracleMessageConstants

MG_ERR_USER_REQ_OP_CANCEL, MSG_E_AUTO_IDX_MUST_SOLR_OR_LUCENE, MSG_E_AUTO_IDX_NOT_EXIST, MSG_ERR_ARG_CANNOT_BE_NULL, MSG_ERR_ID_CANNOT_BE_NULL, MSG_ERR_NO_ELEM_TO_BE_CONSUMED, MSG_ERR_NO_OP_SUPPORTED, MSG_IDX_KEY_NOT_NULL, MSG_IDX_KEYS_NOT_NULL, MSG_TXT_IDX_NOT_FOUND, MSG_V_AUTO_IDX_MUST_SOLR_OR_LUCENE, MSG_V_AUTO_IDX_NOT_EXIST

Fields inherited from interface oracle.pg.common.messages.MesgConsts

ERR_ARG_CANNOT_BE_NULL, ERR_CODE_NOT_IN_LIST, ERR_DOP_CONSTRAINT, ERR_E_AUTO_IDX_MUST_SOLR_OR_LUCENE, ERR_E_AUTO_IDX_NOT_EXIST, ERR_E_DS_MUSTBE_FILE_OR_IS, ERR_E_MAX_LINE_CONSTRAINT, ERR_EDGE_DIR_NOT_BOTH, ERR_EDGE_WORK_TBL_NOT_CLEAN, ERR_GRAPH_NAME_CANT_BENULL, ERR_HIT_EXCEPTION, ERR_HIT_INTERRUPT, ERR_HIT_IO, ERR_HIT_OPGE, ERR_HIT_PARSEE, ERR_ID_CANNOT_BE_NULL, ERR_IDX_KEY_CANNOT_BE_NULL, ERR_IDX_KEY_NOT_NULL, ERR_IDX_KEYS_NOT_NULL, ERR_IN_DS_EDGE_CANTBE_NULL, ERR_IN_DS_VERT_CANTBE_NULL, ERR_IN_E_FLAT_DOESNT_EXIST, ERR_IN_V_FLAT_DOESNT_EXIST, ERR_INPUT_LIST_NOT_NULL, ERR_INPUT_NOT_NULL, ERR_INVALID_CHANGE_ACTION, ERR_INVALID_NUM_FIELDS, ERR_INVALID_NUM_NOARG, ERR_INVALID_NUMBER, ERR_INVALID_VAL, ERR_KEY_CANNOT_BE_ID, ERR_KEY_CANNOT_BE_NULL, ERR_KEY_CANNOT_BE_NULL_EMPTY, ERR_KEY_NOT_SET_TO, ERR_NO_ELEM_TO_BE_CONSUMED, ERR_NO_OP_SUPPORTED, ERR_NOT_IMPLEMENTED_YET, ERR_NOT_ORACLE_V_E, ERR_NUM_PART_CONSTRAINT, ERR_OFFSET_CONSTRAINT, ERR_OFFSET_E_MUST_BE_POS, ERR_OFFSET_V_MUST_BE_POS, ERR_OPT_FLAG_EDGE_INVALID, ERR_OPT_FLAG_VERT_INVALID, ERR_ORA_TEXT_INVALID_KEY, ERR_ORA_TEXT_ONLY_AUTO, ERR_OUTPUT_LIST_NOT_NULL, ERR_PG_ALREADY_EXISTS, ERR_PG_NOT_EMPTY, ERR_PG_NOT_NULL, ERR_SIZE_MUST_BE_POS, ERR_TIMEOUT_TAB_CREATION, ERR_TXT_IDX_NOT_FOUND, ERR_TYPE_ID_NOT_RECO, ERR_UNSUPPORTED_VAL_TYP, ERR_USER_REQ_OP_CANCEL, ERR_V_AUTO_IDX_MUST_SOLR_OR_LUCENE, ERR_V_AUTO_IDX_NOT_EXIST, ERR_V_DS_MUSTBE_FILE_OR_IS, ERR_V_MAX_LINE_CONSTRAINT, ERR_VALUE_CANNOT_BE_NULL, INFO_EMP_K_NO_P_VAL, INFO_NUL_K_NO_P_VAL

Fields inherited from interface oracle.pg.common.OraclePropertyGraphConstants

VERTEX_LABEL

Fields inherited from interface oracle.pg.common.OracleCSVDatatypeConstants

CSV_DATE_TIME_FORMATS, CSV_DT_KEYWORD_BOOLEAN, CSV_DT_KEYWORD_BYTE, CSV_DT_KEYWORD_CHAR, CSV_DT_KEYWORD_DATE, CSV_DT_KEYWORD_DOUBLE, CSV_DT_KEYWORD_FLOAT, CSV_DT_KEYWORD_INTEGER, CSV_DT_KEYWORD_JSON, CSV_DT_KEYWORD_LONG, CSV_DT_KEYWORD_RDF, CSV_DT_KEYWORD_SHORT, CSV_DT_KEYWORD_SPATIAL, CSV_DT_KEYWORD_STRING, OPG_DATETIME_FORMAT

Fields inherited from interface oracle.pg.common.OraclePropertyGraphDatatypeConstants

TYPE_DT_BOOL, TYPE_DT_BYTE, TYPE_DT_CHAR, TYPE_DT_DATE, TYPE_DT_DOUBLE, TYPE_DT_EMPTY, TYPE_DT_FLOAT, TYPE_DT_INTEGER, TYPE_DT_JSON, TYPE_DT_LONG, TYPE_DT_SERI, TYPE_DT_SHORT, TYPE_DT_SPATIAL, TYPE_DT_STRING, TYPE_DT_URI

Constructor Summary

Constructors

Constructor and Description
SQLAnalyticsGraph()
SQLAnalyticsGraph(PgRdbmsGraphConfig config)

