Graph Class Reference

The Graph class This is the main class for a Graph, used in conjuction with GraphVertex, Parser and Matrix objects. Graph class methods are the interface to various analysis algorithms GraphVertex class holds each vertex data (colors, strings, statistics, etc) Matrix class holds the adjacency matrix of the network. Parser class loads files of networks. More...

#include <graph.h>

Inheritance diagram for Graph:

Collaboration diagram for Graph:

Classes
struct	GraphVisibilitySnapshot
struct	NodeConnectivityResult

Public Types
enum	ModStatus {   NewNet = -1 , SavedUnchanged = 0 , MinorOptions = 1 , VertexMetadata = 2 ,   EdgeMetadata = 3 , VertexPositions = 4 , MajorChanges = 10 , VertexCount = 11 ,   EdgeCount = 12 , VertexEdgeCount = 13 }
enum	Clustering { Single_Linkage = 0 , Complete_Linkage = 1 , Average_Linkage = 2 }
enum class	NodeConnectivityStatus { Ok , Adjacent , Invalid }

Public Slots
int	relationCurrent ()
	Gets the current relation number.
QString	relationCurrentName () const
	Gets the current relation name.
void	relationCurrentRename (const QString &newName)
	Overload. Renames current relation to newName, without emitting any signal to MW.
void	relationCurrentRename (const QString &newName, const bool &signalMW)
	Renames current relation to newName, optionally emitting a signal to MW.
void	vertexCreate (const int &number, const int &size, const QString &color, const QString &numColor, const int &numSize, const QString &label, const QString &labelColor, const int &labelSize, const QPointF &p, const QString &shape, const QString &iconPath=QString(), const bool &signalMW=false, const QHash< QString, QString > &customAttributes=QHash< QString, QString >())
	Creates a new vertex.
void	graphFileLoaded (const int &fileType, const QString &fName=QString(), const QString &netName=QString(), const int &totalNodes=0, const int &totalLinks=0, const int &edgeDirType=0, const qint64 &elapsedTime=0, const QString &message=QString())
	Stores loaded file name, graph name, sets edge direction type and signals MW to update the UI.
void	vertexRemoveDummyNode (int)
	Deletes a dummy node.
void	graphLoadedTerminateParserThreads (QString reason)
	Graph::graphLoadedTerminateParserThreads.
void	setSelectionChanged (const QList< int > selectedVertices, const QList< SelectedEdge > selectedEdges)
	Sets the user-selected vertices and edges.
void	graphClickedEmptySpace (const QPointF &p)
	Resets the clicked edge and node.
bool	edgeCreate (const int &v1, const int &v2, const qreal &weight, const QString &color, const int &type=0, const bool &drawArrows=true, const bool &bezier=false, const QString &label=QString(), const bool &signalMW=true, const QHash< QString, QString > &edgeCustomAttributes=QHash< QString, QString >())
	Checks if edge (v1,v2) already exists, then creates it and signals the UI to draw it.
void	edgeCreateWebCrawler (const int &source, const int &target)
	Called from WebCrawler when it finds an new link Calls edgeCreate() method with initEdgeColor.
void	vertexCreateAtPos (const QPointF &p)
	Creates a new vertex in the given position.
void	vertexCreateAtPosRandom (const bool &signalMW=false)
	Creates a new randomly positioned vertex with default values.
void	vertexCreateAtPosRandomWithLabel (const int &i, const QString &label, const bool &signalMW=false)
	Creates a new randomly positioned vertex with specific number and label. All other values are from the defaults.
void	relationSet (int relNum=RAND_MAX, const bool &updateUI=true)
	Changes the current relation, and optionally emits signals to MW/GW (default: true).
void	relationNext ()
	Changes graph to next relation.
void	relationPrev ()
	Changes graph to previous relation.
void	canvasSizeSet (const int &width, const int &height)
	Sets the size of the canvas.
void	canvasSizeSetQuiet (const qreal &width, const qreal &height)
	Sets the canvas size without rescaling node positions.
double	canvasMaxRadius () const
	Gets the max radius of the canvas.
qreal	canvasMinDimension () const
	Gets the min dimensions of the canvas.
double	canvasVisibleX (const double &x) const
	Checks if x is visible inside the canvas usable area and if not returns an adjusted x-coordinate.
double	canvasVisibleY (const double &y) const
	Checks if y is visible inside the canvas usable area and if not returns an adjusted y-coordinate.
double	canvasRandomX () const
	Returns a random x-coordinate adjusted to be visible inside the canvas usable area.
double	canvasRandomY () const
	Returns a random y-coordinate adjusted to be visible inside the canvas usable area.
void	vertexIsolatedAllToggle (const bool &toggle)
	Toggles the status of all isolated vertices (those without links).
void	vertexClickedSet (const int &v, const QPointF &p)
	Sets the clicked vertex.
void	edgeClickedSet (const int &v1, const int &v2, const bool &openMenu=false)
	Sets the clicked edge.
void	vertexFilterByCentrality (const float threshold, const bool overThreshold, const IndexType centralityIndex)
	Filters vertices by their score on the given centrality or prestige index.
void	vertexFilterByEgoNetwork (const int v1, const int depth=1)
	Saves current visibility state and shows only the ego network of vertex v1 at the given depth.
void	vertexFilterBySelection (const QList< int > &selectedVertices)
	Saves current visibility state and shows only the selected vertices and the edges between them.
void	vertexFilterByAttribute (const FilterCondition &cond)
	Shows only vertices whose custom attribute satisfies cond; all other vertices are hidden.
void	edgeFilterByAttribute (const FilterCondition &cond)
	Hides all edges whose custom attribute does not satisfy cond.
void	vertexFilterRestoreAll ()
	Restores vertex and edge visibility from the top snapshot on the history stack.
void	vertexFilterRemoveAt (int stackIndex)
	Removes the filter at stackIndex (0 = oldest) and replays the rest.
QList< FilterSpec >	filterSpecList () const
	Returns the FilterSpec list in application order (oldest first).
bool	visibilityHistoryEmpty () const
	Returns true if the visibility history stack is empty.
void	vertexFilterByQuery (const GraphQuery &query)
	Shows only vertices whose custom attributes satisfy ALL conditions in query.
void	edgeFilterByQuery (const GraphQuery &query)
	Shows only edges whose custom attributes satisfy ALL conditions in query.
void	edgeFilterByWeight (const qreal, const bool)
	Hides edges whose weight does not satisfy the threshold condition.
void	edgeFilterReset ()
	Re-enables all edges in the current relation.
void	edgeFilterUnilateral (const bool &toggle)
	Enables or disables unilateral edges in current relationship.
Graph *	subgraphExtract (const QString &name, const bool &includeCustomAttributes=true)
	Extracts currently visible (non-filtered) nodes and their inter-edges into an independent Graph copy.
Graph *	subgraphExtractFromSelection (const QString &name, const bool &includeCustomAttributes=true)
	Extracts currently selected nodes and their inter-edges into an independent Graph copy.
void	startWebCrawler (const QUrl &startUrl, const QStringList &urlPatternsIncluded, const QStringList &urlPatternsExcluded, const QStringList &linkClasses, const int &maxNodes, const int &maxLinksPerPage, const bool &intLinks, const bool &childLinks, const bool &parentLinks, const bool &selfLinks, const bool &extLinksIncluded, const bool &extLinksCrawl, const bool &socialLinks, const bool &delayedRequests)
	Creates a new WebCrawler, that will parse the downloaded HTML code of each webpage we download. Moves the WebCrawler to a new thread and starts the thread. Then creates the fist node (initial url), and starts the web spider to download the first HTML page. Called by MW with user options.
void	slotHandleCrawlerRequestReply ()
	Gets the reply of a MW network request made by Web Crawler, and emits that reply as is to the Web Crawler.
void	webSpider ()
	A loop, that takes urls awaiting in front of the urlQueue, and signals to the MW to make the network request.
void	slotCancelComputation ()
	Slot called by MainWindow when the user clicks Cancel in the progress dialog.
QString	htmlEscaped (QString str) const
	Helper method, returns a nice qstring where all html special chars are encoded.

Signals
void	signalWebCrawlParse (QNetworkReply *reply)
void	signalNetworkManagerRequest (const QUrl &currentUrl, const NetworkRequestType &type)
void	signalGraphSavedStatus (const int &status)
void	signalGraphModified (const bool &undirected, const int &vertices, const int &edges, const qreal &density, const bool &notSaved=true)
void	signalGraphLoaded (const int &fileType, const QString &fileName=QString(), const QString &netName=QString(), const int &totalNodes=0, const int &totalLinks=0, const qreal &density=0, const qint64 &elapsedTime=0, const QString &message=QString())
void	statusMessage (const QString &message)
void	signalDatasetDescription (QString)
void	signalNodeClickedInfo (const int &number=0, const QPointF &p=QPointF(), const QString &label=QString(), const int &inDegree=0, const int &outDegree=0)
void	signalEdgeClicked (const MyEdge &edge=MyEdge(), const bool &openMenu=false)
void	signalRelationAddToMW (const QString &newRelation)
void	signalRelationsClear ()
void	signalRelationRenamedToMW (const QString &newRelName)
void	signalRelationChangedToGW (int)
void	signalRelationChangedToMW (const int &relIndex=RAND_MAX)
void	signalGraphDirectedChanged (const bool &directed)
void	signalSelectionChanged (const int &selectedVertices, const int &selectedEdges)
void	signalPromininenceDistributionChartUpdate (QAbstractSeries *series, QAbstractAxis *axisX=Q_NULLPTR, const qreal &min=0, const qreal &max=0, QAbstractAxis *axisY=Q_NULLPTR, const qreal &minF=0, const qreal &maxF=0)
void	signalDrawNode (const QPointF &p, const int &num, const int &size, const QString &nodeShape, const QString &nodeIconPath, const QString &nodeColor, const QString &numberColor, const int &numSize, const int &numDistance, const QString &label, const QString &labelColor, const int &labelSize, const int &labelDistance)
void	signalRemoveNode (int)
void	signalDrawEdge (const int &v1, const int &v2, const qreal &weight, const QString &label="", const QString &color="black", const int &type=0, const bool &drawArrows=true, const bool &bezier=false, const bool &weightNumbers=false)
void	signalRemoveEdge (const int &v1, const int &v2, const bool &removeReverse)
void	signalSetEdgeVisibility (const int &relation, const int &source, const int &target, const bool &toggle, const bool &preserveReverseEdge=false, const int &edgeWeight=1, const int &reverseEdgeWeight=1)
void	signalSetEdgesVisibilityBatch (const QList< EdgeVisibilityChange > &changes)
void	setVertexVisibility (const int &number, const bool &toggle)
void	setNodePos (const int &, const qreal &, const qreal &)
void	signalLayoutFinished ()
void	signalNodesFound (const QList< int > foundList)
void	setNodeSize (const int &v, const int &size)
void	setNodeShape (const int &v, const QString &shape, const QString &iconPath=QString())
void	setNodeColor (const int &v, const QString &color)
void	setNodeLabel (const int &v, const QString &label)
void	setNodeNumberColor (const int &v, const QString &color)
void	setNodeNumberSize (const int &v, const int &size)
void	setNodeNumberDistance (const int &v, const int &distance)
void	setNodeLabelSize (const int &v, const int &size)
void	setNodeLabelColor (const int &v, const QString &color)
void	setNodeLabelDistance (const int &v, const int &distance)
void	setEdgeWeight (const int &v1, const int &v2, const qreal &weight)
void	signalEdgeType (const int &v1, const int &v2, const int &type)
void	setEdgeColor (const int &v1, const int &v2, const QString &color)
void	setEdgeLabel (const int &v1, const int &v2, const QString &label)
void	addGuideCircle (const double &, const double &, const double &)
void	addGuideHLine (const double &y0)