Method Summary

Modifier and Type	Method and Description
void	analyze(float estimatePercent, java.lang.String methodOpt) Collects statistics for this graph.
void	analyze(float estimatePercent, java.lang.String methodOpt, int dop) Collects statistics for this graph.
void	analyze(int dop) Collects statistics for this graph.
void	analyzeAuto(int dop) Collects statistics for this graph.
void	collaborativeFiltering(java.lang.String szEdgeTabName, java.lang.String szLabel, java.lang.String szRatingProperty, int iterations, double minError, int features, double learningRate, double decreaseRate, double regularizationParam, int dop, java.lang.StringBuilder sbLeft, java.lang.StringBuilder sbRight, java.lang.StringBuilder sbLeft1, java.lang.StringBuilder sbRight1, java.lang.StringBuilder sbMatrixProduct, java.lang.StringBuilder sbLeftDelta, java.lang.StringBuilder sbRightDelta, java.lang.String szTBS, java.lang.String szOptions) This method performs collaborative filtering using matrix factorization on the graph.
java.lang.Double	collaborativeFilteringCheckIntermediateError(double regularizationParam, int dop, java.lang.StringBuilder sbMatrixProduct) This method checks the intermediate error in the collaborative filtering using matrix factorization on the graph.
void	collaborativeFilteringCleanup(java.lang.String szEdgeTabName, java.lang.StringBuilder sbLeft, java.lang.StringBuilder sbRight, java.lang.StringBuilder sbLeft1, java.lang.StringBuilder sbRight1, java.lang.StringBuilder sbMatrixProduct, java.lang.StringBuilder sbLeftDelta, java.lang.StringBuilder sbRightDelta, java.lang.String szOptions) This method does cleanup work after running graph collaborative filtering.
java.lang.Double	collaborativeFilteringGetRecommendation(java.lang.String szEdgeTabName, int dop, long srcVertex, long dstVertex, java.lang.StringBuilder sbLeft, java.lang.StringBuilder sbRight) This method gets the recommendation result of the collaborative filtering using matrix factorization on the graph.
void	collaborativeFilteringPrep(java.lang.String szEdgeTabName, java.lang.StringBuilder sbLeft, java.lang.StringBuilder sbRight, java.lang.StringBuilder sbLeft1, java.lang.StringBuilder sbRight1, java.lang.StringBuilder sbMatrixProduct, java.lang.StringBuilder sbLeftDelta, java.lang.StringBuilder sbRightDelta, java.lang.String szOptions) This method does preparation work before running the collaborative filtering on the graph.
static long	countTriangle(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.StringBuilder sbUndirWorkTab, int iNumSubPtns, int iDOP, java.lang.String szTBS, java.lang.String szOptions) Returns the number of triangles that exist in this graph.
long	countTriangle(java.lang.StringBuilder sbUndirWorkTab) Returns the number of triangles that exist in this graph which is treated as undirected.
long	countTriangle(java.lang.StringBuilder sbUndirWorkTab, int iNumSubPtns, int iDOP, java.lang.String szTBS, java.lang.String szOptions) Returns the number of triangles that exist in this graph.
void	countTriangleCleanup(java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, java.lang.String szOptions) This method does cleanup work after running triangle calculations.
void	countTrianglePrep(java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, java.lang.String szOptions) This method does preparation work before running triangle calculations.
static long	countTriangleRenum(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, int iNumSubPtns, int iDOP, java.lang.String szTBS, java.lang.String szOptions) Returns the number of triangles that exist in this graph (treated as undirected).
long	countTriangleRenum(java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, int iNumSubPtns, int iDOP, java.lang.String szTBS, java.lang.String szOptions) Returns the number of triangles that exist in this graph (treated as undirected).
void	createStatTable(java.lang.String statTableName) Creates a stat table using DBMS_STATS package.
static long	estimateTriangleRenum(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, int iNumSubPtns, int iChunk, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method estimates the number of triangles that exist in this graph (treated as undirected).
long	estimateTriangleRenum(java.lang.StringBuilder sbUndirBMTab, java.lang.StringBuilder sbMapWorkTab, java.lang.StringBuilder sbUndirAMTab, int iNumSubPtns, int iChunk, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method estimates the number of triangles that exist in this graph (treated as undirected).
void	expEdgeStat(java.lang.String statTableName, java.lang.String statid) Exports the edge table statistics into the given stat table.
void	expVertexStat(java.lang.String statTableName, java.lang.String statid) Exports the vertex table statistics into the given stat table.
static void	findCluster(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.StringBuilder sbClusters, java.lang.StringBuilder sbUndir, java.lang.StringBuilder sbCluas, java.lang.StringBuilder sbNewas, java.lang.StringBuilder sbDelta, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method finds clusters in a given graph (treated as undirected).
void	findCluster(java.lang.StringBuilder sbClusters, java.lang.StringBuilder sbUndir, java.lang.StringBuilder sbCluas, java.lang.StringBuilder sbNewas, java.lang.StringBuilder sbDelta, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method finds clusters in this property graph (treated as undirected).
void	findClusterCleanup(java.lang.StringBuilder sbClusters, java.lang.StringBuilder sbUndir, java.lang.StringBuilder sbCluas, java.lang.StringBuilder sbNewas, java.lang.StringBuilder sbDelta, java.lang.String szOptions) This method does cleanup work after running graph cluster detection.
void	findClusterPrep(java.lang.StringBuilder sbClusters, java.lang.StringBuilder sbUndir, java.lang.StringBuilder sbCluas, java.lang.StringBuilder sbNewas, java.lang.StringBuilder sbDelta, java.lang.String szOptions) This method does preparation work before running graph cluster detection.
abstract java.lang.String	getEdgeTabName()
abstract Oracle	getOracle()
static java.sql.ResultSet	getTopK(java.sql.Connection conn, java.lang.StringBuilder sbTabName, int iColIdx, SQLAnalyticsGraph.Sort sort, int k, int iDOP) Returns the top K rows of a given table.
java.sql.ResultSet	getTopK(java.lang.StringBuilder sbTabName, int iColIdx, SQLAnalyticsGraph.Sort sort, int k) Returns the top K rows of a given table.
java.sql.ResultSet	getTopK(java.lang.StringBuilder sbTabName, int iColIdx, SQLAnalyticsGraph.Sort sort, int k, int iDOP) Returns the top K rows of a given table.
void	impEdgeStat(java.lang.String statTableName, java.lang.String statid) Imports the edge table statistics from the given stat table.
void	impVertexStat(java.lang.String statTableName, java.lang.String statid) Imports the vertex table statistics from the given stat table.
static long	pageRank(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.StringBuilder sbNodePR, java.lang.StringBuilder sbNodeNextPR, java.lang.StringBuilder sbEdgeTabDeg, java.lang.StringBuilder sbDelta, float d, int iNumIterations, float convergence, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method calculates the page rank values of vertices in the given graph.
long	pageRank(java.lang.StringBuilder sbNodePR, java.lang.StringBuilder sbNodeNextPR, java.lang.StringBuilder sbEdgeTabDeg, java.lang.StringBuilder sbDelta, float d, int iNumIterations, float convergence, int iDOP, java.lang.String szTBS, java.lang.String szOptions) This method calculates the page rank values of vertices in this graph.
void	pageRankCleanup(java.lang.StringBuilder sbNodePR, java.lang.StringBuilder sbNodeNextPR, java.lang.StringBuilder sbEdgeTabDeg, java.lang.StringBuilder sbDelta, java.lang.String szOptions) This method does cleanup work before running page rank calculation.
void	pageRankPrep(java.lang.StringBuilder sbNodePR, java.lang.StringBuilder sbNodeNextPR, java.lang.StringBuilder sbEdgeTabDeg, java.lang.StringBuilder sbDelta, java.lang.String szOptions) This method does preparation work before running page rank calculation.
static java.lang.String	shortestPath(java.sql.Connection conn, java.lang.String szEdgeTabName, java.lang.Long lStart, java.lang.Long lEnd, java.lang.StringBuilder sbExpTab, java.lang.StringBuilder sbPathWeights, java.lang.String szOptions, java.lang.Long lSCN) Returns a shortest path in a string format.
java.lang.String	shortestPath(java.lang.Long lStart, java.lang.Long lEnd, java.lang.StringBuilder sbExpTab) Returns a shortest path in a string format.
java.lang.String	shortestPath(java.lang.Long lStart, java.lang.Long lEnd, java.lang.StringBuilder sbExpTab, java.lang.StringBuilder sbPathWeights, java.lang.String szOptions) Returns a shortest path in a string format.
java.lang.String	shortestPath(java.lang.Long lStart, java.lang.Long lEnd, java.lang.StringBuilder sbExpTab, java.lang.String szEdgeTabName, java.lang.StringBuilder sbPathWeights, java.lang.String szOptions, java.lang.Long lSCN) Returns a shortest path in a string format.
java.lang.String	shortestPath(OracleVertex vStart, OracleVertex vEnd, java.lang.StringBuilder sbExpTab) Returns a shortest path in a string format.
java.lang.String	shortestPath(OracleVertex vStart, OracleVertex vEnd, java.lang.StringBuilder sbExpTab, java.lang.StringBuilder sbPathWeights, java.lang.String szOptions) Returns a shortest path in a string format.
void	shortestPathCleanup(java.lang.StringBuilder sbExpTab, java.lang.String szOptions) This method does clean up after running shortest path calculations.
void	shortestPathPrep(java.lang.StringBuilder sbExpTab, java.lang.String szOptions) This method does some preparation work before running one or multiple shortest path calculations.
static void	sparsifyGraph(java.sql.Connection conn, java.lang.String szEdgeTabName, float fThreshold, int iMinKeep, int iDOP, java.lang.StringBuilder sbOutTab, java.lang.StringBuilder sbUndTab, java.lang.StringBuilder sbHshTab, java.lang.StringBuilder sbMchTab, java.lang.String szTBS, java.lang.String szOptions) This method sparsifies a given graph (treated as undirected).
void	sparsifyGraph(float fThreshold, int iMinKeep, int iDOP, java.lang.StringBuilder sbOutTab, java.lang.StringBuilder sbUndTab, java.lang.StringBuilder sbHshTab, java.lang.StringBuilder sbMchTab, java.lang.String szTBS, java.lang.String szOptions) This method sparsifies this graph instance (treated as undirected).
void	sparsifyGraphCleanup(java.lang.StringBuilder sbOutTab, java.lang.StringBuilder sbUndTab, java.lang.StringBuilder sbHshTab, java.lang.StringBuilder sbMchTab, java.lang.String szOptions) This method does cleanup work before running graph sparsification.
void	sparsifyGraphPrep(java.lang.StringBuilder sbOutTab, java.lang.StringBuilder sbUndTab, java.lang.StringBuilder sbHshTab, java.lang.StringBuilder sbMchTab, java.lang.String szOptions) This method does preparation work before running graph sparsification.

Methods inherited from class oracle.pg.common.OraclePropertyGraphBase

addEdge, addUpdatedEdge, addUpdatedVertex, addVertex, addVertex, addVertexById, adjacentEdgesFromVertices, adjacentEdgesFromVertices, clearCache, clearRepository, clearUpdateQueue, commit, commitIndices, compute, compute, configuration, createIndex, decodeObject, edges, edges, edgesByIds, edgesByIds, edgesByProperty, edgesByProperty, edgesByProperty, edgesByProperty, edgesByProperty, edgesByPropertyKey, edgesByPropertyKey, edgesByPropertyKey, edgesByPropertyKey, getBatchSize, getCacheStatus, getConfig, getDatatypeClassFromValueObjForTextIndex, getDateTimeFormatter, getDefaultEdgeOptFlag, getDefaultIndexParameters, getDefaultVertexOptFlag, getEdge, getEdgeFilterCallback, getEdgeInstance, getEdgeInstance, getElementsProperties, getGraphName, getJustEdgeIdOptFlag, getJustVertexIdOptFlag, getMaxLabelLengthAllowedInBytes, getMaxPropertyNameLengthAllowedInBytes, getMaxValueLengthAllowedInBytes, getOperationQueueSize, getOracleIndexManager, getProgressListener, getQueueSize, getShortestPath, getShortestPath, getSimpleDateFormat, getStringForObj, getTextIndexDirectoryPrefix, getVertex, getVertexFilterCallback, getVertexInstance, io, isBuiltinElementsCacheDisabled, isDisposed, isEmpty, isShutdown, isSkipRefreshIndicesEnabled, isTransientPropertyGraphInstance, removeEdge, removeEdges, removeEdges, removeEdges, removeEdges, removeVertex, removeVertices, removeVertices, removeVertices, removeVertices, rollback, serializableToStr, setBatchSize, setCacheStatus, setDefaultEdgeOptFlag, setDefaultIndexParameters, setDefaultVertexOptFlag, setEdgeFilterCallback, setProgressListener, setQueueSize, setSkipRefreshIndices, setSolrJavaBinFormat, setTransientPropertyGraphInstance, setVertexFilterCallback, startTransaction, strToSerializable, sum, supportVertexLabels, toString, traversal, traversal, tx, useSolrJavaBinFormat, variables, vertices, vertices, verticesByIds, verticesByIds, verticesByProperty, verticesByProperty, verticesByProperty, verticesByProperty, verticesByProperty, verticesByPropertyKey, verticesByPropertyKey, verticesByPropertyKey, verticesByPropertyKey

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface org.apache.tinkerpop.gremlin.structure.Graph

addVertex, close, features

Methods inherited from interface org.apache.tinkerpop.gremlin.structure.util.Host

getHostingVertex

Methods inherited from interface oracle.pg.common.OracleIndexableGraph

createKeyIndex, createKeyIndex, createKeyIndex, dropAllAutoIndices, dropAllIndices, dropAllManualIndices, dropIndex, dropKeyIndex, dropKeyIndex, dropKeyIndex, edgesUsingTextSearch, getAutoIndex, getIndex, getIndexedKeys, getIndexParameters, getIndices, isEdgeAutoIndexEnabled, isVertexAutoIndexEnabled, refreshAutoIndices, validateIndexManager, verticesUsingTextSearch

Methods inherited from interface oracle.pg.common.OracleMetadataGraph

countEdges, countEdges, countEdges, countVertices, countVertices, countVertices, getEdgePropertyNames, getEdgePropertyNames, getEdgePropertyNames, getEdgePropertyNames, getElementsProperties, getInvalidEdgeIDSet, getInvalidEdgeIDSet, getInvalidEdgeIDSet, getMaxEdgeID, getMaxEdgeID, getMaxEdgeID, getMaxVertexID, getMaxVertexID, getMaxVertexID, getMinEdgeID, getMinEdgeID, getMinEdgeID, getMinVertexID, getMinVertexID, getMinVertexID, getVertexIDSet, getVertexIDSet, getVertexIDSet, getVertexPropertyNames, getVertexPropertyNames, getVertexPropertyNames, getVertexPropertyNames

Methods inherited from interface oracle.pg.common.OracleGraphAttributeUpdater

addAttributeToAllEdges, addAttributeToAllEdges, addAttributeToAllEdges, addAttributeToAllVertices, addAttributeToAllVertices, addAttributeToAllVertices, getDOPForMetadataOperations, removeAttributeFromAllEdges, removeAttributeFromAllEdges, removeAttributeFromAllEdges, removeAttributeFromAllVertices, removeAttributeFromAllVertices, removeAttributeFromAllVertices, updateGraph

Methods inherited from interface java.io.Closeable

close

Constructor Detail

SQLAnalyticsGraph

public SQLAnalyticsGraph()

SQLAnalyticsGraph

public SQLAnalyticsGraph(PgRdbmsGraphConfig config)

Method Detail

analyze

public void analyze(float estimatePercent,
                    java.lang.String methodOpt)
             throws java.sql.SQLException

Collects statistics for this graph. Behind the scenes, Oracle database will invoke DBMS_STATS package and collect statistics for the underlying vertex and edge tables of this graph.

It is very important that up-to-date statistics be maintained for a graph to ensure efficient execution of query. The SQL Optimizer needs accurate information to choose a good execution plan.