Public Member Functions
	Graph (const int &reserveVerticesSize=5000, const int &reserveEdgesPerVertexSize=500)
	Constructs a Graph.
	~Graph ()
	Destroys the Graph object.
QThread *	getThread () const
void	moveToThreadFacade (QThread *thread)
void	clear (const QString &reason="")
	Clears all vertices.
QString	getFileName () const
	Returns the file name of the current graph, if any.
void	setFileName (const QString &fileName)
	Sets the file name of the current graph.
QString	getName () const
	Returns the name of the current graph.
void	setName (const QString &graphName)
	Sets the name of the current graph.
bool	isSaved () const
	Returns true if the graph is saved.
bool	isLoaded () const
	Returns true if a graph has been loaded from a file.
int	getFileFormat () const
	Returns the format of the last file opened.
void	setFileFormat (const int &fileFormat)
bool	isFileFormatExportSupported (const int &fileFormat) const
	Returns true if the fileFormat is supported for saving.
void	setModStatus (const int &graphNewStatus, const bool &signalMW=true)
	Sets the graph modification status.
bool	isModified () const
	Returns true of graph is modified (edges/vertices added/removed).
void	loadFile (const QString fileName, const QString codecName, const int format, const QString delimiter=QString(), const int sm_two_mode=1, const bool sm_has_labels=false)
	Loads a graph from a given file.
void	saveToFile (const QString &fileName, const int &fileType, const bool &saveEdgeWeights=true, const bool &saveZeroWeightEdges=false)
	Saves the current graph to a file.
bool	saveToPajekFormat (const QString &fileName, QString networkName="", int maxWidth=0, int maxHeight=0)
	Save the current graph to a Pajek (.paj) file.
bool	saveToAdjacencyFormat (const QString &fileName, const bool &saveEdgeWeights=true)
	Saves the active graph to an adjacency-formatted file.
bool	saveToGraphMLFormat (const QString &fileName, const bool &saveZeroWeightEdges=false, QString networkName="", int maxWidth=0, int maxHeight=0)
	Saves the current graph to a GraphML-formatted file.
bool	saveToDotFormat (const QString &fileName)
	Saves the active graph (current relation) to a GraphViz DOT file.
bool	saveToUCINETDLFormat (const QString &fileName)
	Saves the current graph to a UCINET DL file (FULLMATRIX format).
bool	saveToEdgeListWeightedFormat (const QString &fileName)
	Saves the active relation to a weighted edge list file.
bool	saveToEdgeListSimpleFormat (const QString &fileName)
	Saves the active relation to a simple (unweighted) edge list file.
QString	graphMatrixTypeToString (const int &matrixType) const
	Helper method, return the human readable name of matrix type.
int	graphMatrixStrToType (const QString &matrix) const
	Helper method, return the matrix type of human readable matrix name .
QString	graphMetricTypeToString (const int &metricType) const
	Helper method, return the human readable name of metric type.
int	graphMetricStrToType (const QString &metricStr) const
	Helper method, return the identifier of a metric.
QString	graphClusteringMethodTypeToString (const int &methodType) const
	Helper method, return the human readable name of clustering method type.
int	graphClusteringMethodStrToType (const QString &method) const
	Helper method, return clustering method type from the human readable name of it.
int	relations ()
	Returns the count of relationships in this Graph.
void	relationsClear ()
	Clears relationships in this Graph.
void	relationAdd (const QString &relName, const bool &changeRelation=false)
	Adds a new relation to the graph.
int	vertexIndexByNumber (int v) const
	Returns the index of a vertex by its number.
GraphVertex *	vertexAtIndex (int idx)
	Returns the vertex at a given index.
const GraphVertex *	vertexAtIndex (int idx) const
	Returns the vertex at a given index.
VList::const_iterator	verticesBegin () const
	iterator helpers
VList::const_iterator	verticesEnd () const
int	vertexNumberMax ()
	Returns the number of the last vertex in the graph.
int	vertexNumberMin ()
	Returns the number of the first vertex in the graph.
int	vertexDegreeOut (int)
	Returns the outDegree (sum of outbound edge weights) of vertex v1.
int	vertexDegreeIn (int)
	Returns the inDegree (sum of inbound edge weights) of vertex v1.
QList< int >	vertexReciprocalNeighborsList (const int &v1)
	Returns a list of all vertices reciprocally connected to vertex v1 in the current relation.
QSet< int >	vertexReciprocalNeighborsSet (const int &v1)
	Returns the set of all vertices reciprocally connected to vertex v1 in the current relation.
QSet< int >	vertexOutNeighborsSet (const int &v1, const bool includeInEdges=false)
	Returns the set of all 1-hop neighbors of vertex v1 in the current relation.
bool	vertexIsolated (const int &v1) const
	Checks if vertex is isolated.
bool	vertexExists (const int &v1)
	Checks if the given vertex exists in the graph.
int	vertexIndexIfExists (const int &v1)
	Returns the internal position of the given vertex in m_graph, if it exists.
bool	vertexExists (const QString &label)
	Checks if there is a vertex with a specific label in the graph.
int	vertexIndexIfExists (const QString &label)
	Returns the internal position of the vertex with the given label, if it exists.
bool	vertexFindByNumber (const QStringList &numList)
	Finds vertices in strList by their number.
bool	vertexFindByLabel (const QStringList &labelList)
	Finds vertices by their label.
bool	vertexFindByIndexScore (const int &index, const QStringList &thresholds, const bool &considerWeights, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Finds vertices by their index score.
void	vertexRemove (const int &v1)
	Removes the vertex v1 from the graph First, it removes all edges to doomed from other vertices Then it changes the vpos of all subsequent vertices inside m_graph Finally, it removes the vertex.
void	vertexSizeInit (const int)
	Sets the initial vertex size.
void	vertexSizeSet (const int &v, const int &newsize)
	Changes the size of a vertex v or all vertices if v=0.
int	vertexSize (const int &v) const
	Returns the size of vertex v.
void	vertexShapeSetDefault (const QString, const QString &iconPath=QString())
	Sets the default vertex shape and iconPath.
void	vertexShapeSet (const int &v, const QString &shape, const QString &iconPath=QString())
	Changes the shape and iconPath of vertex v1, or all vertices if v1=-1.
QString	vertexShape (const int &v)
	Returns the shape of this vertex.
QString	vertexShapeIconPath (const int &v)
	Returns the IconPath of vertex v1.
bool	graphHasVertexCustomIcons () const
	Returns true if at least one vertex has a 'custom' shape (therefore a custom icon).
QStringList	graphHasVertexCustomAttributes () const
	Returns true if at least one vertex has a 'custom' attribute.
void	vertexColorInit (const QString &color)
	Graph::vertexColorInit default vertex color initialization.
void	vertexColorSet (const int &v, const QString &color)
	Changes the color of vertex v1.
QColor	vertexColor (const int &v) const
	Graph::vertexColor.
void	vertexNumberColorInit (const QString &color)
	Changes the initial color of the vertex numbers.
void	vertexNumberColorSet (const int &v=0, const QString &color="#000000")
	Graph::vertexColorSet Changes the color of vertex v1.
void	vertexNumberSizeInit (const int &size)
	Changes the initial size of vertex numbers.
void	vertexNumberSizeSet (const int &v, const int &newsize)
	Changes the size of vertex v number.
void	vertexNumberDistanceInit (const int &distance)
	Changes the initial distance of vertex numbers.
void	vertexNumberDistanceSet (const int &v, const int &newDistance)
	Changes the distance.of vertex v number from the vertex.
void	vertexLabelSet (const int &v, const QString &label)
	Changes the label of a vertex v1.
QString	vertexLabel (const int &v) const
	Returns the label of a vertex v1.
void	vertexLabelsVisibilitySet (bool toggle)
void	vertexLabelSizeInit (int newSize)
	Graph::vertexLabelSizeInit Changes the default size of vertex labels.
void	vertexLabelSizeSet (const int &v, const int &labelSize)
	Changes the label size of vertex v1 or all vertices if v1=0.
void	vertexLabelColorInit (QString color)
	Graph::vertexLabelColorInit Changes the default vertex label color.
void	vertexLabelColorSet (const int &v1, const QString &color)
	Changes the label color of vertex v1 or all vertices if v1 = 0.
void	vertexLabelDistanceInit (const int &distance)
	Changes the default distance of vertex labels.
void	vertexLabelDistanceSet (const int &v, const int &newDistance)
	Changes the distance.of vertex v label from the vertex.
void	vertexLabelDistanceAllSet (const int &newDistance)
	Changes the distance of all vertex labels from their vertices.
void	vertexCustomAttributesSet (const int &v1, const QHash< QString, QString > &customAttributes)
	Sets custom attributes for a specified vertex.
void	vertexCustomAttributeSet (const int &v1, const QString &key, const QString &value)
	Sets a single custom attribute key/value on vertex v1, leaving all other attributes untouched.
void	vertexCustomAttributeRemove (const int &v1, const QString &key)
	Removes a single custom attribute key from vertex v1. Does nothing if the key does not exist.
QHash< QString, QString >	vertexCustomAttributes (const int &v1) const
	Retrieves the vertex at the position specified by the index v1 from the vpos map and calls its customAttributes method.
int	vertexAttributesImport (const QStringList &headers, const QVector< QStringList > &rows, int idColumn, bool matchByLabel)
	Imports custom attributes from a parsed table into existing vertices.
void	vertexPosSet (const int &v, const int &x, const int &y)
	Changes the position of the given vertex.
QPointF	vertexPos (const int &v1) const
	Returns the position of the given vertex.
int	vertexClicked () const
	Returns the number of the clicked vertex.
int	vertices (const bool &dropIsolates=false, const bool &countAll=false, const bool &recount=false)
	Gets the number of vertices in the graph.
int	vertexEdgesOutbound (int i)
	Returns the number of outbound edges (arcs) from vertex v1.
int	vertexEdgesInbound (int i)
	Returns the number of inbound edges (arcs) to vertex v1.
int	verticesWithOutboundEdges ()
	Returns the sum of vertices having edgesOutbound.
int	verticesWithInboundEdges ()
	Returns the sum of vertices having edgesInbound.
int	verticesWithReciprocalEdges ()
	Returns the sum of vertices having reciprocal edges.
QList< int >	verticesListIsolated ()
	Returns a list of all isolated vertices inside the graph.
QList< int >	verticesList ()
	Returns a list of all vertices numbers inside the graph.
QSet< int >	verticesSet ()
	Returns a QSet of all vertices numbers inside the graph.
void	verticesCreateSubgraph (QList< int > vList, const int &type=SUBGRAPH_CLIQUE, const int &center=0)
	Creates a subgraph (clique, star, cycle, line) with vertices in vList Iff vList is empty, then fallbacks to the m_verticesSelected.
GraphVertex *	vertexPtr (const int v)
int	edgesEnabled ()
	Returns the number of enabled ties in the current relation.
MyEdge	edgeClicked ()
	Returns clicked edge.
qreal	edgeExists (const int &v1, const int &v2, const bool &checkReciprocal=false)
	Checks if there is an edge from v1 to v2 and returns the weight, if the edge exists.
qreal	edgeExistsVirtual (const int &v1, const int &v2)
	Checks if there is an edge from v1 to v2, even weight = 0 and returns the weight, if the edge exists or RAND_MAX if the edge does not exist at all.
void	edgeOutboundStatusSet (const int &source, const int &target, const bool &toggle=false)
	Toggles the status of outbound edge source -> target at source vertex.
void	edgeInboundStatusSet (const int &target, const int &source, const bool &toggle=false)
	Toggles the status of inbound edge target <- source at target vertex.
void	notifyEdgeVisibilityChanged (const int &relation, const int &source, const int &target, const bool &toggle, const bool &preserveReverseEdge=false, const int &edgeWeight=1, const int &reverseEdgeWeight=1)
	Plain relay to signalSetEdgeVisibility, for GraphVertex (a QtCore-only value class, not a QObject) to notify the UI layer of a single edge's visibility change. See #WS3 M2.
void	notifyEdgesVisibilityBatch (const QList< EdgeVisibilityChange > &changes)
	Plain relay to signalSetEdgesVisibilityBatch, for GraphVertex to notify the UI layer of many edges' visibility changes as a single queued dispatch. See #WS3 M2.
void	edgeRemove (const int &v1, const int &v2, const bool &removeReverse=false)
	Removes the directed arc v1->v2 or, if the graph is undirected, the edge v1 <->v2.
void	edgeRemoveSelected (SelectedEdge &selectedEdge, const bool &removeReverse)
	Removes a SelectedEdge.
void	edgeRemoveSelectedAll ()
	Removes all selected edges.
bool	edgeSymmetric (const int &v1, const int &v2)
	Returns TRUE if edge(v1, v2) is symmetric, i.e. (v1,v2) == (v2,v1).
void	edgeTypeSet (const int &v1, const int &v2, const qreal &w, const int &dirType=EdgeType::Directed)
	Changes the direction type of an existing edge.
void	edgeWeightSet (const int &v1, const int &v2, const qreal &w, const bool &undirected=false)
	Changes the weight of the edge from vertex v1 to v2 (and optionally of the reverse edge).
qreal	edgeWeight (const int &v1, const int &v2) const
	Returns the weight of the edge v1->v2.
void	edgeWeightNumbersVisibilitySet (const bool &toggle)
	Changes the visibility of edge weight numbers.
void	edgeLabelSet (const int &v1, const int &v2, const QString &label)
	Changes the label of edge v1->v2.
QString	edgeLabel (const int &v1, const int &v2) const
	Returns the label of edge v1->v2.
void	edgeLabelsVisibilitySet (const bool &toggle)
	Toggles the visibility of edge labels.
void	edgeColorInit (const QString &)
	Saves the default edge color Used by random network creation methods.
void	edgeColorZeroInit (const QString &)
void	showZeroWeightEdgesSet (const bool &toggle)
void	edgeColorSet (const int &v1, const int &v2, const QString &color)
	Changes the color of edge v1->v2.
QString	edgeColor (const int &v1, const int &v2)
	Returns the color of the directed edge v1->v2.
bool	edgeColorAllSet (const QString &color, const int &threshold=RAND_MAX)
	Changes the color of all enabled edges.
void	edgeCustomAttributesSet (const int &v1, const int &v2, const QHash< QString, QString > &attrs)
	Sets all custom attributes on edge v1→v2, replacing any previously stored attributes for that edge.
QHash< QString, QString >	edgeCustomAttributes (const int &v1, const int &v2) const
	Returns the custom attributes stored on edge v1→v2. Returns an empty hash if no attributes have been set for that edge.
int	edgeAttributesImport (const QStringList &headers, const QVector< QStringList > &rows, int srcColumn, int tgtColumn)
	Imports custom attributes from a parsed table into existing edges.
QStringList	graphHasEdgeCustomAttributes () const
	Returns a list of all unique custom attribute keys present across all enabled edges in the current graph.
bool	isEmpty () const
	Returns true if the current graph has no vertices at all.
QList< int >	getSelectedVertices () const
	Returns a QList of user-selected vertices.
int	getSelectedVerticesCount () const
	Returns count of user-selected vertices.
int	getSelectedVerticesMin () const
	Returns min of user-selected vertices.
int	getSelectedVerticesMax () const
	Returns max of user-selected vertices.
QList< SelectedEdge >	getSelectedEdges () const
	Returns a QList of user-selected edges in pair<int,int>.
int	getSelectedEdgesCount () const
	Returns the count of user-selected edges.
int	getGeodesicsCount ()
	Returns the number of geodesics (shortest-paths) in the graph.
qreal	graphDensity ()
	Gets the graph density (if computed) or computes it again.
bool	isWeighted ()
	Returns true if the current relation has at least one edge with weight other than 0 or 1 (i.e. the relation is valued/weighted).
bool	isAnyRelationWeighted ()
	Returns true if any relation in the graph has at least one edge with weight other than 0 or 1.
void	setWeighted (const bool &toggle=true)
	Sets the graph to be weighted ( valued edges ).
qreal	graphReciprocity ()
	Gets the arc reciprocity of the graph.
bool	isSymmetric ()
	Returns TRUE if the adjacency matrix of the current relation is symmetric.
void	setSymmetric ()
	Transforms the graph to symmetric (all edges reciprocal).
void	addRelationSymmetricStrongTies (const bool &allRelations=false)
	Creates a new symmetric relation by keeping only strong-ties (mutual links) in the current relation. In the new relation, two actors are connected only if they are mutually connected in the current relation.
void	relationAddCocitation ()
	Creates a new symmetric relation by connecting those actors that are cocitated by others. In the new relation, an edge will exist between actor i and actor j only if C(i,j) > 0, where C the Cocitation Matrix. Thus the actor pairs cited by more common neighbors will appear with a stronger tie between them than pairs those cited by fewer common neighbors. The resulting relation is symmetric.
void	graphDichotomization (const qreal threshold)
	Creates a new binary relation in a valued network using edge dichotomization according to the threshold parameter.
void	setDirected (const bool &toggle=true, const bool &signalMW=true)
	Toggles the graph directed or undirected.
void	setUndirected (const bool &toggle=true, const bool &signalMW=true)
	Makes the graph undirected or directed.
bool	isDirected ()
	Returns true if graph is directed.
bool	isUndirected ()
	Returns true if graph is undirected.
bool	isConnected ()
	Checks if the graph is connected, in the sense of a topological space, i.e., there is a path from any vertex to any other vertex in the graph.
bool	isConnectedCached () const
	Returns true if the graph is connected, without recalculating it.
int	graphWeaklyConnectedComponents ()
	Counts weakly connected components using BFS.
int	graphWeaklyConnectedComponentsCached () const
int	graphStronglyConnectedComponents ()
	Counts strongly connected components using Tarjan's algorithm.
int	graphStronglyConnectedComponentsCached () const
const QHash< int, int > &	vertexComponentId () const
NodeConnectivityResult	graphNodeConnectivity (int source, int target, bool respectDirection)
	Local vertex connectivity kappa(source, target): the minimum number of nodes, other than source and target themselves, whose removal disconnects target from source.
int	graphConnectivity (bool respectDirection)
	Global vertex connectivity kappa(G): the minimum, over every non-adjacent pair of vertices, of their local vertex connectivity (graphNodeConnectivity()) - the network's worst-case robustness to node removal, i.e. the fewest nodes that would need to be removed to disconnect the network at its weakest point.
Matrix &	matrixAdjacency ()
Matrix &	matrixAdjacencyInverse ()
Matrix &	matrixDistances ()
Matrix &	matrixReachability ()
Matrix &	matrixWalks ()
Matrix &	matrixTotalWalks ()
Matrix &	matrixCliqueCoMembership ()
void	createMatrixAdjacency (const bool dropIsolates=false, const bool considerWeights=true, const bool inverseWeights=false, const bool symmetrize=false)
	Creates an adjacency matrix AM where AM(i,j)=1 if i is connected to j and AM(i,j)=0 if i not connected to j.
bool	createMatrixAdjacencyInverse (const QString &method="lu")
	Computes the inverse of the current adjacency matrix.
void	createMatrixSimilarityMatching (Matrix &AM, Matrix &SEM, const int &measure=METRIC_SIMPLE_MATCHING, const QString &varLocation="Rows", const bool &diagonal=false, const bool &considerWeights=true)
	Calls Matrix:similarityMatrix to compute the similarity matrix SCM of the variables (rows, columns, both) in given input matrix using the selected matching measure.
void	createMatrixSimilarityPearson (Matrix &AM, Matrix &PCC, const QString &varLocation="Rows", const bool &diagonal=false)
	The Pearson product-moment correlation coefficient (PPMCC, PCC or Pearson's r) is a measure of the linear dependence between two variables X and Y.
void	createMatrixDissimilarities (Matrix &INPUT_MATRIX, Matrix &DSM, const int &metric, const QString &varLocation, const bool &diagonal, const bool &considerWeights)
	Calls Matrix:distancesMatrix to compute the dissimilarities matrix DSM of the variables (rows, columns, both) in given input matrix using the user defined metric.
void	setReportsDataDir (const QString &reportsDir)
	Sets the directory where reports are saved This is used when exporting prominence distribution images to be used in HTML reports.
void	setReportsRealNumberPrecision (const int &precision)
	Sets the precision (number of fraction digits) the app will use when writing real numbers in reports.
void	setReportsLabelLength (const int &length)
	Sets the length of labels in reports.
void	setReportsChartType (const int &type)
	Sets the chart type in reports.
void	setReportsOutputFormat (const int &format)
	Sets the output format (HTML or CSV) new reports are written in.
void	writeDataSetToFile (const QString dir, const QString)
	Writes a "famous" dataset to the given file Datasets are hardcoded! They are exported in the given fileName...
void	writeMatrixAdjacencyTo (QTextStream &os, const bool &saveEdgeWeights=true)
bool	writeReciprocity (const QString fileName, const bool considerWeights=false, const int &format=ReportFormat::Html)
	Writes reciprocity report to filename.
bool	writeMatrix (const QString &fileName, const int &matrix=MATRIX_ADJACENCY, const bool &considerWeights=true, const bool &inverseWeights=false, const bool &dropIsolates=false, const QString &varLocation="Rows", const bool &simpler=false, const int &format=ReportFormat::Html)
	Computes and writes the specified matrix of the social network to an HTML report file.
void	writeMatrixHTMLTable (QTextStream &outText, Matrix &M, const bool &markDiag=true, const bool &plain=false, const bool &printInfinity=true, const bool &dropIsolates=false)
	Writes the matrix M as HTML.
void	writeMatrixCSVTable (QTextStream &outText, Matrix &M, const bool &printInfinity=true, const bool &dropIsolates=false)
	Writes the matrix M as a comma-separated table to the specified text stream outText.
bool	writeMatrixAdjacency (const QString fileName, const bool &markDiag=true, const int &format=ReportFormat::Html)
void	writeMatrixAdjacencyPlot (const QString fileName, const bool &simpler=false)
	Writes a visual representation of the adjacency matrix of the graph to the specified file.
bool	writeMatrixDissimilarities (const QString fileName, const QString &metricStr, const QString &varLocation, const bool &diagonal, const bool &considerWeights, const int &format=ReportFormat::Html)
	Writes dissimilarity matrix based on a metric/measure to given html file.
bool	writeMatrixSimilarityMatching (const QString fileName, const QString &measure="Simple", const QString &matrix="adjacency", const QString &varLocation="rows", const bool &diagonal=false, const bool &considerWeights=true, const int &format=ReportFormat::Html)
	Writes similarity matrix based on a matching measure to given html file.
bool	writeMatrixSimilarityPearson (const QString fileName, const bool considerWeights, const QString &matrix="adjacency", const QString &varLocation="rows", const bool &diagonal=false, const int &format=ReportFormat::Html)
	Calls Graph::createMatrixSimilarityPearson() and writes Pearson Correlation Coefficients to given file.
bool	writeEccentricity (const QString fileName, const bool considerWeights=false, const bool inverseWeights=false, const bool dropIsolates=false, const int &format=ReportFormat::Html)
	Writes the Eccentricity report to file.
void	writeScoreTableHTML (QTextStream &outText, const QStringList &dataColumnHeaders, const std::function< QVector< qreal >(GraphVertex *)> &rowValues, const std::function< bool(GraphVertex *)> &isBlanked=nullptr, const std::function< bool(GraphVertex *)> &isSkipped=nullptr)
	Renders a per-node score table as a sortable HTML.
void	writeScoreTableCSV (QTextStream &outText, const QStringList &dataColumnHeaders, const std::function< QVector< qreal >(GraphVertex *)> &rowValues, const std::function< bool(GraphVertex *)> &isBlanked=nullptr, const std::function< bool(GraphVertex *)> &isSkipped=nullptr)
	CSV sibling of writeScoreTableHTML() - same column shape and semantics, comma-delimited, no HTML markup, same RAND_MAX-as-infinity convention. Reuses TableExport::csvQuote() for the Label column, the only free-text field (Node numbers and scores are always numeric, so they never need escaping).
bool	writeCentralityDegree (const QString, const bool weights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the Degree Centrality to a file.
bool	writeCentralityCloseness (const QString, const bool weights, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the closeness centralities to a file.
bool	writeCentralityClosenessInfluenceRange (const QString, const bool weights, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the "improved" closeness centrality indices to a file.
bool	writeCentralityBetweenness (const QString, const bool weights, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes Betweenness centralities to file.
bool	writeCentralityPower (const QString, const bool weigths, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes Power Centralities to a file.
bool	writeCentralityStress (const QString, const bool weigths, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the Stress centralities to a file.
bool	writeCentralityEccentricity (const QString, const bool weigths, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the Eccentricity centralities (aka Harary Graph Centrality) to a file.
bool	writeCentralityInformation (const QString, const bool weigths, const bool inverseWeights, const int &format=ReportFormat::Html)
	Writes the information centralities to file.
bool	writeCentralityEigenvector (const QString, const bool &weigths=true, const bool &inverseWeights=false, const bool &dropIsolates=false, const int &format=ReportFormat::Html)
	Writes the eigenvector centralities to a file.
bool	writeCentralityKatz (const QString, const qreal &alpha, const bool &weigths=false, const bool &inverseWeights=false, const bool &dropIsolates=false, const int &format=ReportFormat::Html)
	Writes the Katz Centrality report to a file, then displays it.
bool	writeCentralityBonacich (const QString, const qreal &alpha, const qreal &beta, const bool &weigths=false, const bool &inverseWeights=false, const bool &dropIsolates=false, const int &format=ReportFormat::Html)
bool	writePrestigeDegree (const QString, const bool weights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the Degree Prestige of each node to a file.
bool	writePrestigeProximity (const QString, const bool weights, const bool inverseWeights, const bool dropIsolates, const int &format=ReportFormat::Html)
	Writes the proximity prestige indices to a file.
bool	writePrestigePageRank (const QString, const bool Isolates=false, const int &format=ReportFormat::Html)
	Writes the PageRank scores of vertices to a file.
bool	writeClusteringHierarchical (const QString &fileName, const QString &varLocation, const QString &matrix="Adjacency", const QString &metric="Manhattan", const QString &method="Complete", const bool &diagonal=false, const bool &dendrogram=false, const bool &considerWeights=true, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Performs Hierarchical Cluster Analysis (HCA) and writes the results to an HTML report file.
void	writeClusteringHierarchicalResultsToStream (QTextStream &outText, const int N, const bool &dendrogram=false)
	Writes Hierarchical Clustering results to given output stream Before running this method, the method Graph::graphClusteringHierarchical() must execute and return true. Otherwise, the result is unpredictable...
bool	writeCliqueCensus (const QString &fileName, const bool considerWeights)
	Calls graphCliques() to compute all cliques (maximal connected subgraphs) of the network. Then writes the results into a file, along with the Actor by clique analysis, the Co-membership matrix and the Hierarchical clustering of overlap matrix.
bool	writeClusteringCoefficient (const QString, const bool, const int &format=ReportFormat::Html)
	Writes the clustering coefficients to a file.
bool	writeTriadCensus (const QString, const bool, const int &format=ReportFormat::Html)
int	graphConnectednessFull (const bool updateProgress=false)
bool	graphReachable (const int &v1, const int &v2)
	Returns true if vertices v1 and v2 are reachable.
void	createMatrixReachability ()
	Creates the reachability matrix XRM.
int	graphDiameter (const bool considerWeights, const bool inverseWeights)
	Returns the diameter of the graph, aka the largest geodesic distance between any two vertices.
int	graphDiameterCached () const
	Returns the number of geodesics (shortest paths) in the graph, without recalculating it.
qreal	graphSumDistanceCached () const
	Returns the sum of all finite geodesic distances accumulated by DistanceEngine, without recalculating anything.
qreal	graphGeodesicsCountCached () const
	Returns the number of geodesics (shortest paths) accumulated by DistanceEngine, without recalculating anything.
int	graphDistanceGeodesic (const int &v1, const int &v2, const bool &considerWeights=false, const bool &inverseWeights=true)
	Returns the geodesic distance (length of shortest path) from vertex v1 to vertex v2.
qreal	apspDistance (const int &v1, const int &v2)
	Returns the already-computed geodesic distance from vertex v1 to vertex v2, for the current relation, without triggering a recompute (unlike graphDistanceGeodesic() above). RAND_MAX if either vertex is unknown or nothing has been computed yet for this relation.
int	apspShortestPaths (const int &v1, const int &v2)
	Returns the already-computed number of shortest paths from vertex v1 to vertex v2, for the current relation, without triggering a recompute. 0 if either vertex is unknown or nothing has been computed yet for this relation.
QMap< int, int >	graphGeodesicDistanceDistribution (const bool &considerWeights=false, const bool &inverseWeights=false)
	Returns a histogram of geodesic distances across all ordered vertex pairs.
bool	writeGeodesicDistribution (const QString &fileName, const bool &considerWeights=false, const bool &inverseWeights=false)
	Writes a geodesic distance distribution report to an HTML file.
QList< int >	graphGeodesicShortestPath (const int &v1, const int &v2, const bool &considerWeights=false, const bool &inverseWeights=false)
	Reconstructs one shortest path from vertex v1 to vertex v2.
qreal	graphDistanceGeodesicAverage (const bool considerWeights, const bool inverseWeights, const bool dropIsolates)
	Returns the average distance of the graph.
qreal	graphDistanceGeodesicAverageCached () const
	Returns the average geodesic distance of the graph, without recalculating it.
void	graphDistancesGeodesic (const bool &computeCentralities=false, const bool &considerWeights=false, const bool &inverseWeights=true, const bool &dropIsolates=false)
	Computes the geodesic distances between all vertices: In the process, it also computes many other centrality/prestige metrics:
void	notConnectedPairsClear ()
void	notConnectedPairsInsert (int from, int to)
int	notConnectedPairsSize () const
void	resetDistanceCentralityCacheFlags ()
void	setSymmetricCached (bool v)
bool	symmetricCached () const
void	setConnectedCached (bool v)
void	setDiameterCached (int v)
void	resetDistanceAggregates ()
void	addToDistanceSum (qreal delta)
void	incGeodesicsCount ()
void	addGeodesicsCount (int n)
void	setAverageDistanceCached (qreal v)
bool	graphMatrixDistanceGeodesicCreate (const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Creates the matrix DM of geodesic distances between vertices.
void	graphMatrixShortestPathsCreate (const bool &considerWeights=false, const bool &inverseWeights=true, const bool &dropIsolates=false)
	Creates the matrix SIGMA of shortest paths (geodesics) between vertices Each SIGMA(i,j) is the number of shortest paths (geodesics) from i and j.
int	getProminenceIndexByName (const QString &prominenceIndexName)
	Returns the IndexType of the given prominence index name Called from MW::slotEditNodeFind, MW::slotLayoutRadialByProminenceIndex etc.
void	prominenceDistribution (const int &index, const ChartType &type, const QString &distImageFileName=QString())
	Computes the distribution of a centrality index score. The distribution is stored as Qt Series depending on the SeriesType parameter type It is send to MW through signal/slot.
void	prominenceDistributionBars (const H_StrToInt &discreteClasses, const QString &name, const QString &distImageFileName)
	Computes the prominence distribution and delegates Bar chart rendering.
void	prominenceDistributionArea (const H_StrToInt &discreteClasses, const QString &name, const QString &distImageFileName)
	Computes prominence distribution data and delegates Area chart rendering.
void	prominenceDistributionSpline (const H_StrToInt &discreteClasses, const QString &seriesName, const QString &distImageFileName)
	Computes prominence distribution data and delegates Spline chart rendering.
void	centralityDegree (const bool &considerWeights=true, const bool &dropIsolates=false)
	Calculates the degree (outDegree) centrality of each vertex - diagonal included.
void	centralityInformation (const bool considerWeights=false, const bool inverseWeights=false)
	Computes the Information centrality of each vertex - diagonal included Note that there is no known generalization of Stephenson & Zelen's theory for information centrality to directional data.
void	centralityEigenvector (const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Computes Eigenvector centrality.
qreal	estimateSpectralRadius (const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Estimates the adjacency matrix's spectral radius (dominant eigenvalue magnitude).
void	centralityKatz (const qreal &alpha, const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Computes Katz Centrality of each vertex, with attenuation factor alpha.
void	centralityBonacich (const qreal &alpha, const qreal &beta, const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Computes Bonacich Power Centrality of each vertex, with parameters alpha and beta.
void	centralityClosenessIR (const bool considerWeights=false, const bool inverseWeights=false, const bool dropIsolates=false)
	Computes an "improved" closeness centrality index, IRCC, which can be used on disconnected graphs. IRCC is an improved node-level centrality closeness index which focuses on the influence range of each node (the set of nodes that are reachable from it) For each node v, this index calculates the fraction of nodes in its influence range and divides it by the average distance of those nodes from v, ignoring nodes that are not reachable from it.
void	prestigeDegree (const bool &considerWeights, const bool &dropIsolates=false)
	Computes the Degree Prestige (in-degree) of each vertex - diagonal included Also the mean value and the variance of the in-degrees.
void	prestigePageRank (const bool &dropIsolates=false)
	Calculates the PageRank Prestige of each vertex.
void	prestigeProximity (const bool considerWeights=false, const bool inverseWeights=false, const bool dropIsolates=false)
	Computes Proximity Prestige of each vertex Also the mean value and the variance of it..
bool	isCentralityIndexComputed (const IndexType index) const
	Returns true if the given centrality/prestige index has been computed.
int	walksBetween (int v1, int v2, int length)
	Calculates and returns the number of walks of a given length between v1 and v2.
void	graphWalksMatrixCreate (const int &N=0, const int &length=0, const bool &dropIsolates=false, const bool &considerWeights=false, const bool &inverseWeights=false, const bool &symmetrize=false)
	Computes either the "Walks of given length" or the "Total Walks" matrix. If length>0, it computes the Walks of given length matrix, XM=AM^l where each element (i,j) denotes the number of walks of length l between vertex i and j. If length=0, it computes the Total Walks matrix, XSM=Sum{AM^n} where each (i,j) denotes the total number of walks of any length between vertices i and j. NOTE: In the latter case, this function is VERY SLOW on large networks (n>50), since it will calculate all powers of the sociomatrix up to n-1 in order to find out all possible walks.
bool	writeMatrixWalks (const QString &fn, const int &length=0, const bool &simpler=false, const int &format=ReportFormat::Html)
	Writes the walks of given length matrix to a file in HTML. If length = 0, it writes the Total Walks matrix.
qreal	numberOfTriples (int v1)
void	graphCliques (QSet< int > R=QSet< int >(), QSet< int > P=QSet< int >(), QSet< int > X=QSet< int >())
	Finds all maximal cliques in the graph using the Bron–Kerbosch algorithm with Tomita et al. (2006) pivot selection.
void	graphCliqueAdd (const QList< int > &clique)
	Called from Graph::graphCliques to add a new clique (list of vertices) Adds clique info to each clique member and updates co-membership matrix CLQM .
int	graphCliquesContaining (const int &actor, const int &size=0)
int	graphCliquesOfSize (const int &size)
	Graph::graphCliquesOfSize Returns the number of maximal cliques of a given size.
bool	graphClusteringHierarchical (Matrix &STR_EQUIV, const QString &varLocation, const int &metric, const int &method, const bool &diagonal=false, const bool &diagram=false, const bool &considerWeights=true, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Performs an hierarchical clustering process (Johnson, 1967) on a given NxN distance/dissimilarity matrix. The input matrix can be the the adjacency matrix, the geodesic distance matrix or a derived from them dissimilarities matrix using a user-specified metric, i.e. euclidean distance. The method parameter defines how to compute distances (similarities) between a new cluster the old clusters. Valid values can be:
qreal	clusteringCoefficientLocal (const int &v1)
	Returns the local clustering coefficient (CLUCOF) of vertex v1.
qreal	clusteringCoefficient ()
	Computes local clustering coefficients and returns the network average Clustering Coefficient.
bool	graphTriadCensus ()
	Conducts a triad census and updates QList::triadTypeFreqs, which is the list carrying all triad type frequencies Complexity: O(n³) — three nested loops each bounded by N.
void	triadType_examine_MAN_label (int, int, int, GraphVertex *, GraphVertex *, GraphVertex *)
const QList< int > &	graphTriadTypeFreqs () const
bool	hasCalculatedTriadCensus () const
void	layoutRandom ()
	Repositions all nodes on random positions Emits setNodePos(i, x,y) to tell GW that the node item should be moved.
void	layoutRadialRandom (const bool &guides=true)
	Repositions all nodes on the periphery of different circles with random radius.
void	layoutEgoRadial (const int egoVertex)
	Ego-centered radial layout.
void	layoutCircular (const double &x0, const double &y0, const double &newRadius, const bool &guides=false)
	Repositions all nodes on the periphery of a circle with given radius.
void	layoutByProminenceIndex (int prominenceIndex, int layoutType, const bool &considerWeights=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
	Applies a layout according to each actor's prominence index score. The layout type can be radial (0), level (1), node sizes (2) or node colors (3), as follows: layoutType=0 - Repositions all nodes on the periphery of concentric circles with radius analogous to their prominence index layoutType=1 - Repositions all nodes on different top-down levels according to their centrality layoutType=2 - Changes node sizes to be proportional to their prominence index score layoutType=2 - Changes node colors to reflect their prominence index score (from red to green).
void	layoutVertexSizeByIndegree ()
	Convenience method Changes the size of all nodes to be proportional to their InDegree (Degree Prestige) Calls layoutByProminenceIndex.
void	layoutVertexSizeByOutdegree ()
	Convenience method Changes the size of all nodes to be proportional to their outDegree (Degree Centrality) Calls layoutByProminenceIndex.
void	layoutForceDirectedSpringEmbedder (const int maxIterations)
	Embeds a Force Directed Placement layout according to the initial Spring Embedder model proposed by Eades.
void	layoutForceDirectedFruchtermanReingold (const int maxIterations)
	Embeds a Force Directed Placement layout according to the Fruchterman-Reingold model. Fruchterman and Reingold (1991) refined the Spring Embedder model by replacing the forces. In this model, "the vertices behave as atomic particles or celestial bodies, exerting attractive and repulsive forces on one another." (ibid). Again, only vertices that are neighbours attract each other but, unlike Spring Embedder, all vertices repel each other. These forces induce movement. The algorithm might resemble molecular or planetary simulations, sometimes called n-body problems.
void	layoutForceDirectedKamadaKawai (const int maxIterations=500, const bool considerWeights=false, const bool inverseWeights=false, const bool dropIsolates=false, const QString &initialPositions="current")
	Embeds a Force Directed Placement layout according to the Kamada-Kawai model. In this model, the network is considered to be a dynamic system where every two actors are 'particles' mutually connected by a 'spring'. Each spring has a desirable length, which corresponds to their graph theoretic distance. In this way, the optimal layout of the graph is the state with the minimum imbalance. The degree of imbalance is formulated as the total spring energy: the square summation of the differences between desirable distances and real ones for all pairs of particles Initially, the particles/actors are placed on the vertices of a regular n-polygon.
qreal	graphDistanceEuclidean (const QPointF &a, const QPointF &b)
	Computes the euclideian distance between QPointF a and b.
qreal	graphDistanceEuclidean (const QPointF &a)
	the euclideian distance of QPointF a (where a is difference vector)
int	sign (const qreal &D)
	Graph::sign returns the sign of number D as integer (1 or -1).
qreal	layoutForceDirected_F_rep (const QString model, const qreal &dist, const qreal &optimalDistance)
	Computes the repulsive force between two vertices for force-directed layouts.
qreal	layoutForceDirected_F_att (const QString model, const qreal &dist, const qreal &optimalDistance)
	Computes the attractive (spring) force between two adjacent vertices.
qreal	layoutForceDirected_Eades_moveNodes (const qreal &c4)
	Moves all vertices to their new positions as computed by the Eades Spring Embedder model.
qreal	layoutForceDirected_FR_moveNodes (const qreal &temperature)
	Moves all vertices to their new positions as computed by the Fruchterman-Reingold model.
qreal	layoutForceDirected_FR_temperature (const int iteration) const
	Reduces the temperature as the layout approaches a better configuration.
qreal	computeOptimalDistance (const int &V)
	Computes the optimal inter-vertex distance for force-directed layouts.
void	compute_angles (const QPointF &Delta, const qreal &dist, qreal &angle1, qreal &angle2, qreal &degrees1, qreal &degrees2)
	Graph::compute_angles Computes the two angles of the orthogonal triangle shaped by two points of difference vector DV and distance dist A = 90 B = angle1 C = angle2.
void	webCrawlTerminateThreads (QString reason)
	called from Graph, when closing network, to terminate all crawler processes Also called indirectly when wc_spider finishes
void	randomizeThings ()
	Adds a little universal randomness :).
bool	randomNetErdosCreate (const int &N, const QString &model, const int &m, const qreal &p, const QString &mode, const bool &diag)
	Creates an Erdős–Rényi random network.
bool	randomNetScaleFreeCreate (const int &N, const int &power, const int &m0, const int &m, const qreal &alpha, const QString &mode)
	Creates a Barabási–Albert scale-free random network.
bool	randomNetSmallWorldCreate (const int &N, const int &degree, const double &beta, const QString &mode)
	Creates a Watts–Strogatz small-world random network.
bool	randomNetRingLatticeCreate (const int &N, const int &degree, const bool updateProgress=false)
	Creates a random ring lattice network.
bool	randomNetRegularCreate (const int &N, const int &degree, const QString &mode, const bool &diag)
	Creates a pseudo-random d-regular network.
bool	randomNetLatticeCreate (const int &N, const int &length, const int &dimension, const int &nei, const QString &mode, const bool &circular)
	Creates a lattice (mesh/grid) random network.
int	factorial (int)
bool	progressCanceled () const
	Returns true if the user has requested cancellation via the progress dialog.
void	resetProgressCanceled ()
	Resets the cancellation status.