After a significant amount of change has been done to a graph, this method should be invoked. Notice that however, statistics collection is a time/resource consuming process, so only run it when necessary.

See Oracle documentation "PL/SQL Packages and Types Reference" for detailed explanations of the parameters.

Parameters:

estimatePercent - has a valid range of [0.000001,100].

methodOpt - A typical value for methodOpt is "FOR ALL COLUMNS SIZE AUTO"

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

analyze

public void analyze(float estimatePercent,
                    java.lang.String methodOpt,
                    int dop)
             throws java.sql.SQLException

Collects statistics for this graph. Behind the scenes, Oracle database will invoke DBMS_STATS package and collect statistics for the underlying vertex and edge tables of this graph.

It is very important that up-to-date statistics be maintained for a graph to ensure efficient execution of query. The SQL Optimizer needs accurate information to choose a good execution plan.

After a significant amount of change has been done to a graph, this method should be invoked. Notice that however, statistics collection is a time/resource consuming process, so only run it when necessary.

See Oracle documentation "PL/SQL Packages and Types Reference" for detailed explanations of the parameters.

Parameters:

estimatePercent - has a valid range of [0.000001,100].

methodOpt - A typical value for methodOpt is "FOR ALL COLUMNS SIZE AUTO"

dop - degree of parallelism.

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

analyze

public void analyze(int dop)
             throws java.sql.SQLException

Collects statistics for this graph. Behind the scenes, Oracle database will invoke DBMS_STATS package and collect statistics for the underlying vertex and edge tables of this graph.

It is very important that up-to-date statistics be maintained for a graph to ensure efficient execution of query. The SQL Optimizer needs accurate information to choose a good execution plan.

After a significant amount of change has been done to a graph, this method should be invoked. Notice that however, statistics collection is a time/resource consuming process, so only run it when necessary.

See Oracle documentation "PL/SQL Packages and Types Reference" for detailed explanations of the parameters.

Parameters:

dop - degree of parallelism.

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

analyzeAuto

public void analyzeAuto(int dop)
                 throws java.sql.SQLException

Collects statistics for this graph. Behind the scenes, Oracle database will invoke DBMS_STATS package and collect statistics for the underlying vertex and edge tables of this graph.

It is very important that up-to-date statistics be maintained for a graph to ensure efficient execution of query. The SQL Optimizer needs accurate information to choose a good execution plan.

After a significant amount of change has been done to a graph, this method should be invoked. Notice that however, statistics collection is a time/resource consuming process, so only run it when necessary.

See Oracle documentation "PL/SQL Packages and Types Reference" for detailed explanations of the parameters.

Parameters:

dop - degree of parallelism.

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

collaborativeFiltering

public void collaborativeFiltering(java.lang.String szEdgeTabName,
                                   java.lang.String szLabel,
                                   java.lang.String szRatingProperty,
                                   int iterations,
                                   double minError,
                                   int features,
                                   double learningRate,
                                   double decreaseRate,
                                   double regularizationParam,
                                   int dop,
                                   java.lang.StringBuilder sbLeft,
                                   java.lang.StringBuilder sbRight,
                                   java.lang.StringBuilder sbLeft1,
                                   java.lang.StringBuilder sbRight1,
                                   java.lang.StringBuilder sbMatrixProduct,
                                   java.lang.StringBuilder sbLeftDelta,
                                   java.lang.StringBuilder sbRightDelta,
                                   java.lang.String szTBS,
                                   java.lang.String szOptions)
                            throws java.sql.SQLException

This method performs collaborative filtering using matrix factorization on the graph. The resulting factors of the matrix product will be stored on the left and right tables.

Specified by:

collaborativeFiltering in interface OraclePropertyGraphAnalytics

Parameters:

szEdgeTabName - a String with the name of the graph edges table.

szLabel - the label of the edges that hold the rating property.

szRatingProperty - the name of the rating property.

iterations - the maximum number of iterations.

minError - the minimum error to reach to finish the iterations.

features - the number of features for the left and right side products.

learningRate - the learning rate for the gradient descent.

decreaseRate - the decrease rate if the learning rate is too big for the intermediate error.

regularizationParam - an additional parameter to avoid over fitting.

dop - degree of paralellism.

sbLeft - a place holder for the name of the working table that holds the left side factor of the matrix factorization.

sbRight - a place holder for the name of the working table that holds the right side factor of the matrix factorization.

sbLeft1 - a place holder for the name of the working table that holds the left side intermediate step in the gradient descent.

sbRight1 - a place holder for the name of the working table that holds the right side intermediate step in the gradient descent.

sbMatrixProduct - a place holder for the name of the working table that holds intermediate matrix product.

sbLeftDelta - a place holder for the name of the working table that holds intermediate left side delta in gradient descent.

sbRightDelta - a place holder for the name of the working table that holds intermediate right side delta in gradient descent.

szTBS - the name of the tablespace. It can be null.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

collaborativeFilteringCheckIntermediateError

public java.lang.Double collaborativeFilteringCheckIntermediateError(double regularizationParam,
                                                                     int dop,
                                                                     java.lang.StringBuilder sbMatrixProduct)
                                                              throws java.sql.SQLException

This method checks the intermediate error in the collaborative filtering using matrix factorization on the graph.

Parameters:

regularizationParam - the additional parameter to avoid over fitting.

dop - degree of paralellism.

sbMatrixProduct - a place holder for the name of the working table that holds intermediate matrix product.

Throws:

java.sql.SQLException

collaborativeFilteringCleanup

public void collaborativeFilteringCleanup(java.lang.String szEdgeTabName,
                                          java.lang.StringBuilder sbLeft,
                                          java.lang.StringBuilder sbRight,
                                          java.lang.StringBuilder sbLeft1,
                                          java.lang.StringBuilder sbRight1,
                                          java.lang.StringBuilder sbMatrixProduct,
                                          java.lang.StringBuilder sbLeftDelta,
                                          java.lang.StringBuilder sbRightDelta,
                                          java.lang.String szOptions)
                                   throws java.sql.SQLException

This method does cleanup work after running graph collaborative filtering. All the tables that hold intermediate data and the output matrix factors will be dropped.

Specified by:

collaborativeFilteringCleanup in interface OraclePropertyGraphAnalytics

Parameters:

szEdgeTabName - a String with the name of the graph edges table

sbLeft - a place holder for the name of the working table that holds the left side factor of the matrix factorization

sbRight - a place holder for the name of the working table that holds the right side factor of the matrix factorization

sbLeft1 - a place holder for the name of the working table that holds the left side intermediate step in the gradient descent

sbRight1 - a place holder for the name of the working table that holds the right side intermediate step in the gradient descent

sbMatrixProduct - a place holder for the name of the working table that holds intermediate matrix product

sbLeftDelta - a place holder for the name of the working table that holds intermediate left side delta in gradient descent

sbRightDelta - a place holder for the name of the working table that holds intermediate right side delta in gradient descent

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

collaborativeFilteringGetRecommendation

public java.lang.Double collaborativeFilteringGetRecommendation(java.lang.String szEdgeTabName,
                                                                int dop,
                                                                long srcVertex,
                                                                long dstVertex,
                                                                java.lang.StringBuilder sbLeft,
                                                                java.lang.StringBuilder sbRight)
                                                         throws java.sql.SQLException

This method gets the recommendation result of the collaborative filtering using matrix factorization on the graph.

Parameters:

szEdgeTabName - a String with the name of the graph edges table.

dop - degree of paralellism.

srcVertex - the source vertex to the make the recommendation for (e.g. a customer vertex)

dstVertex - the destination vertex of the recommendation (e.g. a product vertex)

sbLeft - a place holder for the name of the working table that holds the left side factor of the matrix factorization.

sbRight - a place holder for the name of the working table that holds the right side factor of the matrix factorization.

Throws:

java.sql.SQLException

collaborativeFilteringPrep

public void collaborativeFilteringPrep(java.lang.String szEdgeTabName,
                                       java.lang.StringBuilder sbLeft,
                                       java.lang.StringBuilder sbRight,
                                       java.lang.StringBuilder sbLeft1,
                                       java.lang.StringBuilder sbRight1,
                                       java.lang.StringBuilder sbMatrixProduct,
                                       java.lang.StringBuilder sbLeftDelta,
                                       java.lang.StringBuilder sbRightDelta,
                                       java.lang.String szOptions)
                                throws java.sql.SQLException

This method does preparation work before running the collaborative filtering on the graph.

Specified by:

collaborativeFilteringPrep in interface OraclePropertyGraphAnalytics

Parameters:

szEdgeTabName - a String with the name of the graph edges table

sbLeft - a place holder for the name of the working table that holds the left side factor of the matrix factorization

sbRight - a place holder for the name of the working table that holds the right side factor of the matrix factorization

sbLeft1 - a place holder for the name of the working table that holds the left side intermediate step in the gradient descent

sbRight1 - a place holder for the name of the working table that holds the right side intermediate step in the gradient descent

sbMatrixProduct - a place holder for the name of the working table that holds intermediate matrix product

sbLeftDelta - a place holder for the name of the working table that holds intermediate left side delta in gradient descent

sbRightDelta - a place holder for the name of the working table that holds intermediate right side delta in gradient descent

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTriangle

public static long countTriangle(java.sql.Connection conn,
                                 java.lang.String szEdgeTabName,
                                 java.lang.StringBuilder sbUndirWorkTab,
                                 int iNumSubPtns,
                                 int iDOP,
                                 java.lang.String szTBS,
                                 java.lang.String szOptions)
                          throws java.sql.SQLException

Returns the number of triangles that exist in this graph. By default, the underlying logic will NOT create an undirected graph before counting the number of triangles. To change that behavior, one can add ",CREATE_UNDIRECTED=T," to the szOptions.

Parameters:

conn - a JDBC Connection instance

szEdgeTabName - name of the edge table (or view)

sbUndirWorkTab - a place holder for the name of the undirected working table to be used for calculating triangles. This argument is relevant when "CREATE_UNDIRECTED=T" option is set.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. Set the value to 1 to avoid using partitioned approach.

iDOP - degree of parallelism to be used

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTriangle

public long countTriangle(java.lang.StringBuilder sbUndirWorkTab)
                   throws java.sql.SQLException

Returns the number of triangles that exist in this graph which is treated as undirected. Note that an undirected graph will be created before counting the number of triangles.

Specified by:

countTriangle in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirWorkTab - a place holder for the name of the undirected working table to be used for calculating triangles.

Throws:

java.sql.SQLException

countTriangle

public long countTriangle(java.lang.StringBuilder sbUndirWorkTab,
                          int iNumSubPtns,
                          int iDOP,
                          java.lang.String szTBS,
                          java.lang.String szOptions)
                   throws java.sql.SQLException

Returns the number of triangles that exist in this graph. By default, the underlying logic will NOT create an undirected graph before counting the number of triangles. To change that behavior, one can add ",CREATE_UNDIRECTED=T," to the szOptions.

Specified by:

countTriangle in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirWorkTab - a place holder for the name of the undirected working table to be used for calculating triangles. This argument is relevant when "CREATE_UNDIRECTED=T" option is set.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. Set the value to 1 to avoid using partitioned approach.

iDOP - degree of parallelism to be used

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTriangleCleanup

public void countTriangleCleanup(java.lang.StringBuilder sbUndirBMTab,
                                 java.lang.StringBuilder sbMapWorkTab,
                                 java.lang.StringBuilder sbUndirAMTab,
                                 java.lang.String szOptions)
                          throws java.sql.SQLException

This method does cleanup work after running triangle calculations. Basically, all intermediate working tables will be truncated and dropped.

Specified by:

countTriangleCleanup in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTrianglePrep

public void countTrianglePrep(java.lang.StringBuilder sbUndirBMTab,
                              java.lang.StringBuilder sbMapWorkTab,
                              java.lang.StringBuilder sbUndirAMTab,
                              java.lang.String szOptions)
                       throws java.sql.SQLException

This method does preparation work before running triangle calculations.

Specified by:

countTrianglePrep in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles. The map working table is for renumbering vertices.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTriangleRenum

public static long countTriangleRenum(java.sql.Connection conn,
                                      java.lang.String szEdgeTabName,
                                      java.lang.StringBuilder sbUndirBMTab,
                                      java.lang.StringBuilder sbMapWorkTab,
                                      java.lang.StringBuilder sbUndirAMTab,
                                      int iNumSubPtns,
                                      int iDOP,
                                      java.lang.String szTBS,
                                      java.lang.String szOptions)
                               throws java.sql.SQLException

Returns the number of triangles that exist in this graph (treated as undirected). A re-numbering of vertices will be carried out before the actual work of counting is performed.

Parameters:

conn - a JDBC connection to Oracle Database

szEdgeTabName - name of a graph edge table (or view)

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. Set the value to 1 to avoid using partitioned approach.

iDOP - degree of parallelism to be used for creating a working table holding undirected graph data and for executing triangle counting query.

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

countTriangleRenum

public long countTriangleRenum(java.lang.StringBuilder sbUndirBMTab,
                               java.lang.StringBuilder sbMapWorkTab,
                               java.lang.StringBuilder sbUndirAMTab,
                               int iNumSubPtns,
                               int iDOP,
                               java.lang.String szTBS,
                               java.lang.String szOptions)
                        throws java.sql.SQLException

Returns the number of triangles that exist in this graph (treated as undirected). A re-numbering of vertices will be carried out before the actual work of counting is performed.

Specified by:

countTriangleRenum in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. Set the value to 1 to avoid using partitioned approach.

iDOP - degree of parallelism to be used for creating a working table holding undirected graph data and for executing triangle counting query.

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

createStatTable

public void createStatTable(java.lang.String statTableName)
                     throws java.sql.SQLException

Creates a stat table using DBMS_STATS package.

Parameters:

statTableName - name of the stat table.

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

estimateTriangleRenum

public static long estimateTriangleRenum(java.sql.Connection conn,
                                         java.lang.String szEdgeTabName,
                                         java.lang.StringBuilder sbUndirBMTab,
                                         java.lang.StringBuilder sbMapWorkTab,
                                         java.lang.StringBuilder sbUndirAMTab,
                                         int iNumSubPtns,
                                         int iChunk,
                                         int iDOP,
                                         java.lang.String szTBS,
                                         java.lang.String szOptions)
                                  throws java.sql.SQLException

This method estimates the number of triangles that exist in this graph (treated as undirected). A re-numbering of vertices will be carried out before the actual work of counting is performed.