Public Attributes
H_Int	vpos

Protected Member Functions
void	progressStatus (const QString &msg)
	Emits a status message to be shown in the UI status bar.
void	runOnGuiThread (std::function< void()> fn)
	Runs fn on the application's main/GUI thread, regardless of the calling thread.
void	uiProminenceDistributionSpline (const QVector< QPair< qreal, qreal > > &points, qreal min, qreal max, qreal minF, qreal maxF, const QString &seriesName, const QString &distImageFileName)
void	uiProminenceDistributionArea (const QVector< QPair< qreal, qreal > > &points, const qreal min, const qreal max, const qreal minF, const qreal maxF, const QString &name, const QString &distImageFileName)
void	uiProminenceDistributionBars (const QStringList &categories, const QVector< qreal > &frequencies, const qreal min, const qreal max, const qreal minF, const qreal maxF, const QString &name, const QString &distImageFileName)

Private Member Functions
Graph *	subgraphFromVertexList (const QList< int > &vertexNums, const QString &name, const bool &includeCustomAttributes=true)
	Core helper: builds an independent Graph copy from an explicit list of vertex numbers.
void	applyVisibilitySnapshot (const GraphVisibilitySnapshot &snap)
	Restores vertex and edge visibility to the state recorded in snap.
void	vertexFilterReplaySpec (const FilterSpec &spec)
	Re-applies one filter from its stored replay parameters.
void	edgeAdd (const int &v1, const int &v2, const qreal &weight, const int &type, const QString &label, const QString &color)
	Adds a directed arc from v1 to v2 into the internal graph data structures.
void	dijkstra (const int &s, const int &si, const bool &computeCentralities=false, const bool &inverseWeights=false, const bool &dropIsolates=false)
void	minmax (qreal C, GraphVertex *v, qreal &max, qreal &min, int &maxNode, int &minNode)
	Computes minimum and maximum centralities during graphDistancesGeodesic().
void	resolveClasses (qreal C, H_StrToInt &discreteClasses, int &classes)
	Checks if score C is a new prominence class If yes, it stores that number in a QHash<QString,int> type where the score is the key. If no, increases the frequency of this prominence score by 1 Called from graphDistancesGeodesic().
void	resolveClasses (qreal C, H_StrToInt &discreteClasses, int &classes, int name)
	Overloaded method. It only adds displaying current vertex for debugging purposes.
void	layoutRandomInMemory ()
	Repositions all vertices at random coordinates without emitting any signals.

Private Attributes
QThread	file_parserThread
QThread	webcrawlerThread
VList	m_graph
Parser *	file_parser
WebCrawler *	web_crawler
QQueue< QUrl > *	urlQueue
int	m_crawler_max_urls
int	m_crawler_visited_urls
QList< QString >	m_relationsList
QList< bool >	m_relationsDirected
QList< int >	m_graphFileFormatExportSupported
QList< int >	triadTypeFreqs
QList< int >	m_verticesList
QList< int >	m_verticesIsolatedList
QList< int >	m_verticesSelected
QSet< int >	m_verticesSet
QList< SelectedEdge >	m_selectedEdges
QStack< GraphVisibilitySnapshot >	m_visibilityHistory
QMultiHash< int, int >	m_vertexPairsNotConnected
QHash< int, int >	m_vertexPairsUnilaterallyConnected
QMultiMap< int, L_int >	m_cliques
QHash< int, QSet< int > >	neighboursHash
QList< qreal >	m_clusteringLevel
QMap< int, V_int >	m_clustersPerSequence
QMap< QString, V_int >	m_clustersByName
QMap< int, V_str >	m_clusterPairNamesPerSeq
Matrix	SIGMA
Matrix	DM
Matrix	invAM
Matrix	AM
Matrix	invM
Matrix	WM
Matrix	XM
Matrix	XSM
Matrix	XRM
Matrix	CLQM
QHash< int, Matrix >	m_apspDist
QHash< int, Matrix >	m_apspSigma
H_StrToInt	discreteDPs
H_StrToInt	discreteSDCs
H_StrToInt	discreteCCs
H_StrToInt	discreteBCs
H_StrToInt	discreteSCs
H_StrToInt	discreteIRCCs
H_StrToInt	discreteECs
H_StrToInt	discreteEccentricities
H_StrToInt	discretePCs
H_StrToInt	discreteICs
H_StrToInt	discretePRPs
H_StrToInt	discretePPs
H_StrToInt	discreteEVCs
H_StrToInt	discreteCLCs
H_StrToInt	discreteKCs
H_StrToInt	discreteBPCs
QString	m_reportsDataDir
int	m_reportsRealPrecision
int	m_reportsLabelLength
ChartType	m_reportsChartType
ReportFormat	m_reportsOutputFormat
int	m_fieldWidth
int	m_curRelation
int	m_fileFormat
int	m_vertexClicked
MyEdge	m_clickedEdge
qreal	edgeWeightTemp
qreal	edgeReverseWeightTemp
qreal	meanSDC
qreal	varianceSDC
qreal	meanSCC
qreal	varianceSCC
qreal	meanIRCC
qreal	varianceIRCC
qreal	meanSBC
qreal	varianceSBC
qreal	meanSSC
qreal	varianceSSC
qreal	meanEC
qreal	varianceEC
qreal	meanSPC
qreal	varianceSPC
qreal	meanIC
qreal	varianceIC
qreal	meanEVC
qreal	varianceEVC
qreal	meanKC
qreal	varianceKC
qreal	meanBPC
qreal	varianceBPC
qreal	meanSDP
qreal	varianceSDP
qreal	meanPP
qreal	variancePP
qreal	meanPRP
qreal	variancePRP
qreal	minEccentricity
qreal	maxEccentricity
qreal	minSDP
qreal	maxSDP
qreal	sumDP
qreal	sumSDP
qreal	groupDP
qreal	minSDC
qreal	maxSDC
qreal	sumDC
qreal	sumSDC
qreal	groupDC
qreal	minSCC
qreal	maxSCC
qreal	nomSCC
qreal	denomSCC
qreal	sumCC
qreal	sumSCC
qreal	groupCC
qreal	maxIndexCC
qreal	minIRCC
qreal	maxIRCC
qreal	nomIRCC
qreal	denomIRCC
qreal	sumIRCC
qreal	groupIRCC
qreal	minSBC
qreal	maxSBC
qreal	nomSBC
qreal	denomSBC
qreal	sumBC
qreal	sumSBC
qreal	groupSBC
qreal	maxIndexBC
qreal	minSPC
qreal	maxSPC
qreal	nomSPC
qreal	denomSPC
qreal	t_sumIC
qreal	sumSPC
qreal	groupSPC
qreal	maxIndexPC
qreal	minSSC
qreal	maxSSC
qreal	sumSC
qreal	sumSSC
qreal	groupSC
qreal	maxIndexSC
qreal	minEC
qreal	maxEC
qreal	nomEC
qreal	denomEC
qreal	sumEC
qreal	groupEC
qreal	maxIndexEC
qreal	minIC
qreal	maxIC
qreal	nomIC
qreal	denomIC
qreal	sumIC
qreal	maxIndexIC
qreal	minEVC
qreal	maxEVC
qreal	nomEVC
qreal	denomEVC
qreal	sumEVC
qreal	sumSEVC
qreal	groupEVC
qreal	minKC
qreal	maxKC
qreal	sumKC
qreal	m_lastKatzAlpha = -1
qreal	minBPC
qreal	maxBPC
qreal	sumBPC
qreal	m_lastBonacichAlpha = -1
qreal	m_lastBonacichBeta = 0
qreal	minPRP
qreal	maxPRP
qreal	nomPRC
qreal	denomPRC
qreal	sumPC
qreal	t_sumPRP
qreal	sumPRP
qreal	minPP
qreal	maxPP
qreal	nomPP
qreal	denomPP
qreal	sumPP
qreal	groupPP
qreal	minCLC
qreal	maxCLC
qreal	averageCLC
qreal	varianceCLC
qreal	d_factor
int	maxNodeCLC
int	minNodeCLC
int	classesSDP
int	maxNodeDP
int	minNodeDP
int	classesSDC
int	maxNodeSDC
int	minNodeSDC
int	classesSCC
int	maxNodeSCC
int	minNodeSCC
int	classesIRCC
int	maxNodeIRCC
int	minNodeIRCC
int	classesSBC
int	maxNodeSBC
int	minNodeSBC
int	classesSPC
int	maxNodeSPC
int	minNodeSPC
int	classesSSC
int	maxNodeSSC
int	minNodeSSC
int	classesEC
int	maxNodeEC
int	minNodeEC
int	classesEccentricity
int	maxNodeEccentricity
int	minNodeEccentricity
int	classesIC
int	maxNodeIC
int	minNodeIC
int	classesPRP
int	maxNodePRP
int	minNodePRP
int	classesPP
int	maxNodePP
int	minNodePP
int	classesCLC
int	classesEVC
int	maxNodeEVC
int	minNodeEVC
int	classesKC
int	maxNodeKC
int	minNodeKC
int	classesBPC
int	maxNodeBPC
int	minNodeBPC
int	m_graphModStatus
int	m_reserveEdgesPerVertexSize
int	m_totalVertices
int	m_totalEdges
int	m_graphDiameter
int	initVertexSize
int	initVertexLabelSize
int	initVertexNumberSize
int	initVertexNumberDistance
int	initVertexLabelDistance
bool	order
bool	initEdgeWeightNumbers
bool	initEdgeLabels
qreal	m_graphAverageDistance
qreal	m_graphGeodesicsCount
qreal	m_graphDensity
qreal	m_graphSumDistance
qreal	m_graphReciprocityArc
qreal	m_graphReciprocityDyad
int	m_graphReciprocityTiesReciprocated
int	m_graphReciprocityTiesNonSymmetric
int	m_graphReciprocityTiesTotal
int	m_graphReciprocityPairsReciprocated
int	m_graphReciprocityPairsTotal
bool	m_graphHasVertexCustomIcons
int	outboundEdgesVert
int	inboundEdgesVert
int	reciprocalEdgesVert
qreal	canvasWidth
qreal	canvasHeight
bool	calculatedEdges
bool	calculatedVertices
bool	calculatedVerticesList
bool	calculatedVerticesSet
bool	m_verticesCacheDropIsolates = false
bool	m_verticesCacheCountAll = false
bool	calculatedAdjacencyMatrix
bool	calculatedDistances
bool	calculatedCentralities
bool	calculatedIsolates
bool	calculatedEVC
bool	calculatedKC
bool	calculatedBPC
bool	calculatedDP
bool	calculatedDC
bool	calculatedPP
bool	calculatedIRCC
bool	calculatedIC
bool	calculatedPRP
bool	calculatedTriad
bool	calculatedGraphSymmetry
bool	calculatedGraphReciprocity
bool	calculatedGraphDensity
bool	calculatedGraphWeighted
std::atomic< bool >	m_progressCanceled
bool	m_graphIsDirected
bool	m_graphIsSymmetric
bool	m_graphIsWeighted
bool	m_graphIsConnected
int	m_graphWeaklyConnectedComponents
int	m_graphStronglyConnectedComponents
QHash< int, int >	m_vertexComponentId
int	csRecDepth
QString	m_fileName
QString	m_graphName
QString	initEdgeColor
QString	initEdgeColorZero
QString	initVertexColor
QString	initVertexNumberColor
QString	initVertexLabelColor
bool	initShowZeroWeightEdges
QString	initVertexShape
QString	initVertexIconPath
QString	htmlHead
QString	htmlHeadLight
QString	htmlEnd
QDateTime	actualDateTime

Friends
class	DistanceEngine

Detailed Description

The Graph class This is the main class for a Graph, used in conjuction with GraphVertex, Parser and Matrix objects. Graph class methods are the interface to various analysis algorithms GraphVertex class holds each vertex data (colors, strings, statistics, etc) Matrix class holds the adjacency matrix of the network. Parser class loads files of networks.

Member Enumeration Documentation

Clustering

enum Graph::Clustering

Enumerator
Single_Linkage
Complete_Linkage
Average_Linkage

ModStatus

enum Graph::ModStatus

Enumerator
NewNet
SavedUnchanged
MinorOptions
VertexMetadata
EdgeMetadata
VertexPositions
MajorChanges
VertexCount
EdgeCount
VertexEdgeCount

NodeConnectivityStatus

enum class Graph::NodeConnectivityStatus

strong

Enumerator
Ok
Adjacent
Invalid

Constructor & Destructor Documentation

Graph()

Graph::Graph	(	const int &	reserveVerticesSize = 5000,
		const int &	reserveEdgesPerVertexSize = 500 )

Constructs a Graph.

~Graph()

Graph::~Graph

(

)

Destroys the Graph object.

Member Function Documentation

addGeodesicsCount()

void Graph::addGeodesicsCount

(

int

n

)

addGuideCircle

void Graph::addGuideCircle ( const double & , const double & , const double &  )

signal

addGuideHLine

void Graph::addGuideHLine ( const double & y0 )

signal

addRelationSymmetricStrongTies()

void Graph::addRelationSymmetricStrongTies

(

const bool &

allRelations = false

)

Creates a new symmetric relation by keeping only strong-ties (mutual links) in the current relation. In the new relation, two actors are connected only if they are mutually connected in the current relation.

Parameters

allRelations

addToDistanceSum()

void Graph::addToDistanceSum

(

qreal

delta

)

applyVisibilitySnapshot()

void Graph::applyVisibilitySnapshot ( const GraphVisibilitySnapshot & snap )

private

Restores vertex and edge visibility to the state recorded in snap.

Private helper shared by vertexFilterRestoreAll() and vertexFilterRemoveAt(). Does not touch m_visibilityHistory — callers manage the stack.

apspDistance()

qreal Graph::apspDistance	(	const int &	v1,
		const int &	v2 )

Returns the already-computed geodesic distance from vertex v1 to vertex v2, for the current relation, without triggering a recompute (unlike graphDistanceGeodesic() above). RAND_MAX if either vertex is unknown or nothing has been computed yet for this relation.

apspShortestPaths()

int Graph::apspShortestPaths	(	const int &	v1,
		const int &	v2 )

Returns the already-computed number of shortest paths from vertex v1 to vertex v2, for the current relation, without triggering a recompute. 0 if either vertex is unknown or nothing has been computed yet for this relation.

canvasMaxRadius

double Graph::canvasMaxRadius ( ) const

slot

Gets the max radius of the canvas.

Returns

double

canvasMinDimension

qreal Graph::canvasMinDimension ( ) const

slot

Gets the min dimensions of the canvas.

Returns

qreal

canvasRandomX

double Graph::canvasRandomX ( ) const

slot

Returns a random x-coordinate adjusted to be visible inside the canvas usable area.

Returns

double

canvasRandomY

double Graph::canvasRandomY ( ) const

slot

Returns a random y-coordinate adjusted to be visible inside the canvas usable area.

Returns

double

canvasSizeSet

void Graph::canvasSizeSet ( const int & width, const int & height )

slot

Sets the size of the canvas.

Called when the MW is resized to update canvasWidth/canvasHeight, and node positions

Parameters

w
h

canvasSizeSetQuiet

void Graph::canvasSizeSetQuiet ( const qreal & width, const qreal & height )

slot

Sets the canvas size without rescaling node positions.

Used by Graph::subgraphExtract() to propagate the source graph's canvas dimensions to the new Graph object so that IO routines normalize coordinates correctly on export.

Parameters

width	New canvas width in pixels.
height	New canvas height in pixels.

canvasVisibleX

double Graph::canvasVisibleX ( const double & x ) const

slot

Checks if x is visible inside the canvas usable area and if not returns an adjusted x-coordinate.

Parameters

x

Returns

double

canvasVisibleY

double Graph::canvasVisibleY ( const double & y ) const

slot

Checks if y is visible inside the canvas usable area and if not returns an adjusted y-coordinate.

Parameters

y

Returns

double

centralityBonacich()

void Graph::centralityBonacich	(	const qreal &	alpha,
		const qreal &	beta,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Computes Bonacich Power Centrality of each vertex, with parameters alpha and beta.

Meaning: like Katz Centrality, credit for indirect connections discounted by distance - but with a twist: beta can be negative, in which case being tied to well-connected others can hurt rather than help a node's score (e.g. a buyer connected to powerful sellers has less bargaining power the more powerful those sellers are). alpha is a free overall scale factor that doesn't change the relative ranking, only the numbers' size.

When to use: whenever the sign of "connections to well-connected others" itself matters (bargaining power, exchange networks) rather than always being a good thing - Katz Centrality and Eigenvector Centrality both assume more indirect connections is always better.

Weights: like Katz and Eigenvector Centrality, this is walk-counting, not shortest-path routing - don't invert a strength-type weight (it would make your strongest ties contribute the least); only invert if the weight genuinely represents a cost.

Compare to: Katz Centrality (KC, see centralityKatz()) is the same walk-counting idea with a single always-positive decay parameter; Bonacich generalizes it with a second, possibly negative parameter beta and a free scale factor alpha. Not to be confused with Power Centrality (PC, Gil-Schmidt, see graphDistancesGeodesic()), a same-sounding but unrelated measure.

Math: b = alpha * (I - beta*R)^-1 * R * 1, where R = A^T (the transpose only matters for directed graphs, same convention as Katz). Only beta must satisfy |beta| < 1/lambda_max(A) for the underlying geometric series to converge; alpha is an unconstrained outer scale factor. Unlike every other centrality measure in this app, b(i) can come out negative when beta is negative - that is a genuine property of the measure, not a bug, and SBPC (b divided by max b) is not guaranteed to land in [0,1] in that case.

Parameters

alpha
beta
considerWeights
inverseWeights
dropIsolates

centralityClosenessIR()

void Graph::centralityClosenessIR	(	const bool	considerWeights = false,
		const bool	inverseWeights = false,
		const bool	dropIsolates = false )

Computes an "improved" closeness centrality index, IRCC, which can be used on disconnected graphs. IRCC is an improved node-level centrality closeness index which focuses on the influence range of each node (the set of nodes that are reachable from it) For each node v, this index calculates the fraction of nodes in its influence range and divides it by the average distance of those nodes from v, ignoring nodes that are not reachable from it.

Meaning: how efficiently can an actor reach the part of the network it can reach? Plain closeness centrality breaks down on a disconnected network, since "distance to everyone" is undefined once some actors are unreachable. IRCC sidesteps that by only ever averaging distances to actors that are actually reachable, so it stays meaningful even when the network is split into several disconnected pieces.

When to use: any time the network might not be fully connected - social media crawls, partial datasets, or naturally fragmented networks (e.g. isolated friend clusters) - where plain Closeness Centrality (CC) would otherwise be undefined or require dropping actors.

Weights: shortest-path-based, same as CC - if a weight represents value/strength, invert it so a strong tie behaves like a short/cheap path.

Compare to: Closeness Centrality (CC, see graphDistancesGeodesic()) is the classic version this generalizes - use CC directly once the graph is known to be fully connected. Proximity Prestige (PP, see prestigeProximity()) is prestige's directed-graph counterpart of this same idea (distance from others, rather than to them).

Math: for actor i, let J_i be the set of nodes reachable from i (its influence range). IRCC(i) = [ |J_i| / (N-1) ] / [ (sum of d(i,j) for j in J_i) / |J_i| ] - the fraction of the network i can reach, divided by the average distance to that reachable set.

Parameters

considerWeights
inverseWeights
dropIsolates

centralityDegree()

void Graph::centralityDegree	(	const bool &	considerWeights = true,
		const bool &	dropIsolates = false )

Calculates the degree (outDegree) centrality of each vertex - diagonal included.

Meaning: the simplest centrality there is - how many direct connections (ties) does this actor have? More connections means more prominence, full stop; no attention is paid to who those connections are or how the rest of the network is shaped.

When to use: a fast first-pass screen for "who's active/popular" on any graph, directed or not, connected or not - cheap to compute and easy to explain to a non-technical audience. Weak at finding brokers or long-range influence; pair with betweenness or eigenvector centrality when those matter.

Compare to: Degree Prestige (DP, see prestigeDegree()) is this same idea restricted to inbound ties only - meaningful on directed graphs, where "connections this actor made" and "connections this actor received" can differ a lot.

Weights: no inversion choice here (considerWeights only) - when considered, weights are summed directly, so a stronger tie always adds more, regardless of what the weight means.

Math: DC(i) = number of edges incident to i (or the sum of their weights, if weights are considered). Standardized SDC(i) = DC(i) / (N-1), the fraction of all other actors i is directly tied to.

Parameters

considerWeights
dropIsolates

centralityEigenvector()

void Graph::centralityEigenvector	(	const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Computes Eigenvector centrality.

Meaning: not just how many connections you have, but how important your connections are - being tied to a few highly-connected people can outscore being tied to many poorly-connected ones. It is a "popularity feeds back on itself" measure: your score depends on your neighbors' scores, which depend on their neighbors' scores, and so on until the whole network settles into one stable ranking.

When to use: ranking actors by prestige/influence rather than raw activity - e.g. finding the truly influential accounts in a social network (as opposed to merely the most active ones), or citation-network-style "important papers cited by other important papers" reasoning. Needs a connected (or largely connected) graph to produce a meaningful ranking.

Weights: not shortest-path routing - the adjacency matrix entry is used directly as "how much this tie contributes." Don't invert a strength-type weight (that would make your strongest ties count least); only invert if the weight genuinely represents a cost.

Compare to: Katz Centrality (see centralityKatz()) and Bonacich Power Centrality (see centralityBonacich()) generalize this same "connections to well-connected others matter" idea by summing actual walks with a distance-based decay factor instead of finding the dominant eigenvector directly, which makes the decay directly tunable by the user rather than fixed by the network's own structure. Power Centrality (PC, Gil-Schmidt, see graphDistancesGeodesic()) is a different, older, similarly-named but unrelated measure computed straight from vertex degrees rather than via eigen-decomposition.

Math: the eigenvector centrality vector x is the dominant eigenvector of the adjacency matrix A, i.e. the vector solving A*x = lambda_max*x for the largest eigenvalue lambda_max. Computed here via power iteration (Matrix::powerIteration()): start from any vector, repeatedly multiply by A and rescale to unit length - it converges to that eigenvector.

Parameters

considerWeights
inverseWeights

centralityInformation()

void Graph::centralityInformation	(	const bool	considerWeights = false,
		const bool	inverseWeights = false )

Computes the Information centrality of each vertex - diagonal included Note that there is no known generalization of Stephenson & Zelen's theory for information centrality to directional data.

Meaning: most centrality measures only look at the single shortest path between two actors. Information centrality gives an actor credit for every path connecting it to others, not just the best one - shorter, less roundabout paths count for more - so an actor sitting on many decent alternative routes can score as centrally as one sitting on the single best route.

When to use: when redundancy of connection matters more than the single fastest route - e.g. assessing how robust an actor's information access is to a single link failing, or comparing actors whose "one best path" scores (closeness, betweenness) look similar but whose actual web of alternative routes differs a lot.

Compare to: Closeness Centrality (CC) also scores actors by distance to others, but only via the single shortest path; Betweenness (BC) counts how often an actor sits on shortest paths between others. Information Centrality is the only one of the three that uses every connecting path, not just the shortest one.

Weights: unlike CC/BC/SC/EC/PC/IRCC, this is not shortest-path routing - the raw weight feeds directly into the B matrix (B_ij = 1 - w_ij), where a larger raw weight already pulls two actors closer together. Don't invert a strength-type weight here (that would weaken your strongest ties); only invert if the weight genuinely represents a cost.

Math: build matrix B where B_ii = 1 + (sum of i's edge weights) and B_ij = 1 - w_ij for i != j, then invert it to get C = B^-1. Information centrality is IC(i) = 1 / [ C_ii + (tr(C) - 2*R) / n ], where tr(C) is the trace of C and R is the sum of any one row of C (interchangeable by construction of B).

Parameters

considerWeights
inverseWeights

centralityKatz()

void Graph::centralityKatz	(	const qreal &	alpha,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Computes Katz Centrality of each vertex, with attenuation factor alpha.

Meaning: credit for indirect connections, discounted the further away they are. Unlike degree centrality, which only counts direct ties, Katz centrality also counts friends-of-friends, friends-of-friends-of-friends, and so on - but each extra hop counts for less, shrunk by alpha (a direct tie counts full price, a 2-hop tie counts alpha times as much, a 3-hop tie counts alpha^2 times as much).

When to use: ranking actors by total reach through the network, with a user-tunable "how much does distance matter" knob - useful when the natural decay eigenvector centrality settles on isn't the one you want, or the graph is disconnected/sparse enough that eigenvector centrality's power iteration doesn't produce a useful ranking.

Weights: like Eigenvector Centrality, this is walk-counting, not shortest-path routing - the adjacency matrix entry is used directly as tie strength. Don't invert a strength-type weight (that would make your strongest ties contribute the least to every walk); only invert if the weight genuinely represents a cost.

Compare to: Eigenvector Centrality (EVC, see centralityEigenvector()) generalizes the same "connections to well-connected others matter" idea via eigen-decomposition instead of an explicit decay parameter. Bonacich Power Centrality (see centralityBonacich()) extends this same formula with a second parameter beta that can be negative, flipping whether being connected to well-connected others helps or hurts. Not to be confused with Power Centrality (PC, Gil-Schmidt, see graphDistancesGeodesic()), a same-sounding but unrelated measure.

Math: C_Katz(i) = sum_{k=1}^inf sum_j alpha^k (A^k)_ji, i.e. the discounted count of walks of every length arriving at i, since (A^k)_ij is the number of length-k walks between i and j. Computed in closed form via the matrix identity I + xA + x^2*A^2 + x^3*A^3 + ... = (I - xA)^-1 (valid for |x| < 1/lambda_max, the same geometric-series identity used for ordinary numbers, applied to matrices): C_Katz = ((I - alpha*A^T)^-1 - I) * 1 (the transpose only matters for directed graphs; for undirected graphs A = A^T). alpha must satisfy |alpha| < 1/lambda_max(A), where lambda_max is the adjacency matrix's dominant eigenvalue (see estimateSpectralRadius()), or the underlying geometric series does not converge and the inversion is singular/meaningless.

Parameters

alpha
considerWeights
inverseWeights
dropIsolates

clear()

void Graph::clear

(

const QString &

reason = ""

)

Clears all vertices.

Parameters

reason

clusteringCoefficient()

qreal Graph::clusteringCoefficient

(

)

Computes local clustering coefficients and returns the network average Clustering Coefficient.

Parameters

updateProgress

Returns

clusteringCoefficientLocal()

qreal Graph::clusteringCoefficientLocal

(

const int &

v1

)

Returns the local clustering coefficient (CLUCOF) of vertex v1.