Parameters:

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. As the number of logical partitions grows, the estimation will become faster (and less precise).

iChunk - an integer in the range of [0, iNumSubPtns*iNumSubPtns - 1]

iDOP - degree of parallelism to be used

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

estimateTriangleRenum

public long estimateTriangleRenum(java.lang.StringBuilder sbUndirBMTab,
                                  java.lang.StringBuilder sbMapWorkTab,
                                  java.lang.StringBuilder sbUndirAMTab,
                                  int iNumSubPtns,
                                  int iChunk,
                                  int iDOP,
                                  java.lang.String szTBS,
                                  java.lang.String szOptions)
                           throws java.sql.SQLException

This method estimates the number of triangles that exist in this graph (treated as undirected). A re-numbering of vertices will be carried out before the actual work of counting is performed.

Specified by:

estimateTriangleRenum in interface OraclePropertyGraphAnalytics

Parameters:

sbUndirBMTab - a place holder for the name of the undirected working table (before mapping) to be used for calculating triangles.

sbMapWorkTab - a place holder for the name of the map working table to be used for calculating triangles.

sbUndirAMTab - a place holder for the name of the undirected working table (after mapping) to be used for calculating triangles.

iNumSubPtns - has to be a positive integer and has to be power of two. This value represents the number of logical partitions chosen to do the calculation. As the number of logical partitions grows, the estimation will become faster (and less precise).

iChunk - an integer in the range of [0, iNumSubPtns*iNumSubPtns - 1]

iDOP - degree of parallelism to be used

szTBS - tablespace. It can be NULL.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

expEdgeStat

public void expEdgeStat(java.lang.String statTableName,
                        java.lang.String statid)
                 throws java.sql.SQLException

Exports the edge table statistics into the given stat table.

Parameters:

statTableName - name of the stat table.

statid - identifier of statistics

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

expVertexStat

public void expVertexStat(java.lang.String statTableName,
                          java.lang.String statid)
                   throws java.sql.SQLException

Exports the vertex table statistics into the given stat table.

Parameters:

statTableName - name of the stat table.

statid - identifier of statistics

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

findCluster

public static void findCluster(java.sql.Connection conn,
                               java.lang.String szEdgeTabName,
                               java.lang.StringBuilder sbClusters,
                               java.lang.StringBuilder sbUndir,
                               java.lang.StringBuilder sbCluas,
                               java.lang.StringBuilder sbNewas,
                               java.lang.StringBuilder sbDelta,
                               int iDOP,
                               java.lang.String szTBS,
                               java.lang.String szOptions)
                        throws java.sql.SQLException

This method finds clusters in a given graph (treated as undirected).

Parameters:

conn - a JDBC connection to Oracle Database

szEdgeTabName - name of the edge table (or view). Note that one can create a view on top of existing edge table to filter out unwanted edges.

sbClusters - a place holder for the name of the output table that holds cluster assignment for vertices

sbUndir - a place holder for the name of the working table that holds undirected graph

sbCluas - a place holder for the name of the working table that holds intermediate cluster assignment

sbNewas - a place holder for the name of the working table that holds updated cluster assignment

sbDelta - a place holder for the name of the working table that holds delta information

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

findCluster

public void findCluster(java.lang.StringBuilder sbClusters,
                        java.lang.StringBuilder sbUndir,
                        java.lang.StringBuilder sbCluas,
                        java.lang.StringBuilder sbNewas,
                        java.lang.StringBuilder sbDelta,
                        int iDOP,
                        java.lang.String szTBS,
                        java.lang.String szOptions)
                 throws java.sql.SQLException

This method finds clusters in this property graph (treated as undirected).

Specified by:

findCluster in interface OraclePropertyGraphAnalytics

Parameters:

sbClusters - a place holder for the name of the output table that holds cluster assignment for vertices

sbUndir - a place holder for the name of the working table that holds undirected graph

sbCluas - a place holder for the name of the working table that holds intermediate cluster assignment

sbNewas - a place holder for the name of the working table that holds updated cluster assignment

sbDelta - a place holder for the name of the working table that holds delta information

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

findClusterCleanup

public void findClusterCleanup(java.lang.StringBuilder sbClusters,
                               java.lang.StringBuilder sbUndir,
                               java.lang.StringBuilder sbCluas,
                               java.lang.StringBuilder sbNewas,
                               java.lang.StringBuilder sbDelta,
                               java.lang.String szOptions)
                        throws java.sql.SQLException

This method does cleanup work after running graph cluster detection. All the tables that hold intermediate data and the output cluster assignment will be dropped.

Specified by:

findClusterCleanup in interface OraclePropertyGraphAnalytics

Parameters:

sbClusters - a place holder for the name of the output table that holds cluster assignment for vertices

sbUndir - a place holder for the name of the working table that holds undirected graph

sbCluas - a place holder for the name of the working table that holds intermediate cluster assignment

sbNewas - a place holder for the name of the working table that holds updated cluster assignment

sbDelta - a place holder for the name of the working table that holds delta information

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

findClusterPrep

public void findClusterPrep(java.lang.StringBuilder sbClusters,
                            java.lang.StringBuilder sbUndir,
                            java.lang.StringBuilder sbCluas,
                            java.lang.StringBuilder sbNewas,
                            java.lang.StringBuilder sbDelta,
                            java.lang.String szOptions)
                     throws java.sql.SQLException

This method does preparation work before running graph cluster detection.

Specified by:

findClusterPrep in interface OraclePropertyGraphAnalytics

Parameters:

sbClusters - a place holder for the name of the output table that holds cluster assignment for vertices

sbUndir - a place holder for the name of the working table that holds undirected graph

sbCluas - a place holder for the name of the working table that holds intermediate cluster assignment

sbNewas - a place holder for the name of the working table that holds updated cluster assignment

sbDelta - a place holder for the name of the working table that holds delta information

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

getEdgeTabName

public abstract java.lang.String getEdgeTabName()

getOracle

public abstract Oracle getOracle()

getTopK

public static java.sql.ResultSet getTopK(java.sql.Connection conn,
                                         java.lang.StringBuilder sbTabName,
                                         int iColIdx,
                                         SQLAnalyticsGraph.Sort sort,
                                         int k,
                                         int iDOP)
                                  throws java.sql.SQLException

Returns the top K rows of a given table. Note that the result will include an additional column "ROW_COUNT" that can be used for index purpose. This column can be safely ignored. As a contract, the caller of this API needs to close the result set and its associated JDBC statement after consumption.

Parameters:

conn - a JDBC connection to Oracle Database

sbTabName - a place holder for the table name

iColIdx - the column to be sorted

sort - ASC or DESC

k - k rows to be returned. Note that it is an upper bound

iDOP - degree of parallelism to be used

Returns:

a JDBC result set.

Throws:

java.sql.SQLException

getTopK

public java.sql.ResultSet getTopK(java.lang.StringBuilder sbTabName,
                                  int iColIdx,
                                  SQLAnalyticsGraph.Sort sort,
                                  int k)
                           throws java.sql.SQLException

Returns the top K rows of a given table. Note that the result will include an additional column "ROW_COUNT" that can be used for index purpose. This column can be safely ignored.

Parameters:

sbTabName - a place holder for the table name

iColIdx - the column to be sorted

sort - ASC or DESC

k - k rows to be returned. Note that it is an upper bound

Returns:

a JDBC result set. Note that the query will be executed with a DOP of getSessionQueryDOP.