Meaning: not a measure of how prominent an actor is, but of how tightly-knit its neighborhood is - do this actor's friends also know each other, or is the actor the only thing connecting a set of strangers? A score near 1 means the neighborhood is a tight clique; near 0 means the actor bridges otherwise-unconnected people.

When to use: studying local cohesion and "small world" structure - identifying tightly-knit cliques, triadic closure, or (as the actor-level building block of the graph-wide average) assessing how clustered a network is overall. Not a prominence/ranking measure like the centrality and prestige indices above - a high-CLC actor isn't necessarily important, just embedded in a dense neighborhood.

For undirected (symmetric) graphs, uses the Watts–Strogatz formula:

C_i = 2 * |{e_jk : v_j, v_k ∈ N_i, e_jk ∈ E}| / ( k_i * (k_i - 1) )

where N_i is the set of direct neighbours of v_i and k_i = |N_i|. Because the graph is symmetric every undirected edge is counted once (the reverse-edge dedup guard in the inner loop is preserved).

For directed (asymmetric) graphs, uses the generalisation described in the SocNetV manual (equivalent to Watts–Strogatz extended to digraphs):

C_i = |{e_jk : v_j, v_k ∈ N_i, e_jk ∈ E}| / ( k_i * (k_i - 1) )

where N_i is the UNION of in-neighbours and out-neighbours of v_i (excluding v_i itself), and each directed edge e_jk is counted independently – i.e. e_jk and e_kj are distinct.

Bug fix (issue #58): the previous implementation built N_i from reciprocalEdgesHash() (only mutual ties), which gave C_i = 0 for nodes whose in- and out-neighbourhood were not identical. N_i must be the full combined neighbourhood for directed networks.

Parameters

v1	The vertex number whose local CLUCOF is requested.

Returns

The local clustering coefficient in [0, 1], or 0 for isolates and vertices with fewer than 2 neighbours.

compute_angles()

void Graph::compute_angles	(	const QPointF &	DV,
		const qreal &	dist,
		qreal &	angle1,
		qreal &	angle2,
		qreal &	degrees1,
		qreal &	degrees2 )

Graph::compute_angles Computes the two angles of the orthogonal triangle shaped by two points of difference vector DV and distance dist A = 90 B = angle1 C = angle2.

Parameters

DV
dist
angle1
angle2
degrees1
degrees2

computeOptimalDistance()

qreal Graph::computeOptimalDistance

(

const int &

V

)

Computes the optimal inter-vertex distance for force-directed layouts.

The optimal distance (also called k or naturalLength) represents the radius of the empty area ideally surrounding each vertex. It is derived from the available canvas area per vertex, plus the vertex's own diameter.

Formula: vertexArea = ceil( sqrt( canvasArea / V ) ) — side of a square tile per vertex optimalDistance = vertexDiameter + vertexArea

This gives the distance at which repulsive and attractive forces balance for a uniformly distributed graph. For denser graphs or larger N, vertexArea shrinks (fewer pixels per node), so the returned value decreases — callers may apply their own scaling multiplier on top (e.g. FR uses C=0.9).

Parameters

V	Number of active vertices in the graph.

Returns

Optimal inter-vertex spacing in canvas pixels.

createMatrixAdjacency()

void Graph::createMatrixAdjacency	(	const bool	dropIsolates = false,
		const bool	considerWeights = true,
		const bool	inverseWeights = false,
		const bool	symmetrize = false )

Creates an adjacency matrix AM where AM(i,j)=1 if i is connected to j and AM(i,j)=0 if i not connected to j.

Parameters

dropIsolates
considerWeights
inverseWeights
symmetrize

createMatrixAdjacencyInverse()

bool Graph::createMatrixAdjacencyInverse

(

const QString &

method = "lu"

)

Computes the inverse of the current adjacency matrix.

Parameters

method

Returns

createMatrixDissimilarities()

void Graph::createMatrixDissimilarities	(	Matrix &	INPUT_MATRIX,
		Matrix &	DSM,
		const int &	metric,
		const QString &	varLocation,
		const bool &	diagonal,
		const bool &	considerWeights )

Calls Matrix:distancesMatrix to compute the dissimilarities matrix DSM of the variables (rows, columns, both) in given input matrix using the user defined metric.

Parameters

INPUT_MATRIX
DSM
metric
varLocation
diagonal
considerWeights

createMatrixReachability()

void Graph::createMatrixReachability

(

)

Creates the reachability matrix XRM.

createMatrixSimilarityMatching()

void Graph::createMatrixSimilarityMatching	(	Matrix &	AM,
		Matrix &	SCM,
		const int &	measure = METRIC_SIMPLE_MATCHING,
		const QString &	varLocation = "Rows",
		const bool &	diagonal = false,
		const bool &	considerWeights = true )

Calls Matrix:similarityMatrix to compute the similarity matrix SCM of the variables (rows, columns, both) in given input matrix using the selected matching measure.

Parameters

AM
SCM
rows

createMatrixSimilarityPearson()

void Graph::createMatrixSimilarityPearson	(	Matrix &	AM,
		Matrix &	PCC,
		const QString &	varLocation = "Rows",
		const bool &	diagonal = false )

The Pearson product-moment correlation coefficient (PPMCC, PCC or Pearson's r) is a measure of the linear dependence between two variables X and Y.

As a normalized version of the covariance, the PPMCC is computed with the formula: r =\frac{\sum ^n _{i=1}(x_i - \bar{x})(y_i - \bar{y})}{\sqrt{\sum ^n _{i=1}(x_i - \bar{x})^2} \sqrt{\sum ^n _{i=1}(y_i - \bar{y})^2}}

It gives a value between +1 and −1 inclusive, where 1 is total positive linear correlation, 0 is no linear correlation, and −1 is total negative linear correlation.

In SNA, Pearson correlations can be used to track the similarity between actors, in terms of structural equivalence.

This method creates an actor by actor NxN matrix PCC where the (i,j) element is the Pearson correlation coefficient of actor i and actor j. If the input matrix is the adjacency matrix, the PCC of two nodes measures how related (similar, inverse or not related at all) their patterns of ties tend to be. A positive value means there is strong linear association of the two actors, while a negative value means the inverse. For instance a value of -1 means the two actors have exactly opposite ties to other actors, while a value of 1 means the actors have identical patterns of ties to other actors (they are connected to the same actors).

The correlation measure of similarity is particularly useful when the data on ties are valued

Parameters

AM
PCC
rows

dijkstra()

void Graph::dijkstra ( const int & s, const int & si, const bool & computeCentralities = false, const bool & inverseWeights = false, const bool & dropIsolates = false )

private

methods used by graphDistancesGeodesic()

edgeAdd()

void Graph::edgeAdd ( const int & v1, const int & v2, const qreal & weight, const int & type, const QString & label, const QString & color )

private

Adds a directed arc from v1 to v2 into the internal graph data structures.

This is the low-level storage function. It does NOT signal the UI. All UI signaling is handled by the caller (edgeCreate / signalDrawEdge).

Internally, each vertex maintains two adjacency lists:

• OutEdges: arcs going OUT from this vertex
• InEdges: arcs coming IN to this vertex

For a directed arc v1->v2:

• v1's OutEdges gains v2 (with weight, color, label)
• v2's InEdges gains v1 (with weight)

For an undirected edge (EdgeType::Undirected), the edge is stored as two symmetric directed arcs:

• v1->v2 (OutEdge of v1, InEdge of v2)
• v2->v1 (OutEdge of v2, InEdge of v1) This matches SocNetV's internal convention: undirected edges are always represented as a pair of reciprocal directed arcs in the adjacency structure.

For EdgeType::Reciprocated, the reverse arc v2->v1 already exists in the data model (created earlier as a directed arc). No additional storage is needed here; the caller is responsible for updating type semantics if required.

Weight handling: If weight != 1 and weight != 0, the graph is marked as weighted. Weight == 0 is treated as a visual-only "null" edge drawn in initEdgeColorZero (#30).

Parameters

v1	Source node number (external node number, not internal index)
v2	Target node number (external node number, not internal index)
weight	Edge weight
type	EdgeType::Directed, Undirected, or Reciprocated
label	Edge label (optional, stored on the out-arc of v1)
color	Edge color (optional, stored on the out-arc of v1)

edgeAttributesImport()

int Graph::edgeAttributesImport	(	const QStringList &	headers,
		const QVector< QStringList > &	rows,
		int	srcColumn,
		int	tgtColumn )

Imports custom attributes from a parsed table into existing edges.

Each row is matched to an edge by looking up vertex numbers from srcColumn and tgtColumn. All other columns become custom attributes on the matched edge. Rows that do not match any existing edge are skipped.

Example — CSV input with srcColumn=0, tgtColumn=1. Native editable columns (Weight, Label, Color) are routed to their setters; read-only native columns (Relation) are silently skipped:

Source,Target,Weight,Label,Color,relationship,strength
1,2,0.8,,#666666,invested_in,strong   ← Weight/Color updated; relationship/strength → custom attrs
2,3,0.5,,#666666,mentors,medium

Returns

Number of edges that received at least one attribute update.

edgeClicked()

MyEdge Graph::edgeClicked

(

)

Returns clicked edge.

Returns

edgeClickedSet

void Graph::edgeClickedSet ( const int & v1, const int & v2, const bool & openMenu = false )

slot

Sets the clicked edge.

Parameters are the source and target node of the edge. It emits signal to MW, which displays a relevant message on the status bar.

Parameters

v1
v2

edgeColor()

QString Graph::edgeColor	(	const int &	v1,
		const int &	v2 )

Returns the color of the directed edge v1->v2.

Parameters

v1
v2

Returns

edgeColorAllSet()

bool Graph::edgeColorAllSet	(	const QString &	color,
		const int &	threshold = RAND_MAX )

Changes the color of all enabled edges.

Parameters

color

Returns

edgeColorInit()

void Graph::edgeColorInit

(

const QString &

color

)

Saves the default edge color Used by random network creation methods.

Parameters

color

edgeColorSet()

void Graph::edgeColorSet	(	const int &	v1,
		const int &	v2,
		const QString &	color )

Changes the color of edge v1->v2.

Parameters

v1
v2
color

edgeColorZeroInit()

void Graph::edgeColorZeroInit

(

const QString &

color

)

edgeCreate

bool Graph::edgeCreate ( const int & v1, const int & v2, const qreal & weight, const QString & color, const int & type = 0, const bool & drawArrows = true, const bool & bezier = false, const QString & label = QString(), const bool & signalMW = true, const QHash< QString, QString > & edgeCustomAttributes = QHash<QString,QString>() )

slot

Checks if edge (v1,v2) already exists, then creates it and signals the UI to draw it.

Slots to signals from GraphicsWidget and Parser

This is the main entry point for edge creation, called from:

• Parser::signalCreateEdge (when loading a network file)
• MainWindow (when user clicks "add link" button)
• GraphicsWidget (when user middle-clicks two nodes)

Edge type semantics:

• EdgeType::Undirected : stores both v1->v2 and v2->v1 internally (via edgeAdd), draws a single undirected edge in the UI.
• EdgeType::Directed : stores only v1->v2. If v2->v1 already exists, upgrades both to EdgeType::Reciprocated.
• EdgeType::Reciprocated: both directions exist with equal weight.

Mixed-section Pajek files (*Arcs followed by *Edges): Some Pajek files declare directed arcs in an *Arcs section AND undirected edges in a subsequent *Edges section. The same node pair (v1,v2) may appear in both. When *Arcs are processed first, v1->v2 is stored as a directed arc. When *Edges is processed later, edgeExists(v1,v2) returns non-zero, so we must NOT silently drop the request. If v2->v1 is still missing, we store the reverse arc and upgrade the existing forward arc visual to Reciprocated.

In that special case:

• We store the reverse arc v2->v1 using the undirected edge's weight/color/label. The forward arc (from *Arcs) is left untouched to preserve its original data.
• We emit signalDrawEdge(v2, v1, ..., Reciprocated) — note the swapped arguments. GraphicsWidget::drawEdge() with Reciprocated looks up the existing arc via edgeKey(targetNum, sourceNum). Since edgeKey() is order-sensitive (packs relation/v1/v2 in that order), passing (v2,v1) makes targetNum=v1, sourceNum=v2, so the lookup resolves to edgeKey(v1,v2) which matches the key stored during *Arcs processing. No new GraphicsEdge is created; setDirectionType() is called instead.
• drawArrows is forced true: a reciprocated edge always shows arrows on both ends regardless of what the undirected request's drawArrows value was.
• We do NOT set m_graphIsDirected: adding the missing reverse makes the pair more symmetric, not less. Directedness is determined by the overall graph state after full import completes.
• We return true because a new arc was genuinely added to the graph.

Parameters

v1	Source node number
v2	Target node number
weight	Edge weight
color	Edge color
type	Edge type: EdgeType::Undirected, Directed, or Reciprocated
drawArrows	Whether to draw arrowheads in the UI
bezier	Whether to draw the edge as a bezier curve
label	Edge label (optional)
signalMW	Whether to signal MainWindow after modifying graph state

Returns

true if a new edge (or reverse arc) was created, false if fully skipped

edgeCreateWebCrawler

void Graph::edgeCreateWebCrawler ( const int & source, const int & target )

slot

Called from WebCrawler when it finds an new link Calls edgeCreate() method with initEdgeColor.

Parameters

source
target

edgeCustomAttributes()

QHash< QString, QString > Graph::edgeCustomAttributes	(	const int &	v1,
		const int &	v2 ) const

Returns the custom attributes stored on edge v1→v2. Returns an empty hash if no attributes have been set for that edge.

Parameters

v1	Source vertex number.
v2	Target vertex number.

edgeCustomAttributesSet()

void Graph::edgeCustomAttributesSet	(	const int &	v1,
		const int &	v2,
		const QHash< QString, QString > &	attrs )

Sets all custom attributes on edge v1→v2, replacing any previously stored attributes for that edge.

Parameters

v1	Source vertex number.
v2	Target vertex number.
attrs	Key/value map of custom attributes.

edgeExists()

qreal Graph::edgeExists	(	const int &	v1,
		const int &	v2,
		const bool &	checkReciprocal = false )

Checks if there is an edge from v1 to v2 and returns the weight, if the edge exists.

Complexity: O(logN) for vpos retrieval + O(1) for QList index retrieval + O(logN) for checking edge(v2)

Parameters

v1
v2
reciprocated	if true, checks if the edge is reciprocated (v1<->v2) with the same weight

Returns

zero if edge or reciprocated edge does not exist or non-zero if arc /reciprocated edge exists

edgeExistsVirtual()

qreal Graph::edgeExistsVirtual	(	const int &	v1,
		const int &	v2 )

Checks if there is an edge from v1 to v2, even weight = 0 and returns the weight, if the edge exists or RAND_MAX if the edge does not exist at all.

This is only used in GraphML saving if the user has selected the Settings option to save zero-weight edges

See also

https://github.com/socnetv/app/issues/151

Parameters

v1
v2

edgeFilterByAttribute

void Graph::edgeFilterByAttribute ( const FilterCondition & cond )

slot

Hides all edges whose custom attribute does not satisfy cond.

Non-destructive: pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes. Undo via vertexFilterRemoveAt() or vertexFilterRestoreAll().

Returns early with a status message if no edge satisfies the condition.

Parameters

cond	The filter condition (key, operator, value).

edgeFilterByQuery

void Graph::edgeFilterByQuery ( const GraphQuery & query )

slot

Shows only edges whose custom attributes satisfy ALL conditions in query.

AND logic: an edge must pass every condition to remain visible. Non-destructive: pushes a GraphVisibilitySnapshot with FilterSpec::Type::EdgeQuery. Undo via vertexFilterRemoveAt() or vertexFilterRestoreAll().

Parameters

query

Compound query; all conditions should have Scope::Edges.

edgeFilterByWeight

void Graph::edgeFilterByWeight ( const qreal m_threshold, const bool overThreshold )

slot

Hides edges whose weight does not satisfy the threshold condition.

Non-destructive: pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes, so the filter can be undone via vertexFilterRemoveAt().

Parameters

m_threshold	The weight threshold.
overThreshold	If true, keep edges with weight >= threshold (hide the rest); if false, keep edges with weight <= threshold.

edgeFilterReset

void Graph::edgeFilterReset ( )

slot

Re-enables all edges in the current relation.

Restores full edge visibility after a weight filter has been applied. Non-destructive: only visibility is changed, no data is modified.

edgeFilterUnilateral

void Graph::edgeFilterUnilateral ( const bool & toggle )

slot

Enables or disables unilateral edges in current relationship.

If toggle=true, all non-reciprocal edges are disabled, effectively making the network symmetric.

Parameters

toggle

edgeInboundStatusSet()

void Graph::edgeInboundStatusSet	(	const int &	target,
		const int &	source,
		const bool &	toggle = false )

Toggles the status of inbound edge target <- source at target vertex.

Parameters

v1
v2
toggle

Returns

edgeLabel()

QString Graph::edgeLabel	(	const int &	v1,
		const int &	v2 ) const

Returns the label of edge v1->v2.

Parameters

v1
v2

Returns

edgeLabelSet()

void Graph::edgeLabelSet	(	const int &	v1,
		const int &	v2,
		const QString &	label )

Changes the label of edge v1->v2.

Parameters

v1
v2
weight

edgeLabelsVisibilitySet()

void Graph::edgeLabelsVisibilitySet

(

const bool &

toggle

)

Toggles the visibility of edge labels.

Parameters

toggle

edgeOutboundStatusSet()

void Graph::edgeOutboundStatusSet	(	const int &	source,
		const int &	target,
		const bool &	toggle = false )

Toggles the status of outbound edge source -> target at source vertex.

Parameters

v1
v2
toggle

Returns

edgeRemove()

void Graph::edgeRemove	(	const int &	v1,
		const int &	v2,
		const bool &	removeReverse = false )

Removes the directed arc v1->v2 or, if the graph is undirected, the edge v1 <->v2.

Emits signal to GW to delete the graphics item.

Parameters

v1
v2
removeReverse	if true also removes the reverse edge

edgeRemoveSelected()

void Graph::edgeRemoveSelected	(	SelectedEdge &	selectedEdge,
		const bool &	removeReverse )

Removes a SelectedEdge.

Parameters

selectedEdge
removeReverse

edgeRemoveSelectedAll()

void Graph::edgeRemoveSelectedAll

(

)

Removes all selected edges.

edgesEnabled()

int Graph::edgesEnabled

(

)

Returns the number of enabled ties in the current relation.

IMPORTANT: Naming vs semantics

• Internally, SocNetV stores adjacency as directed arcs in each vertex's out-edges.
• For an UNDIRECTED graph, each undirected edge is represented as TWO symmetric arcs (v1->v2 and v2->v1). Therefore, summing outEdgesCount() over all vertices yields 2*E, and we must divide by 2 to return the logical undirected edge count E.
• For a DIRECTED graph, summing outEdgesCount() over all vertices yields A (the arc count), and we return it as-is.

Caching:

• m_totalEdges caches the internal count (sum of enabled out-arcs).
• edgesEnabled() returns the logical count:
  • E for undirected graphs
  • A for directed graphs

TODO / THINK: Self-loops (v->v)

• A self-loop contributes exactly 1 outbound arc in outEdgesCount().
• In an UNDIRECTED graph, dividing m_totalEdges by 2 assumes every tie is a symmetric pair. A loop is NOT a symmetric pair, so it would be mishandled by the /2 rule.
• Decide on a loop policy: (a) forbid loops in undirected graphs (and filter them out here), or (b) count loops separately and adjust the formula to: E = (nonLoopArcs/2) + loopArcs (where loopArcs is the number of enabled v->v arcs).

Returns

int Logical enabled ties: E (undirected) or A (directed)

edgeSymmetric()

bool Graph::edgeSymmetric	(	const int &	v1,
		const int &	v2 )

Returns TRUE if edge(v1, v2) is symmetric, i.e. (v1,v2) == (v2,v1).

Parameters

v1
v2

Returns

edgeTypeSet()

void Graph::edgeTypeSet	(	const int &	v1,
		const int &	v2,
		const qreal &	weight,
		const int &	dirType = EdgeType::Directed )

Changes the direction type of an existing edge.

Parameters

v1
v2
weight

edgeWeight()

qreal Graph::edgeWeight	(	const int &	v1,
		const int &	v2 ) const

Returns the weight of the edge v1->v2.

Parameters

v1
v2

Returns

qreal

edgeWeightNumbersVisibilitySet()

void Graph::edgeWeightNumbersVisibilitySet

(

const bool &

toggle

)

Changes the visibility of edge weight numbers.

Parameters

toggle

edgeWeightSet()

void Graph::edgeWeightSet	(	const int &	v1,
		const int &	v2,
		const qreal &	weight,
		const bool &	undirected = false )

Changes the weight of the edge from vertex v1 to v2 (and optionally of the reverse edge).

Parameters

v1
v2
weight
undirected

estimateSpectralRadius()

qreal Graph::estimateSpectralRadius	(	const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Estimates the adjacency matrix's spectral radius (dominant eigenvalue magnitude).

Meaning: how "amplifying" the network's connectivity structure is - the factor by which a signal spreading through the network grows per step, in the worst case. Used to validate the convergence bound for measures built on a geometric series of the adjacency matrix (Katz Centrality, Bonacich Power Centrality): their attenuation parameter must satisfy |x| < 1/lambda_max, or the underlying series - and the matrix inversion computing it in closed form - does not converge.

When to use: before running Katz or Bonacich, to know (or show the user) the actual valid range for alpha/beta on this specific network, rather than a generic guess - dense/large networks can have a lambda_max well into double digits, making the valid range far smaller than it would be for a sparse one.

Compare to: centralityEigenvector() also power-iterates the adjacency matrix to find its dominant eigenvector, and computes this same lambda_max as a side effect - this method exists separately because callers here (Katz, Bonacich, and their dialogs) only need the eigenvalue, not the ranking eigenvector itself.

Math: via power iteration (Matrix::powerIteration()), starting from a unit seed vector - ||Ax|| approximates lambda_max once the iteration converges (Perron-Frobenius: for a non-negative matrix like an adjacency matrix, this is exact for a connected network).

Parameters

considerWeights
inverseWeights
dropIsolates

Returns

the estimated spectral radius, or 0 if the network is empty or the computation was canceled. 0 conventionally means "no bound - any value converges" to callers.

factorial()

int Graph::factorial

(

int

x

)

Calculates and returns x! factorial... used in (n 2)p edges calculation

filterSpecList

QList< FilterSpec > Graph::filterSpecList ( ) const

slot

Returns the FilterSpec list in application order (oldest first).

Available for consumers that need to inspect the active filter sequence (e.g. Phase 1 DialogQueryBuilder prefill). MainWindow currently tracks chip labels separately via m_nodeFilterChips. QStack<T> inherits QVector<T>, so at(0) is the bottom (oldest) entry.

getFileFormat()

int Graph::getFileFormat

(

)

const

Returns the format of the last file opened.

Returns

int

getFileName()

QString Graph::getFileName

(

)

const

Returns the file name of the current graph, if any.

Returns

QString

getGeodesicsCount()

int Graph::getGeodesicsCount

(

)

Returns the number of geodesics (shortest-paths) in the graph.

Returns

int

getName()

QString Graph::getName

(

)

const

Returns the name of the current graph.

If graph name is empty, then returns current relation name. If no relation exists, returns "noname"

Returns

QString

getProminenceIndexByName()

int Graph::getProminenceIndexByName

(

const QString &

prominenceIndexName

)

Returns the IndexType of the given prominence index name Called from MW::slotEditNodeFind, MW::slotLayoutRadialByProminenceIndex etc.

Parameters

prominenceIndexName

getSelectedEdges()

QList< SelectedEdge > Graph::getSelectedEdges

(

)

const

Returns a QList of user-selected edges in pair<int,int>.

Returns

getSelectedEdgesCount()

int Graph::getSelectedEdgesCount

(

)

const

Returns the count of user-selected edges.

Returns

getSelectedVertices()

QList< int > Graph::getSelectedVertices

(

)

const

Returns a QList of user-selected vertices.

Returns

getSelectedVerticesCount()

int Graph::getSelectedVerticesCount

(

)

const

Returns count of user-selected vertices.

Returns

getSelectedVerticesMax()

int Graph::getSelectedVerticesMax

(

)

const

Returns max of user-selected vertices.

Returns

getSelectedVerticesMin()

int Graph::getSelectedVerticesMin

(

)

const

Returns min of user-selected vertices.

Returns

getThread()

QThread * Graph::getThread

(

)

const

graphClickedEmptySpace

void Graph::graphClickedEmptySpace ( const QPointF & p )

slot

Resets the clicked edge and node.

Usually, called when the user clicks on an empty space.

Parameters

p

graphCliqueAdd()

void Graph::graphCliqueAdd

(

const QList< int > &

clique

)

Called from Graph::graphCliques to add a new clique (list of vertices) Adds clique info to each clique member and updates co-membership matrix CLQM .

Parameters

list

Returns

graphCliques()

void Graph::graphCliques	(	QSet< int >	R = QSet<int>(),
		QSet< int >	P = QSet<int>(),
		QSet< int >	X = QSet<int>() )

Finds all maximal cliques in the graph using the Bron–Kerbosch algorithm with Tomita et al. (2006) pivot selection.

— Algorithm overview —

The Bron–Kerbosch algorithm [1] is a recursive backtracking procedure that maintains three disjoint vertex sets at each call:

R — the clique built so far (all vertices in R are mutually adjacent). P — candidate vertices that can still extend R (each is adjacent to all of R). X — vertices already processed that are also adjacent to all of R (used to avoid reporting the same clique more than once).

When both P and X are empty, R cannot be extended and no super-set of R was reported before — so R is a maximal clique.

— Pivot selection (Tomita et al., 2006) —

Without pivoting the algorithm iterates over every vertex in P at each level, leading to a worst-case exponential blow-up even for graphs with few cliques.

Tomita, Tanaka & Takahashi [2] proved that choosing a pivot vertex u ∈ P∪X that maximises |N(u) ∩ P| (the size of u's neighbourhood intersected with P) allows the main loop to enumerate only the vertices in P \ N(u) — the non-neighbours of u inside the candidate set.

Why this is correct: any maximal clique that extends R must contain at least one vertex from P \ N(u), because if a clique contained only neighbours of u it could be extended by u itself (since u is adjacent to all of them and to all of R), contradicting maximality. So we lose no cliques by restricting the loop to P \ N(u).

Why this is faster: the pivot u was chosen to maximise |N(u) ∩ P|, which minimises |P \ N(u)|. In the best case (a dense graph) |P \ N(u)| ≈ 1, reducing each level of recursion to a single branch. For sparse graphs the improvement is smaller but still significant in practice.

References: [1] Bron, C. & Kerbosch, J. (1973). "Algorithm 457: Finding all cliques of an undirected graph." Commun. ACM, 16(9), 575–577. [2] Tomita, E., Tanaka, A. & Takahashi, H. (2006). "The worst-case time complexity for generating all maximal cliques and computational experiments." Theoretical Computer Science, 363(1), 28–42. https://doi.org/10.1016/j.tcs.2006.06.015

Parameters

R	Current clique under construction (vertices already chosen).
P	Candidate vertices that can extend R.
X	Excluded vertices (already processed at this level).

graphCliquesContaining()

int Graph::graphCliquesContaining	(	const int &	actor,
		const int &	size = 0 )

Returns the number of maximal cliques which include a given actor

graphCliquesOfSize()

int Graph::graphCliquesOfSize

(

const int &

size

)

Graph::graphCliquesOfSize Returns the number of maximal cliques of a given size.

Parameters

size

Returns

graphClusteringHierarchical()

bool Graph::graphClusteringHierarchical	(	Matrix &	STR_EQUIV,
		const QString &	varLocation,
		const int &	metric,
		const int &	method,
		const bool &	diagonal = false,
		const bool &	diagram = false,
		const bool &	considerWeights = true,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Performs an hierarchical clustering process (Johnson, 1967) on a given NxN distance/dissimilarity matrix. The input matrix can be the the adjacency matrix, the geodesic distance matrix or a derived from them dissimilarities matrix using a user-specified metric, i.e. euclidean distance. The method parameter defines how to compute distances (similarities) between a new cluster the old clusters. Valid values can be:

• Clustering::Single_Linkage: "single-link" or "connectedness" or "minimum"
• Clustering::Complete_Linkage: "complete-link" or "diameter" or "maximum"
• Clustering::Average_Linkage: "average-link" or UPGMA

  Parameters

  matrix
  metric
  method
  considerWeights
  inverseWeights
  dropIsolates

graphClusteringMethodStrToType()

int Graph::graphClusteringMethodStrToType

(

const QString &

method

)

const

Helper method, return clustering method type from the human readable name of it.

Parameters

method

Returns

graphClusteringMethodTypeToString()

QString Graph::graphClusteringMethodTypeToString

(

const int &

methodType

)

const

Helper method, return the human readable name of clustering method type.

Returns

graphConnectednessFull()

int Graph::graphConnectednessFull

(

const bool

updateProgress = false

)

graphConnectivity()

int Graph::graphConnectivity

(

bool

respectDirection

)

Global vertex connectivity kappa(G): the minimum, over every non-adjacent pair of vertices, of their local vertex connectivity (graphNodeConnectivity()) - the network's worst-case robustness to node removal, i.e. the fewest nodes that would need to be removed to disconnect the network at its weakest point.

Naive pairwise-minimum algorithm: iterate all non-adjacent pairs, tracking the minimum local connectivity seen so far, starting that minimum at the cheap degree bound below and pruning with an early exit once it can't go any lower. This is O(n^2) local-connectivity computations in the worst case (O(n) for directed "strong" mode, since kappa(s,t) can differ from kappa(t,s) so ordered pairs must be tested) - deliberately the simple approach rather than the smarter O(n) algorithm (Even 1975, which fixes one vertex and reuses far fewer max-flow computations): SocNetV's networks and this feature's usage pattern (an occasional, user-triggered analysis, not a hot path) don't call for that extra complexity yet, and the two pruning steps below already cut the common cases down a lot in practice.

Two pruning steps, both used here:

• Fast path: if the graph is already disconnected (per graphWeaklyConnectedComponents() / graphStronglyConnectedComponents(), whichever matches respectDirection), kappa(G) = 0 immediately, with no max-flow computation at all. This is purely an optimization, not a correctness requirement: the pair loop below would reach the same answer on its own the moment it tests the unreachable pair, since local connectivity between an unreachable pair is 0 by construction (max-flow finds no augmenting path) - but checking the already-cached component count first avoids running any flow computation for the common case of an obviously fragmented network.
• Degree bound (Whitney's inequality): kappa(G) <= delta(G), the minimum vertex degree (its directed analogue: kappa(D) <= min over v of min(indeg(v), outdeg(v))). Seeding the running minimum with this bound and only ever lowering it means a complete graph - where no non-adjacent pair exists at all to test - needs no special case: the pair loop simply never executes, leaving the initial degree bound as the final answer, which is exactly correct (kappa(K_n) = n-1, the minimum degree of K_n).

Parameters

respectDirection

true for strong connectivity (ordered pairs, directed reachability); false for weak (unordered pairs, every edge treated as bidirectional). Only meaningful to vary on a directed graph.

Returns

kappa(G).

Note

Precondition: at least 2 enabled vertices. Callers (the GUI's Graph Connectivity action) special-case 0/1-vertex networks the same way Connectedness already does, so this is never invoked otherwise.

graphDensity()

qreal Graph::graphDensity

(

)

Gets the graph density (if computed) or computes it again.

The graph density is the ratio of present ties to total possible ties for the current relation.

IMPORTANT: edgesEnabled() semantics in SocNetV:

• If the graph is UNDIRECTED, edgesEnabled() returns E (undirected edges), even though internally each undirected edge is stored as two symmetric arcs.
• If the graph is DIRECTED, edgesEnabled() returns A (directed arcs).

Therefore:

• Undirected density: 2E / (V*(V-1))
• Directed density: A / (V*(V-1))

TODO / THINK: Self-loops (v->v)

• The denominator V*(V-1) assumes loops are not allowed/considered.
• If self-loops can exist and be enabled, decide whether to: (a) exclude loops from the numerator for density, or (b) use a loop-aware denominator (e.g., V*V for directed with loops).

Returns

qreal

graphDiameter()

int Graph::graphDiameter	(	const bool	considerWeights,
		const bool	inverseWeights )

Returns the diameter of the graph, aka the largest geodesic distance between any two vertices.

Parameters

considerWeights
inverseWeights

Returns

graphDiameterCached()

int Graph::graphDiameterCached

(

)

const

Returns the number of geodesics (shortest paths) in the graph, without recalculating it.

Returns

int

graphDichotomization()

void Graph::graphDichotomization

(

const qreal

threshold

)

Creates a new binary relation in a valued network using edge dichotomization according to the threshold parameter.

Parameters

threshold

graphDistanceEuclidean() [1/2]

qreal Graph::graphDistanceEuclidean

(

const QPointF &

a

)

the euclideian distance of QPointF a (where a is difference vector)

Parameters

a

Returns

graphDistanceEuclidean() [2/2]

qreal Graph::graphDistanceEuclidean	(	const QPointF &	a,
		const QPointF &	b )

Computes the euclideian distance between QPointF a and b.

Parameters

a
b

Returns

graphDistanceGeodesic()

int Graph::graphDistanceGeodesic	(	const int &	v1,
		const int &	v2,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = true )

Returns the geodesic distance (length of shortest path) from vertex v1 to vertex v2.

Parameters

v1
v2
considerWeights
inverseWeights

Returns

graphDistanceGeodesicAverage()

qreal Graph::graphDistanceGeodesicAverage	(	const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates )

Returns the average distance of the graph.

Parameters

considerWeights
inverseWeights
dropIsolates

Returns

graphDistanceGeodesicAverageCached()

qreal Graph::graphDistanceGeodesicAverageCached

(

)

const

Returns the average geodesic distance of the graph, without recalculating it.

Returns

qreal

graphDistancesGeodesic()

void Graph::graphDistancesGeodesic	(	const bool &	computeCentralities = false,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = true,
		const bool &	dropIsolates = false )

Computes the geodesic distances between all vertices: In the process, it also computes many other centrality/prestige metrics:

• The so-called sigma matrix, where the (i,j) element is the number of shortest paths from vertex i to vertex j, called sigma(i,j).
• The Diameter of the graph, m_graphDiameter, which is the length of the longest shortest path between every (i,j)
• The Eccentricity of every node i which is the length of the longest shortest path from i to every other node j
• The InfluenceRange and InfluenceDomain of each node.
• The centralities for every u in V (if centralities=true):
  • Betweenness: BC(u) = Sum ( sigma(i,j,u)/sigma(i,j) ) for every s,t in V Meaning: how often u sits "in between" on the shortest routes connecting other pairs of actors - a broker/gatekeeper measure. High betweenness means removing u would disrupt many people's shortest path to each other. When to use: finding brokers, bottlenecks, or single points of failure in a communication/supply network - who, if removed, would fragment the network's shortest routes the most. Weights: shortest-path-based - if a weight represents value/strength (interaction frequency, trust), invert it (a strong tie should behave like a short/cheap path) so routing favors the strongest ties, not the weakest. Compare to: Stress (SC) below counts the same "sits on a shortest path" event without dividing by how many alternative shortest paths existed - use BC when you want "share of control" over each pair's routing, SC when you want raw path traffic.
  • Stress: SC(u) = Sum ( sigma(i,j) ) for every s,t in V Meaning: like betweenness, but simply counts how many shortest paths pass through u, without dividing by how many alternative shortest paths existed for that pair - so it also rewards actors on many paths even when those paths weren't a pair's only shortest route. When to use: estimating raw path/traffic load through a node (e.g. network routing, load-bearing infrastructure) rather than its exclusive control over routing. Weights: same shortest-path reasoning as BC above - invert a strength-type weight so strong ties route like short paths.
  • Eccentricity: EC(u) = 1/maxDistance(u,t) for some t in V Meaning: a worst-case reachability measure - how far away is u's single most distant counterpart? High eccentricity centrality means even u's "hardest to reach" other actor is nearby. When to use: worst-case reasoning - e.g. picking a broadcast/facility location that minimizes the longest anyone has to wait to be reached, rather than the average case. Weights: same shortest-path reasoning as BC/SC above - invert a strength-type weight. Compare to: Closeness (CC) below is this same distance-based idea using the average distance instead of the worst case.
  • Closeness: CC(u) = 1 / Sum( d(u,t) ) for every t in V Meaning: how close u is, on average, to everyone else - a low total distance to others gives a high closeness score. Only meaningful on a fully connected graph, since an unreachable actor has undefined distance. When to use: identifying actors who can spread something (information, disease, an influence campaign) to the whole network fastest, on a graph known to be connected. Weights: same shortest-path reasoning as BC/SC/EC above - invert a strength-type weight so a strong tie behaves like a short/cheap path. Compare to: Influence Range Closeness Centrality (IRCC, see centralityClosenessIR()) is this same idea adapted to work on disconnected graphs too.
  • Power (Gil-Schmidt): PC(s) = [1/(N-1)] * Sum_i( nthOrder[i] / i ), where nthOrder[i] is the number of nodes at distance i from s (computed in DistanceEngine::compute()). Meaning: a generalized degree measure that gives (shrinking) credit for nodes several steps away too, not just direct neighbors - similar in spirit to eigenvector centrality, but computed directly from how many nodes sit at each distance rather than via eigen-decomposition. When to use: a cheaper, degree-based alternative to Eigenvector Centrality (EVC) for rewarding both direct and indirect reach, when a full eigen-decomposition isn't needed. Weights: same shortest-path reasoning as BC/SC/EC/CC above (nthOrder is computed from DistanceEngine's shortest-path distances) - invert a strength-type weight. Compare to: Eigenvector Centrality (EVC, see centralityEigenvector()) captures a related "reach plus indirect reach" idea via eigen-decomposition instead. Not to be confused with Bonacich's differently-named, unrelated "Power Centrality" measure (BPC, see centralityBonacich()).

    Parameters

    centralities
    considerWeights
    inverseWeights
    dropIsolates

graphFileLoaded

void Graph::graphFileLoaded ( const int & fileType, const QString & fileName = QString(), const QString & netName = QString(), const int & totalNodes = 0, const int & totalLinks = 0, const int & edgeDirType = 0, const qint64 & elapsedTime = 0, const QString & message = QString() )

slot

Stores loaded file name, graph name, sets edge direction type and signals MW to update the UI.

Called from Parser when file parsing ends.

Parameters

fileType
fileName
netName
totalNodes
totalLinks
edgeDirType
elapsedTime
message

graphGeodesicDistanceDistribution()

QMap< int, int > Graph::graphGeodesicDistanceDistribution	(	const bool &	considerWeights = false,
		const bool &	inverseWeights = false )

Returns a histogram of geodesic distances across all ordered vertex pairs.

Ensures the full APSP result is available (uses cache if graph is unchanged), then iterates all enabled vertex pairs and groups them by their geodesic distance into a QMap<int, int> where key = distance and value = number of ordered pairs at that distance. Unreachable pairs (distance == RAND_MAX) are excluded.

This method is called ONLY on explicit user request — it is never triggered as a side-effect of centrality computation or other internal analysis paths.

Parameters

considerWeights	Pass through to graphDistancesGeodesic; selects BFS vs Dijkstra.
inverseWeights	Pass through to graphDistancesGeodesic; inverts edge weights.

Returns

QMap<int, int> Sorted map: distance bucket → count of ordered pairs.

graphGeodesicsCountCached()

qreal Graph::graphGeodesicsCountCached

(

)

const

Returns the number of geodesics (shortest paths) accumulated by DistanceEngine, without recalculating anything.

graphGeodesicShortestPath()

QList< int > Graph::graphGeodesicShortestPath	(	const int &	v1,
		const int &	v2,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false )

Reconstructs one shortest path from vertex v1 to vertex v2.

Runs a single-source BFS (unweighted) or Dijkstra (weighted) from v1, keeping a predecessor array, then traces the path back from v2. This is an on-demand, interactive call — it does NOT touch the APSP cache and is safe to call at any time regardless of calculatedDistances.