Throws:

java.sql.SQLException

getTopK

public java.sql.ResultSet getTopK(java.lang.StringBuilder sbTabName,
                                  int iColIdx,
                                  SQLAnalyticsGraph.Sort sort,
                                  int k,
                                  int iDOP)
                           throws java.sql.SQLException

Returns the top K rows of a given table. Note that the result will include an additional column "ROW_COUNT" that can be used for index purpose. This column can be safely ignored.

Parameters:

sbTabName - a place holder for the table name

iColIdx - the column to be sorted

sort - ASC or DESC

k - k rows to be returned. Note that it is an upper bound

iDOP - degree of parallelism to be used

Returns:

a JDBC result set.

Throws:

java.sql.SQLException

impEdgeStat

public void impEdgeStat(java.lang.String statTableName,
                        java.lang.String statid)
                 throws java.sql.SQLException

Imports the edge table statistics from the given stat table.

Parameters:

statTableName - name of the stat table.

statid - identifier of statistics

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

impVertexStat

public void impVertexStat(java.lang.String statTableName,
                          java.lang.String statid)
                   throws java.sql.SQLException

Imports the vertex table statistics from the given stat table.

Parameters:

statTableName - name of the stat table.

statid - identifier of statistics

Throws:

java.sql.SQLException - when there is error generated from DBMS_STATS package.

pageRank

public static long pageRank(java.sql.Connection conn,
                            java.lang.String szEdgeTabName,
                            java.lang.StringBuilder sbNodePR,
                            java.lang.StringBuilder sbNodeNextPR,
                            java.lang.StringBuilder sbEdgeTabDeg,
                            java.lang.StringBuilder sbDelta,
                            float d,
                            int iNumIterations,
                            float convergence,
                            int iDOP,
                            java.lang.String szTBS,
                            java.lang.String szOptions)
                     throws java.sql.SQLException

This method calculates the page rank values of vertices in the given graph.

Parameters:

conn - a JDBC connection to Oracle Database

szEdgeTabName - name of a graph edge table (or view)

sbNodePR - a place holder for the name of the working table that holds the page rank values of the vertices

sbNodeNextPR - a place holder for the name of the working table that holds the page rank values in the next iteration

sbEdgeTabDeg - a place holder for the name of the working table that holds edges and node degree information

sbDelta - a place holder for the name of the working table that holds information about some special vertices

d - damping factor, 0.85f is often used.

iNumIterations - number of iterations to perform PR calculations.

convergence - if positive, then continue PR calculations till sum of PR value changes is less than this convergence setting.

szOptions - options to be used by the underlying PL/SQL API.

iDOP - degree of parallelism to be used for creating a working table holding undirected graph data and for executing triangle counting query.

szTBS - tablespace. It can be NULL.

Returns:

the number of vertices of the table stored in sbNodePR. The PR values of vertices are stored in the working table sbNodePR

Throws:

java.sql.SQLException

pageRank

public long pageRank(java.lang.StringBuilder sbNodePR,
                     java.lang.StringBuilder sbNodeNextPR,
                     java.lang.StringBuilder sbEdgeTabDeg,
                     java.lang.StringBuilder sbDelta,
                     float d,
                     int iNumIterations,
                     float convergence,
                     int iDOP,
                     java.lang.String szTBS,
                     java.lang.String szOptions)
              throws java.sql.SQLException

This method calculates the page rank values of vertices in this graph.

Specified by:

pageRank in interface OraclePropertyGraphAnalytics

Parameters:

sbNodePR - a place holder for the name of the working table that holds the page rank values of the vertices

sbNodeNextPR - a place holder for the name of the working table that holds the page rank values in the next iteration

sbEdgeTabDeg - a place holder for the name of the working table that holds edges and node degree information

sbDelta - a place holder for the name of the working table that holds information about some special vertices

d - damping factor, 0.85f is often used.

iNumIterations - number of iterations to perform PR calculations.

convergence - if positive, then continue PR calculations till sum of PR value changes is less than this convergence setting.

szOptions - options to be used by the underlying PL/SQL API.

iDOP - degree of parallelism to be used for creating a working table holding undirected graph data and for executing triangle counting query.

szTBS - tablespace. It can be NULL.

Returns:

the number of vertices of the table stored in sbNodePR. The PR values of vertices are stored in the working table sbNodePR

Throws:

java.sql.SQLException

pageRankCleanup

public void pageRankCleanup(java.lang.StringBuilder sbNodePR,
                            java.lang.StringBuilder sbNodeNextPR,
                            java.lang.StringBuilder sbEdgeTabDeg,
                            java.lang.StringBuilder sbDelta,
                            java.lang.String szOptions)
                     throws java.sql.SQLException

This method does cleanup work before running page rank calculation.

Specified by:

pageRankCleanup in interface OraclePropertyGraphAnalytics

Parameters:

sbNodePR - a place holder for the name of the working table that holds the page rank values of the vertices

sbNodeNextPR - a place holder for the name of the working table that holds the page rank values in the next iteration

sbEdgeTabDeg - a place holder for the name of the working table that holds edges and node degree information

sbDelta - a place holder for the name of the working table that holds information about some special vertices

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

pageRankPrep

public void pageRankPrep(java.lang.StringBuilder sbNodePR,
                         java.lang.StringBuilder sbNodeNextPR,
                         java.lang.StringBuilder sbEdgeTabDeg,
                         java.lang.StringBuilder sbDelta,
                         java.lang.String szOptions)
                  throws java.sql.SQLException

This method does preparation work before running page rank calculation.

Specified by:

pageRankPrep in interface OraclePropertyGraphAnalytics

Parameters:

sbNodePR - a place holder for the name of the working table that holds the page rank values of the vertices

sbNodeNextPR - a place holder for the name of the working table that holds the page rank values in the next iteration

sbEdgeTabDeg - a place holder for the name of the working table that holds edges and node degree information

sbDelta - a place holder for the name of the working table that holds information about some special vertices

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

shortestPath

public static java.lang.String shortestPath(java.sql.Connection conn,
                                            java.lang.String szEdgeTabName,
                                            java.lang.Long lStart,
                                            java.lang.Long lEnd,
                                            java.lang.StringBuilder sbExpTab,
                                            java.lang.StringBuilder sbPathWeights,
                                            java.lang.String szOptions,
                                            java.lang.Long lSCN)
                                     throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Parameters:

conn - a JDBC connection to Oracle Database

szEdgeTabName - name of the edge table (or view). Note that one can create a view on top of existing edge table to filter out unwanted edges.

lStart - start vertex integer ID

lEnd - destination vertex integer ID

sbPathWeights - place holder of path weights of the shortest path. The first number represents the number of vertices on the shortest path, the second number represents the total weight and the rest numbers denote weights of edges along the shortest path. All these numbers are delimited by spaces.

szOptions - options to be passed on to the underlying PL/SQL API.

lSCN - SCN for the edge table. It can be NULL.