Parameters

v1	Source vertex number.
v2	Target vertex number.
considerWeights	If true, uses Dijkstra with edge weights; otherwise BFS.
inverseWeights	If true, uses 1/weight as the edge cost (for closeness-style paths).

Returns

QList<int> Ordered list of vertex numbers on the path, inclusive of v1 and v2. Returns an empty list when v1 == v2 or no path exists.

graphHasEdgeCustomAttributes()

QStringList Graph::graphHasEdgeCustomAttributes

(

)

const

Returns a list of all unique custom attribute keys present across all enabled edges in the current graph.

graphHasVertexCustomAttributes()

QStringList Graph::graphHasVertexCustomAttributes

(

)

const

Returns true if at least one vertex has a 'custom' attribute.

Returns

bool

graphHasVertexCustomIcons()

bool Graph::graphHasVertexCustomIcons

(

)

const

Returns true if at least one vertex has a 'custom' shape (therefore a custom icon).

Returns

bool

graphLoadedTerminateParserThreads

void Graph::graphLoadedTerminateParserThreads ( QString reason )

slot

Graph::graphLoadedTerminateParserThreads.

Parameters

reason

graphMatrixDistanceGeodesicCreate()

bool Graph::graphMatrixDistanceGeodesicCreate	(	const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Creates the matrix DM of geodesic distances between vertices.

Phase 1: calls graphDistancesGeodesic() which runs the DistanceEngine.

Phase 2: fills the DM matrix from the cached per-vertex distances - an O(N²) memory-write pass.

Parameters

considerWeights	If true, edge weights are used in distance computations.
inverseWeights	If true, edge weights are inverted before use.
dropIsolates	If true, isolate nodes are excluded from the analysis.

Returns

true on success, false if the computation was cancelled.

graphMatrixShortestPathsCreate()

void Graph::graphMatrixShortestPathsCreate	(	const bool &	considerWeights = false,
		const bool &	inverseWeights = true,
		const bool &	dropIsolates = false )

Creates the matrix SIGMA of shortest paths (geodesics) between vertices Each SIGMA(i,j) is the number of shortest paths (geodesics) from i and j.

Parameters

considerWeights
inverseWeights
dropIsolates

graphMatrixStrToType()

int Graph::graphMatrixStrToType

(

const QString &

matrix

)

const

Helper method, return the matrix type of human readable matrix name .

Parameters

matrix

Returns

graphMatrixTypeToString()

QString Graph::graphMatrixTypeToString

(

const int &

matrixType

)

const

Helper method, return the human readable name of matrix type.

Parameters

matrix

graphMetricStrToType()

int Graph::graphMetricStrToType

(

const QString &

metricStr

)

const

Helper method, return the identifier of a metric.

Parameters

metricStr

graphMetricTypeToString()

QString Graph::graphMetricTypeToString

(

const int &

metricType

)

const

Helper method, return the human readable name of metric type.

Parameters

metric

graphNodeConnectivity()

Graph::NodeConnectivityResult Graph::graphNodeConnectivity	(	int	source,
		int	target,
		bool	respectDirection )

Local vertex connectivity kappa(source, target): the minimum number of nodes, other than source and target themselves, whose removal disconnects target from source.

Computed via Menger's theorem / vertex-split max-flow - see the detailed algorithm comment on localVertexConnectivityFlow() above.

source and target being directly adjacent is reported as NodeConnectivityStatus::Adjacent, not as a number: Menger's theorem requires non-adjacency (an edge is a "path" no vertex removal can ever break), so there is no finite, meaningful cut size to report - reporting some sentinel number here would silently misrepresent "cannot be separated by removing other nodes" as an ordinary connectivity value.

Parameters

source
target
respectDirection	true for the "strong" reading (directed reachability, source must reach target via directed paths); false for "weak" (every edge treated as bidirectional). Only meaningful to vary on a directed graph - see the GUI's Connectedness/Node Connectivity prompts, which only ask when isDirected() is true.

Returns

NodeConnectivityResult - see graph.h. Ok.value == 0 is a normal, valid answer (target is simply unreachable from source under the chosen direction mode).

graphReachable()

bool Graph::graphReachable	(	const int &	v1,
		const int &	v2 )

Returns true if vertices v1 and v2 are reachable.

Parameters

v1
v2

Returns

bool

graphReciprocity()

qreal Graph::graphReciprocity

(

)

Gets the arc reciprocity of the graph.

Also computes the dyad reciprocity and fills parameters with values.

Returns

qreal

graphStronglyConnectedComponents()

int Graph::graphStronglyConnectedComponents

(

)

Counts strongly connected components using Tarjan's algorithm.

Strong connectivity respects edge direction: two nodes are in the same Strongly Connected Component (SCC) only if each is reachable from the other via directed edges. This is a strictly finer partition than weak connectivity (graphWeaklyConnectedComponents(), which treats every edge as undirected) - a graph can be a single weak component while having many strong components, e.g. a directed path a->b->c is one weak component but three strong components, since c cannot reach a. For an undirected graph the two notions coincide (every edge is effectively reciprocal), so this still returns the right answer, just via a slightly more roundabout route than graphWeaklyConnectedComponents().

Algorithm (Tarjan 1972): a single Depth First Search (DFS) assigns each vertex a discovery "index" (the order it was first visited) and a "lowlink" - the smallest index reachable from that vertex by following zero or more tree edges and then at most one edge back into an ancestor still on the DFS stack. Vertices are pushed onto an explicit stack as they're discovered and popped once a SCC is found. A vertex is the root of an SCC exactly when its lowlink equals its own index - meaning nothing below it on the DFS stack can reach back above it - and popping the stack down to and including that root yields exactly the members of one SCC. This is O(V+E), a single DFS pass, with no graph transpose needed - unlike Kosaraju's algorithm, which gets the same complexity but needs two passes over a transposed graph. Tarjan's is the better fit here since SocNetV already has cheap out-edge iteration and no existing transpose-graph structure to reuse.

Implemented as an explicit-stack simulation of the textbook recursive DFS, not recursion itself: SocNetV networks can have a directed path of a few thousand nodes (a plausible worst case for a citation or hierarchy network), which would recurse that deep and risk overflowing the real call stack. Each simulated stack frame remembers which out-edge it was in the middle of examining (Frame::cursor), which is exactly the resumption point a real call stack would give for free after a simulated recursive call "returns" (its frame is popped).

m_vertexComponentId (weak-component IDs, used by the colorize-by-component layout action) is NOT touched here; strong components are not currently exposed per-vertex, only as a count.

Returns

Number of strongly connected components.

graphStronglyConnectedComponentsCached()

int Graph::graphStronglyConnectedComponentsCached

(

)

const

graphSumDistanceCached()

qreal Graph::graphSumDistanceCached

(

)

const

Returns the sum of all finite geodesic distances accumulated by DistanceEngine, without recalculating anything.

graphTriadCensus()

bool Graph::graphTriadCensus

(

)

Conducts a triad census and updates QList::triadTypeFreqs, which is the list carrying all triad type frequencies Complexity: O(n³) — three nested loops each bounded by N.

Returns

graphTriadTypeFreqs()

const QList< int > & Graph::graphTriadTypeFreqs ( ) const

inline

graphWalksMatrixCreate()

void Graph::graphWalksMatrixCreate	(	const int &	N = 0,
		const int &	length = 0,
		const bool &	dropIsolates = false,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	symmetrize = false )

Computes either the "Walks of given length" or the "Total Walks" matrix. If length>0, it computes the Walks of given length matrix, XM=AM^l where each element (i,j) denotes the number of walks of length l between vertex i and j. If length=0, it computes the Total Walks matrix, XSM=Sum{AM^n} where each (i,j) denotes the total number of walks of any length between vertices i and j. NOTE: In the latter case, this function is VERY SLOW on large networks (n>50), since it will calculate all powers of the sociomatrix up to n-1 in order to find out all possible walks.

Parameters

N	- dimension of the sociomatrix (number of vertices). Default is 0, in which case it will be calculated as the number of vertices in the graph.
length	- the length of walks to be calculated. Default is 0, in which case all walks of any length will be calculated.

graphWeaklyConnectedComponents()

int Graph::graphWeaklyConnectedComponents

(

)

Counts weakly connected components using BFS.

Weak connectivity treats every edge as undirected regardless of graph type:

• Undirected: equivalent to ordinary connected components.
• Directed: two nodes are in the same component when there is an undirected path between them (ignoring arrow direction). Answers the practical "how many disconnected islands?" question consistently for both types. Does NOT imply strong connectivity (all-pairs directed reachability); use isConnected() / graphDistancesGeodesic() for that.

m_graphIsConnected is NOT touched here; that flag belongs to the SSSP engine. Component IDs (1-based) are cached in m_vertexComponentId keyed by vertex number, ready for the colorize-by-component layout action.

Returns

Number of weakly connected components.

graphWeaklyConnectedComponentsCached()

int Graph::graphWeaklyConnectedComponentsCached

(

)

const

hasCalculatedTriadCensus()

bool Graph::hasCalculatedTriadCensus ( ) const

inline

htmlEscaped

QString Graph::htmlEscaped ( QString str ) const

slot

Helper method, returns a nice qstring where all html special chars are encoded.

Parameters

str

Returns

incGeodesicsCount()

void Graph::incGeodesicsCount

(

)

isAnyRelationWeighted()

bool Graph::isAnyRelationWeighted

(

)

Returns true if any relation in the graph has at least one edge with weight other than 0 or 1.

Unlike isWeighted(), which only scans the current relation, this method iterates every relation. It is used to report overall graph weightedness independently of which relation is currently active.

isCentralityIndexComputed()

bool Graph::isCentralityIndexComputed

(

const IndexType

index

)

const

Returns true if the given centrality/prestige index has been computed.

Uses the per-index calculated* flags set by each analysis method. Matches the IndexType enum defined in global.h.

isConnected()

bool Graph::isConnected

(

)

Checks if the graph is connected, in the sense of a topological space, i.e., there is a path from any vertex to any other vertex in the graph.

Returns

bool

isConnectedCached()

bool Graph::isConnectedCached

(

)

const

Returns true if the graph is connected, without recalculating it.

Returns

bool

isDirected()

bool Graph::isDirected

(

)

Returns true if graph is directed.

Returns

bool

isEmpty()

bool Graph::isEmpty

(

)

const

Returns true if the current graph has no vertices at all.

isFileFormatExportSupported()

bool Graph::isFileFormatExportSupported

(

const int &

fileFormat

)

const

Returns true if the fileFormat is supported for saving.

Parameters

fileFormat

Returns

isLoaded()

bool Graph::isLoaded

(

)

const

Returns true if a graph has been loaded from a file.

Returns

isModified()

bool Graph::isModified

(

)

const

Returns true of graph is modified (edges/vertices added/removed).

Returns

isSaved()

bool Graph::isSaved

(

)

const

Returns true if the graph is saved.

Returns

isSymmetric()

bool Graph::isSymmetric

(

)

Returns TRUE if the adjacency matrix of the current relation is symmetric.

Returns

bool

isUndirected()

bool Graph::isUndirected

(

)

Returns true if graph is undirected.

Returns

bool

isWeighted()

bool Graph::isWeighted

(

)

Returns true if the current relation has at least one edge with weight other than 0 or 1 (i.e. the relation is valued/weighted).

The result is cached via calculatedGraphWeighted and invalidated on every relation switch (

See also

relationSet()). Consequently, for multi-relation graphs this method only reflects the active relation — switching to a non-weighted relation after a weighted one will return false.

When you need to know whether any relation is weighted, use

See also

isAnyRelationWeighted() instead.

Complexity: O(n²) on cache miss, O(1) on hit.

layoutByProminenceIndex()

void Graph::layoutByProminenceIndex	(	int	prominenceIndex,
		int	layoutType,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Applies a layout according to each actor's prominence index score. The layout type can be radial (0), level (1), node sizes (2) or node colors (3), as follows: layoutType=0 - Repositions all nodes on the periphery of concentric circles with radius analogous to their prominence index layoutType=1 - Repositions all nodes on different top-down levels according to their centrality layoutType=2 - Changes node sizes to be proportional to their prominence index score layoutType=2 - Changes node colors to reflect their prominence index score (from red to green).

Parameters

prominenceIndex,0-12
layoutType	0,1,2,3
considerWeights
inverseWeights
dropIsolates

layoutCircular()

void Graph::layoutCircular	(	const double &	x0,
		const double &	y0,
		const double &	newRadius,
		const bool &	guides = false )

Repositions all nodes on the periphery of a circle with given radius.

Parameters

x0
y0
maxRadius

layoutEgoRadial()

void Graph::layoutEgoRadial

(

const int

egoVertex

)

Ego-centered radial layout.

Places egoVertex at the canvas center, its 1-hop out-neighbors (in the current relation) on an inner ring, and all remaining enabled nodes on an outer ring. Only node positions are changed — graph structure is untouched.

Ring radii are fixed fractions of canvasMaxRadius(): ring1 = 0.35 × maxRadius (ego neighbors) ring2 = 0.75 × maxRadius (all other nodes)

Edge case: if egoVertex does not exist, emits a status message and returns.

Parameters

egoVertex

The vertex number to place at the center.

layoutForceDirected_Eades_moveNodes()

qreal Graph::layoutForceDirected_Eades_moveNodes

(

const qreal &

c4

)

Moves all vertices to their new positions as computed by the Eades Spring Embedder model.

Called once per iteration from layoutForceDirectedSpringEmbedder(). Applies the accumulated displacement vectors to each vertex position, scaled by the normalization factor c4, and clamps the result to the visible canvas area.

Displacement clamping: each component (x, y) is capped at ±(canvasMinDimension * 0.05) before being applied. This prevents the "iteration-2 explosion" caused by near-coincident node pairs at the start of the simulation — without the cap, f_rep = c_rep/dist² diverges for dist→0 and a single node pair can produce a 100px+ spike that locks the layout into a bad configuration for all remaining iterations.

Node positions are updated in-memory only (setX/setY). The caller is responsible for emitting setNodePos signals in a single bulk pass after the iteration loop completes — this avoids N×maxIterations signal emissions to the graphics scene.

Parameters

c4	Displacement normalization factor for this iteration. Supplied by the caller's linear cooling schedule (decreases from c4_init to c4_min over maxIterations) to progressively damp movement and allow the layout to settle.

Returns

The largest Euclidean displacement applied to any single vertex this iteration (canvas pixels). Used by the caller for convergence detection.

layoutForceDirected_F_att()

qreal Graph::layoutForceDirected_F_att	(	const QString	model,
		const qreal &	dist,
		const qreal &	optimalDistance )

Computes the attractive (spring) force between two adjacent vertices.

For the Eades model: logarithmic spring force scaled by c_spring = 2. Note: log10(dist / optimalDistance) is negative when dist < optimalDistance, meaning the spring pushes nodes apart when they are closer than the natural length — this is intentional per Eades (1984), who uses a log spring that both attracts far-apart neighbours and repels too-close ones.

For the FR model: quadratic attraction as per Fruchterman & Reingold (1991).

Parameters

model	Layout model identifier: "Eades" or "FR".
dist	Euclidean distance between the two adjacent vertices (canvas pixels).
optimalDistance	The natural/optimal inter-vertex spacing for the current graph.

Returns

Attractive force magnitude (positive = pull toward each other).

layoutForceDirected_F_rep()

qreal Graph::layoutForceDirected_F_rep	(	const QString	model,
		const qreal &	dist,
		const qreal &	optimalDistance )

Computes the repulsive force between two vertices for force-directed layouts.

For the Eades model: uses an inverse-square law with constant c_rep.

• The previous cutoff (f_rep = 0 when dist > 2 * optimalDistance) has been removed. That cutoff was the primary cause of node clustering: distant nodes received zero repulsion and were never pushed apart. All vertex pairs now repel each other, regardless of distance.
• c_rep = c3 = 1.0 per Eades (1984). The inlined Eades path in layoutForceDirectedSpringEmbedder() uses the same value directly.

For the FR model: uses the grid-variant with a 2*optimalDistance cutoff radius, as described by Fruchterman & Reingold (1991).

Parameters

model	Layout model identifier: "Eades" or "FR".
dist	Euclidean distance between the two vertices (canvas pixels).
optimalDistance	The natural/optimal inter-vertex spacing for the current graph.

Returns

Repulsive force magnitude (negative, i.e. pushing apart).

layoutForceDirected_FR_moveNodes()

qreal Graph::layoutForceDirected_FR_moveNodes

(

const qreal &

temperature

)

Moves all vertices to their new positions as computed by the Fruchterman-Reingold model.

Called once per iteration from layoutForceDirectedFruchtermanReingold(). Limits each vertex's displacement to the current temperature value, which decreases with each iteration to act as a simulated annealing cooling schedule, then clamps the result to the visible canvas area.

Node positions are updated in-memory only (setX/setY). The caller is responsible for emitting setNodePos signals in a single bulk pass after the iteration loop completes — this avoids N×maxIterations signal emissions to the graphics scene.

Parameters

temperature

Current annealing temperature, controlling the maximum displacement per iteration. Computed by layoutForceDirected_FR_temperature().

layoutForceDirected_FR_temperature()

qreal Graph::layoutForceDirected_FR_temperature

(

const int

iteration

)

const

Reduces the temperature as the layout approaches a better configuration.

Returns

qreal temperature

layoutForceDirectedFruchtermanReingold()

void Graph::layoutForceDirectedFruchtermanReingold

(

const int

maxIterations

)

Embeds a Force Directed Placement layout according to the Fruchterman-Reingold model. Fruchterman and Reingold (1991) refined the Spring Embedder model by replacing the forces. In this model, "the vertices behave as atomic particles or celestial bodies, exerting attractive and repulsive forces on one another." (ibid). Again, only vertices that are neighbours attract each other but, unlike Spring Embedder, all vertices repel each other. These forces induce movement. The algorithm might resemble molecular or planetary simulations, sometimes called n-body problems.

Parameters

maxIterations

layoutForceDirectedKamadaKawai()

void Graph::layoutForceDirectedKamadaKawai	(	const int	maxIterations = 500,
		const bool	considerWeights = false,
		const bool	inverseWeights = false,
		const bool	dropIsolates = false,
		const QString &	initialPositions = "current" )

Embeds a Force Directed Placement layout according to the Kamada-Kawai model. In this model, the network is considered to be a dynamic system where every two actors are 'particles' mutually connected by a 'spring'. Each spring has a desirable length, which corresponds to their graph theoretic distance. In this way, the optimal layout of the graph is the state with the minimum imbalance. The degree of imbalance is formulated as the total spring energy: the square summation of the differences between desirable distances and real ones for all pairs of particles Initially, the particles/actors are placed on the vertices of a regular n-polygon.

layoutForceDirectedSpringEmbedder()

void Graph::layoutForceDirectedSpringEmbedder

(

const int

maxIterations

)

Embeds a Force Directed Placement layout according to the initial Spring Embedder model proposed by Eades.

Parameters

maxIterations

Maximum number of iterations to run. The loop may exit earlier if the layout converges (max displacement falls below epsilon).

The Spring Embedder model (Eades, 1984), part of the Force Directed Placement (FDP) family, assigns forces to all vertices and edges: non-adjacent node pairs repel each other via an inverse-square law (analogous to Coulomb's law), while adjacent pairs attract via a logarithmic spring — deliberately not Hooke's law, which Eades argued causes clumping at large distances. These forces are applied to the nodes iteratively, pulling them closer together or pushing them further apart, until the system comes to an equilibrium state (node positions do not change anymore).

Implementation notes:

• Adjacency is pre-cached into a QSet<QPair<int,int>> before the iteration loop to avoid O(log N) edgeExists() calls inside the O(N²) inner loop.
• Repulsion and attraction math is inlined directly (no QString model dispatch per call).
• Vertex numbers are cached to local ints before the inner loop.
• A linear cooling schedule decreases c4 from c4_init to c4_min over maxIterations, damping displacement progressively so the layout settles rather than oscillating.
• Early exit when max displacement in an iteration falls below epsilon (convergence).

layoutRadialRandom()

void Graph::layoutRadialRandom

(

const bool &

guides = true

)

Repositions all nodes on the periphery of different circles with random radius.