Throws:

java.sql.SQLException

shortestPath

public java.lang.String shortestPath(java.lang.Long lStart,
                                     java.lang.Long lEnd,
                                     java.lang.StringBuilder sbExpTab)
                              throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Specified by:

shortestPath in interface OraclePropertyGraphAnalytics

Parameters:

lStart - start vertex integer ID

lEnd - destination vertex integer ID

sbExpTab - place holder of expansion table name

Throws:

java.sql.SQLException

shortestPath

public java.lang.String shortestPath(java.lang.Long lStart,
                                     java.lang.Long lEnd,
                                     java.lang.StringBuilder sbExpTab,
                                     java.lang.StringBuilder sbPathWeights,
                                     java.lang.String szOptions)
                              throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Specified by:

shortestPath in interface OraclePropertyGraphAnalytics

Parameters:

lStart - start vertex integer ID

lEnd - destination vertex integer ID

sbExpTab - place holder of expansion table name

sbPathWeights - place holder of path weights of the shortest path. The first number represents the number of vertices on the shortest path, the second number represents the total weight and the rest numbers denote weights of edges along the shortest path. All these numbers are delimited by spaces.

szOptions - options to be passed on to the underlying PL/SQL API.

Throws:

java.sql.SQLException

shortestPath

public java.lang.String shortestPath(java.lang.Long lStart,
                                     java.lang.Long lEnd,
                                     java.lang.StringBuilder sbExpTab,
                                     java.lang.String szEdgeTabName,
                                     java.lang.StringBuilder sbPathWeights,
                                     java.lang.String szOptions,
                                     java.lang.Long lSCN)
                              throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Specified by:

shortestPath in interface OraclePropertyGraphAnalytics

Parameters:

lStart - start vertex integer ID

lEnd - destination vertex integer ID

szEdgeTabName - name of the edge table (or view). Note that one can create a view on top of existing edge table to filter out unwanted edges.

sbPathWeights - place holder of path weights of the shortest path. The first number represents the number of vertices on the shortest path, the second number represents the total weight and the rest numbers denote weights of edges along the shortest path. All these numbers are delimited by spaces.

szOptions - options to be passed on to the underlying PL/SQL API.

lSCN - SCN for the edge table. It can be NULL.

Throws:

java.sql.SQLException

shortestPath

public java.lang.String shortestPath(OracleVertex vStart,
                                     OracleVertex vEnd,
                                     java.lang.StringBuilder sbExpTab)
                              throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Specified by:

shortestPath in interface OraclePropertyGraphAnalytics

Parameters:

vStart - start vertex object

vEnd - destination vertex object

sbExpTab - place holder of expansion table name

Throws:

java.sql.SQLException

shortestPath

public java.lang.String shortestPath(OracleVertex vStart,
                                     OracleVertex vEnd,
                                     java.lang.StringBuilder sbExpTab,
                                     java.lang.StringBuilder sbPathWeights,
                                     java.lang.String szOptions)
                              throws java.sql.SQLException

Returns a shortest path in a string format. The path is delimited by space. A NULL value will be returned if there is no path found.

Specified by:

shortestPath in interface OraclePropertyGraphAnalytics

Parameters:

vStart - start vertex object

vEnd - destination vertex object

sbExpTab - place holder of expansion table name

sbPathWeights - place holder of path weights of the shortest path. The first number represents the number of vertices on the shortest path, the second number represents the total weight and the rest numbers denote weights of edges along the shortest path. All these numbers are delimited by spaces.

szOptions - options to be passed on to the underlying PL/SQL API.

Throws:

java.sql.SQLException

shortestPathCleanup

public void shortestPathCleanup(java.lang.StringBuilder sbExpTab,
                                java.lang.String szOptions)
                         throws java.sql.SQLException

This method does clean up after running shortest path calculations.

Specified by:

shortestPathCleanup in interface OraclePropertyGraphAnalytics

Parameters:

sbExpTab - a place holder for the name of the expansion table to be used for Bi-directional, Set Style, Dijkstra search.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

shortestPathPrep

public void shortestPathPrep(java.lang.StringBuilder sbExpTab,
                             java.lang.String szOptions)
                      throws java.sql.SQLException

This method does some preparation work before running one or multiple shortest path calculations. This method only needs to be invoked once unless a different expansion table is to be used to calculate shortest paths.

Specified by:

shortestPathPrep in interface OraclePropertyGraphAnalytics

Parameters:

sbExpTab - a place holder for the name of the expansion table to be used for Bi-directional, Set Style, Dijkstra search. If sbExpTab is empty, an intermediate working table will be created and the table name will be returned in the sbExpTab object.

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

sparsifyGraph

public static void sparsifyGraph(java.sql.Connection conn,
                                 java.lang.String szEdgeTabName,
                                 float fThreshold,
                                 int iMinKeep,
                                 int iDOP,
                                 java.lang.StringBuilder sbOutTab,
                                 java.lang.StringBuilder sbUndTab,
                                 java.lang.StringBuilder sbHshTab,
                                 java.lang.StringBuilder sbMchTab,
                                 java.lang.String szTBS,
                                 java.lang.String szOptions)
                          throws java.sql.SQLException

This method sparsifies a given graph (treated as undirected). The output edges will be stored in the table denoted by sbOutTab.

Parameters:

sbOutTab - a place holder for the name of the output table that holds the sparsified graph

sbUndTab - a place holder for the name of the working table that holds undirected graph

sbHshTab - a place holder for the name of the working table that holds min hash values

sbMchTab - a place holder for the name of the working table that holds match counts

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

sparsifyGraph

public void sparsifyGraph(float fThreshold,
                          int iMinKeep,
                          int iDOP,
                          java.lang.StringBuilder sbOutTab,
                          java.lang.StringBuilder sbUndTab,
                          java.lang.StringBuilder sbHshTab,
                          java.lang.StringBuilder sbMchTab,
                          java.lang.String szTBS,
                          java.lang.String szOptions)
                   throws java.sql.SQLException

This method sparsifies this graph instance (treated as undirected). The output edges will be stored in the table denoted by sbOutTab.

Parameters:

sbOutTab - a place holder for the name of the output table that holds the sparsified graph

sbUndTab - a place holder for the name of the working table that holds undirected graph

sbHshTab - a place holder for the name of the working table that holds min hash values

sbMchTab - a place holder for the name of the working table that holds match counts

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

sparsifyGraphCleanup

public void sparsifyGraphCleanup(java.lang.StringBuilder sbOutTab,
                                 java.lang.StringBuilder sbUndTab,
                                 java.lang.StringBuilder sbHshTab,
                                 java.lang.StringBuilder sbMchTab,
                                 java.lang.String szOptions)
                          throws java.sql.SQLException

This method does cleanup work before running graph sparsification.

Parameters:

sbOutTab - a place holder for the name of the output table that holds the sparsified graph

sbUndTab - a place holder for the name of the working table that holds undirected graph

sbHshTab - a place holder for the name of the working table that holds min hash values

sbMchTab - a place holder for the name of the working table that holds match counts

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException

sparsifyGraphPrep

public void sparsifyGraphPrep(java.lang.StringBuilder sbOutTab,
                              java.lang.StringBuilder sbUndTab,
                              java.lang.StringBuilder sbHshTab,
                              java.lang.StringBuilder sbMchTab,
                              java.lang.String szOptions)
                       throws java.sql.SQLException

This method does preparation work before running graph sparsification.

Parameters:

sbOutTab - a place holder for the name of the output table that holds the sparsified graph

sbUndTab - a place holder for the name of the working table that holds undirected graph

sbHshTab - a place holder for the name of the working table that holds min hash values

sbMchTab - a place holder for the name of the working table that holds match counts

szOptions - options to be used by the underlying PL/SQL API.

Throws:

java.sql.SQLException