Parameters

x0
y0
maxRadius

layoutRandom()

void Graph::layoutRandom

(

)

Repositions all nodes on random positions Emits setNodePos(i, x,y) to tell GW that the node item should be moved.

Parameters

maxWidth
maxHeight

layoutRandomInMemory()

void Graph::layoutRandomInMemory ( )

private

Repositions all vertices at random coordinates without emitting any signals.

Used internally by force-directed layout algorithms (Eades, FR) to set up initial particle positions before the iteration loop begins. Since the iteration loop will immediately overwrite these positions and emits a bulk setNodePos pass at the end, there is no need to signal the graphics scene here.

For the standalone random layout action (with progress dialog and scene update), use layoutRandom() instead.

layoutVertexSizeByIndegree()

void Graph::layoutVertexSizeByIndegree

(

)

Convenience method Changes the size of all nodes to be proportional to their InDegree (Degree Prestige) Calls layoutByProminenceIndex.

layoutVertexSizeByOutdegree()

void Graph::layoutVertexSizeByOutdegree

(

)

Convenience method Changes the size of all nodes to be proportional to their outDegree (Degree Centrality) Calls layoutByProminenceIndex.

loadFile()

void Graph::loadFile	(	const QString	fileName,
		const QString	codecName,
		const int	fileFormat,
		const QString	delimiter = QString(),
		const int	sm_two_mode = 1,
		const bool	sm_has_labels = false )

Loads a graph from a given file.

It creates a new Parser object, moves it to a another thread, connects signals and slots and calls its run() method.

Parameters

fileName
codecName
m_showLabels
maxWidth
maxHeight
fileFormat
sm_two_mode

Returns

matrixAdjacency()

Matrix & Graph::matrixAdjacency ( )

inline

matrixAdjacencyInverse()

Matrix & Graph::matrixAdjacencyInverse ( )

inline

matrixCliqueCoMembership()

Matrix & Graph::matrixCliqueCoMembership ( )

inline

matrixDistances()

Matrix & Graph::matrixDistances ( )

inline

matrixReachability()

Matrix & Graph::matrixReachability ( )

inline

matrixTotalWalks()

Matrix & Graph::matrixTotalWalks ( )

inline

matrixWalks()

Matrix & Graph::matrixWalks ( )

inline

minmax()

void Graph::minmax ( qreal C, GraphVertex * v, qreal & max, qreal & min, int & maxNode, int & minNode )

private

Computes minimum and maximum centralities during graphDistancesGeodesic().

Parameters

C
v
max
min
maxNode
minNode

moveToThreadFacade()

void Graph::moveToThreadFacade

(

QThread *

thread

)

notConnectedPairsClear()

void Graph::notConnectedPairsClear

(

)

notConnectedPairsInsert()

void Graph::notConnectedPairsInsert	(	int	from,
		int	to )

notConnectedPairsSize()

int Graph::notConnectedPairsSize

(

)

const

notifyEdgesVisibilityBatch()

void Graph::notifyEdgesVisibilityBatch

(

const QList< EdgeVisibilityChange > &

changes

)

Plain relay to signalSetEdgesVisibilityBatch, for GraphVertex to notify the UI layer of many edges' visibility changes as a single queued dispatch. See #WS3 M2.

notifyEdgeVisibilityChanged()

void Graph::notifyEdgeVisibilityChanged	(	const int &	relation,
		const int &	source,
		const int &	target,
		const bool &	toggle,
		const bool &	preserveReverseEdge = false,
		const int &	edgeWeight = 1,
		const int &	reverseEdgeWeight = 1 )

Plain relay to signalSetEdgeVisibility, for GraphVertex (a QtCore-only value class, not a QObject) to notify the UI layer of a single edge's visibility change. See #WS3 M2.

numberOfTriples()

qreal Graph::numberOfTriples

(

int

v1

)

Returns the number of triples of vertex v1 A triple Υ at a vertex v is a path of length two for which v is the center vertex.

prestigeDegree()

void Graph::prestigeDegree	(	const bool &	considerWeights,
		const bool &	dropIsolates = false )

Computes the Degree Prestige (in-degree) of each vertex - diagonal included Also the mean value and the variance of the in-degrees.

Meaning: prestige measures flip centrality's question around - instead of "how many ties does this actor have," they ask "how many ties point *at* this actor." Degree prestige is the simplest version: how many others chose to connect to this actor, ignoring how many connections the actor made outward. Meaningless on an undirected graph, where in- and out-ties are the same thing.

When to use: directed networks where "being chosen by others" is the thing worth measuring

• citation counts, follower counts, nomination/endorsement data - and a quick raw count is enough (see PageRank Prestige below when the quality of who's choosing you also matters).

Compare to: Degree Centrality (DC, see centralityDegree()) is this same raw-count idea for outbound ties (or all ties, on an undirected graph).

Weights: no inversion choice here (considerWeights only) - when considered, weights are summed directly, so a stronger inbound tie always adds more.

Math: DP(i) = number of inbound edges to i (or their summed weights, if weights are considered). Standardized SDP(i) = DP(i) / (N-1).

Parameters

weights
dropIsolates

prestigePageRank()

void Graph::prestigePageRank

(

const bool &

dropIsolates = false

)

Calculates the PageRank Prestige of each vertex.

Meaning: the same idea Google originally used to rank web pages - an actor is prestigious if prestigious actors point to it. Unlike plain degree prestige (which treats every inbound tie equally), a link from someone with few outbound ties and high prestige counts for much more than one of a hundred outbound ties from someone unremarkable, and that prestige keeps circulating until every score stabilizes.

When to use: directed networks where endorsement quality matters, not just quantity - web links, citation networks, recommendation/referral graphs - anywhere "an endorsement from someone important should count for more."

Compare to: Degree Prestige (DP) counts inbound ties equally; PageRank weighs each one by the endorser's own prestige divided by their out-degree. Eigenvector Centrality (EVC, see centralityEigenvector()) is the closest centrality-side analogue - both are "importance feeds back on itself" measures - but EVC is built for undirected/symmetric graphs while PageRank is built for directed graphs with an explicit damping factor.

Weights: no considerWeights/inverseWeights choice at all - every inbound link always counts as weight 1, normalized by the endorser's out-degree (see Math below), regardless of any edge weight set on the graph.

Math: iteratively, PRP(i) = (1-d)/N + d * Sum_j( PRP(j) / outLinks(j) ) for every j linking to i, where d is the damping factor (0.85, matching Google's original choice) and N is the number of actors. Repeated until scores stop changing by more than a small delta.

Parameters

dropIsolates

prestigeProximity()

void Graph::prestigeProximity	(	const bool	considerWeights = false,
		const bool	inverseWeights = false,
		const bool	dropIsolates = false )

Computes Proximity Prestige of each vertex Also the mean value and the variance of it..

Meaning: how close everyone else who can reach this actor actually is, on average - prestige's answer to closeness centrality. It rewards being easy to reach and being reachable by a large share of the network, so an actor reachable by only a couple of very close others still scores lower than one reachable, at moderate distance, by almost everyone.

When to use: directed networks where you want a refined "how sought-after is this actor" score that accounts for distance, not just a raw inbound-tie count (see Degree Prestige for the simpler version).

Compare to: Influence Range Closeness Centrality (IRCC, see centralityClosenessIR()) is this same idea from the centrality side (distance to others) rather than prestige's (distance from others), and likewise works on disconnected graphs.

Weights: shortest-path-based, same as CC/IRCC - if a weight represents value/strength, invert it so a strong tie behaves like a short/cheap path.

Math: for actor i, let I_i be the set of actors that can reach i (its influence domain). PP(i) = [ |I_i| / (V-1) ] / [ (sum of d(j,i) for j in I_i) / |I_i| ] - the fraction of the network that can reach i, divided by their average distance to i.

progressCanceled()

bool Graph::progressCanceled

(

)

const

Returns true if the user has requested cancellation via the progress dialog.

progressStatus()

void Graph::progressStatus ( const QString & msg )

protected

Emits a status message to be shown in the UI status bar.

Parameters

msg	The message to be shown.

prominenceDistribution()

void Graph::prominenceDistribution	(	const int &	index,
		const ChartType &	type,
		const QString &	distImageFileName = QString() )

Computes the distribution of a centrality index score. The distribution is stored as Qt Series depending on the SeriesType parameter type It is send to MW through signal/slot.

Parameters

index
type

prominenceDistributionArea()

void Graph::prominenceDistributionArea	(	const H_StrToInt &	discreteClasses,
		const QString &	name,
		const QString &	distImageFileName )

Computes prominence distribution data and delegates Area chart rendering.

Performs the algorithmic portion only:

• Orders (value, frequency) pairs derived from discreteClasses
• Computes min/max value and min/max frequency

UI construction (Qt Charts series/axes), optional PNG export, and emission of signalPromininenceDistributionChartUpdate(...) are delegated to the UI façade implementation in graph_ui_prominence_distribution.cpp (WS2/F4).

Behavior and output semantics are preserved.

Parameters

discreteClasses	Map of value (string) -> frequency.
name	Display name of the series.
distImageFileName	If non-empty, export chart to this PNG file.

prominenceDistributionBars()

void Graph::prominenceDistributionBars	(	const H_StrToInt &	discreteClasses,
		const QString &	name,
		const QString &	distImageFileName )

Computes the prominence distribution and delegates Bar chart rendering.

This method performs only the algorithmic portion: it derives ordered category labels (centrality values formatted with 6 decimals), the matching frequencies, and computes basic range statistics:

• min/max value (numeric)
• min/max frequency

UI construction (Qt Charts objects, axes, optional PNG export) and emission of the update signal to MainWindow are delegated to the UI façade layer (graph_ui_prominence_distribution.cpp), following WS2/F4 rules.

Behavior, rendering semantics, and export output remain unchanged.

Parameters

discreteClasses	A map of centrality value (as string) to frequency.
name	The display name of the distribution series.
distImageFileName	If non-empty, the chart is exported to this PNG file.

prominenceDistributionSpline()

void Graph::prominenceDistributionSpline	(	const H_StrToInt &	discreteClasses,
		const QString &	seriesName,
		const QString &	distImageFileName )

Computes prominence distribution data and delegates Spline chart rendering.

Performs the algorithmic portion only:

• Orders (value, frequency) pairs derived from discreteClasses
• Computes min/max value and min/max frequency

UI construction (Qt Charts series/axes), optional PNG export, and emission of signalPromininenceDistributionChartUpdate(...) are delegated to the UI façade implementation in graph_ui_prominence_distribution.cpp (WS2/F4).

Behavior and output semantics are preserved.

Parameters

discreteClasses	Map of value (string) -> frequency.
seriesName	Display name of the series.
distImageFileName	If non-empty, export chart to this PNG file.

randomizeThings()

void Graph::randomizeThings

(

)

Adds a little universal randomness :).

RANDOM NETWORKS

randomNetErdosCreate()

bool Graph::randomNetErdosCreate	(	const int &	N,
		const QString &	model,
		const int &	m,
		const qreal &	p,
		const QString &	mode,
		const bool &	diag )

Creates an Erdős–Rényi random network.

Supports two models:

• G(n,p): each possible edge is included independently with probability p.
• G(n,M): exactly M edges are placed at random among all possible pairs.

Parameters

N	Number of nodes.
model	"G(n,p)" or "G(n,M)".
m	Number of edges (used in G(n,M) model; 0 for G(n,p)).
p	Edge probability (used in G(n,p) model; 0 for G(n,M)).
mode	"graph" for undirected, anything else for directed.
diag	If true, self-loops are allowed.

Returns

true on success, false if the user cancelled.

randomNetLatticeCreate()

bool Graph::randomNetLatticeCreate	(	const int &	N,
		const int &	length,
		const int &	dimension,
		const int &	neighborhoodLength,
		const QString &	mode,
		const bool &	circular )

Creates a lattice (mesh/grid) random network.

Nodes are arranged in a length×length grid. Each node connects to its neighbours within a given neighbourhood distance in both horizontal and vertical directions. The resulting network forms a regular tiling pattern.

Note: the dimension and circular parameters are reserved for future extension to higher-dimensional and toroidal lattices.

Parameters

N	Total number of nodes (should equal length^2).
length	Number of nodes along each dimension of the grid.
dimension	Reserved for future use (higher-dimensional lattices).
neighborhoodLength	Distance within which nodes are connected.
mode	"graph" for undirected, anything else for directed.
circular	Reserved for future use (toroidal/circular lattice).

Returns

true on success, false if the user cancelled.

randomNetRegularCreate()

bool Graph::randomNetRegularCreate	(	const int &	N,
		const int &	degree,
		const QString &	mode,
		const bool &	diag )

Creates a pseudo-random d-regular network.

Every node ends up with exactly degree neighbours. The algorithm:

1. Build an initial ordered edge list where each node i connects to the next degree/2 nodes (undirected) or degree nodes (directed).
2. Repeatedly pick two edges at random and swap their endpoints, ensuring no self-loops or duplicate edges result, until all edges have been processed. This produces a random regular graph.
3. Draw the resulting edges.

Parameters

N	Number of nodes.
degree	Desired degree of every node (must be even for undirected).
mode	"graph" for undirected, anything else for directed.
diag	Reserved, currently unused.

Returns

true on success, false if the user cancelled.

randomNetRingLatticeCreate()

bool Graph::randomNetRingLatticeCreate	(	const int &	N,
		const int &	degree,
		const bool	updateProgress = false )

Creates a random ring lattice network.

Parameters

vert
degree
x0
y0
radius
updateProgress

randomNetScaleFreeCreate()

bool Graph::randomNetScaleFreeCreate	(	const int &	N,
		const int &	power,
		const int &	m0,
		const int &	m,
		const qreal &	alpha,
		const QString &	mode )

Creates a Barabási–Albert scale-free random network.

The algorithm works in two phases:

1. Seed: build a fully connected clique of m0 initial nodes.
2. Growth: add nodes one by one up to N, each connecting to m existing nodes via preferential attachment — nodes with higher degree are more likely to receive new edges (rich-get-richer effect).

The attachment probability for node j is: P(j) = (alpha + degree(j)^power) / sumDegrees

Parameters

N	Total number of nodes in the final network.
power	Exponent of the preferential attachment (typically 1).
m0	Number of nodes in the initial seed clique.
m	Number of edges each new node attaches to existing nodes.
alpha	Additive constant in the attachment probability (zero-appeal).
mode	"graph" for undirected, anything else for directed.

Returns

true on success, false if the user cancelled.

randomNetSmallWorldCreate()

bool Graph::randomNetSmallWorldCreate	(	const int &	N,
		const int &	degree,
		const double &	beta,
		const QString &	mode )

Creates a Watts–Strogatz small-world random network.

The algorithm works in two phases:

1. Build a ring lattice of N nodes each connected to degree neighbours.
2. Rewire each edge with probability beta to a randomly chosen node, producing the characteristic short path lengths and high clustering.

Parameters

N	Number of nodes.
degree	Number of neighbours each node is initially connected to (must be even).
beta	Rewiring probability in [0,1]. 0 = pure lattice, 1 = random graph.
mode	"graph" for undirected, anything else for directed.

Returns

true on success, false if the user cancelled.

relationAdd()

void Graph::relationAdd	(	const QString &	relName,
		const bool &	changeRelation = false )

Adds a new relation to the graph.

Adds a new relation named relName, emitting signal to MW UI, and optionally changing current graph relation to the new one. Called by file parser and various Graph methods, i.e clear() etc.

Parameters

relName

relationAddCocitation()

void Graph::relationAddCocitation

(

)

Creates a new symmetric relation by connecting those actors that are cocitated by others. In the new relation, an edge will exist between actor i and actor j only if C(i,j) > 0, where C the Cocitation Matrix. Thus the actor pairs cited by more common neighbors will appear with a stronger tie between them than pairs those cited by fewer common neighbors. The resulting relation is symmetric.

relationCurrent

int Graph::relationCurrent ( )

slot

Gets the current relation number.

Returns

int

relationCurrentName

QString Graph::relationCurrentName ( ) const

slot

Gets the current relation name.

Returns

string

relationCurrentRename [1/2]

void Graph::relationCurrentRename ( const QString & newName )

slot

Overload. Renames current relation to newName, without emitting any signal to MW.

Parameters

newName

relationCurrentRename [2/2]

void Graph::relationCurrentRename ( const QString & newName, const bool & signalMW )

slot

Renames current relation to newName, optionally emitting a signal to MW.

Parameters

newName

relationNext

void Graph::relationNext ( )

slot

Changes graph to next relation.

relationPrev

void Graph::relationPrev ( )

slot

Changes graph to previous relation.

relations()

int Graph::relations

(

)

Returns the count of relationships in this Graph.

Returns

int

relationsClear()

void Graph::relationsClear

(

)

Clears relationships in this Graph.

relationSet

void Graph::relationSet ( int relNum = RAND_MAX, const bool & updateUI = true )

slot

Changes the current relation, and optionally emits signals to MW/GW (default: true).

Slots to signals from MainWindow

Forces all enabled vertices to disable edges in the old relation and enable edges of the new relation

Then, if updateUI==true (default), it emits signals to MW and GW to update the MW UI and toggle the edges on the GW, respectivelly.

Called from Parser, Graph methods and when the user selects a relation in the MW combo box.

Parameters

relNum	int
updateUI	bool

resetDistanceAggregates()

void Graph::resetDistanceAggregates

(

)

resetDistanceCentralityCacheFlags()

void Graph::resetDistanceCentralityCacheFlags

(

)

resetProgressCanceled()

void Graph::resetProgressCanceled

(

)

Resets the cancellation status.

resolveClasses() [1/2]

void Graph::resolveClasses ( qreal C, H_StrToInt & discreteClasses, int & classes )

private

Checks if score C is a new prominence class If yes, it stores that number in a QHash<QString,int> type where the score is the key. If no, increases the frequency of this prominence score by 1 Called from graphDistancesGeodesic().

Parameters

C
discreteClasses
classes

resolveClasses() [2/2]

void Graph::resolveClasses ( qreal C, H_StrToInt & discreteClasses, int & classes, int vertex )

private

Overloaded method. It only adds displaying current vertex for debugging purposes.

Parameters

C
discreteClasses
classes
vertex

runOnGuiThread()

void Graph::runOnGuiThread ( std::function< void()> fn )

protected

Runs fn on the application's main/GUI thread, regardless of the calling thread.

Exists because Graph's own thread affinity is graphThread (see moveToThreadFacade()), but some UI façade code (currently: the prominence-distribution chart builders in graph_ui_prominence_distribution.cpp) constructs real Qt GUI objects — QtCharts series/axes, QChartView::grab() for PNG export — which is only ever safe on the main/GUI thread, per Qt's single-threaded-GUI rule. Those objects would otherwise inherit whichever thread happens to be calling in, which is graphThread whenever this is reached from algorithm-slice code (layouts, report generation, etc.) that now genuinely executes there (see #254).

Uses QCoreApplication::instance() rather than the qApp macro (which force-casts to QApplication*) so this stays correct in the headless CLI, where only a QCoreApplication is constructed - never a QApplication. Qt::AutoConnection (the default, used explicitly here for clarity) resolves to a direct, synchronous call when already on the target thread - which is always true for the CLI, since it never creates a second thread - so this has no behavioural or performance effect there. In GUI mode, called from graphThread, it resolves to queued delivery on the next GUI event loop iteration.

Parameters

fn	The work to run on the main/GUI thread.

saveToAdjacencyFormat()

bool Graph::saveToAdjacencyFormat	(	const QString &	fileName,
		const bool &	saveEdgeWeights = true )

Saves the active graph to an adjacency-formatted file.

Parameters

fileName

Returns

bool

saveToDotFormat()

bool Graph::saveToDotFormat

(

const QString &

fileName

)

Saves the active graph (current relation) to a GraphViz DOT file.

Produces a valid digraph or graph block. Node identifiers are n<num> to avoid issues with labels that contain spaces or special characters. All standard visual properties (label, color, shape) and custom node/edge attributes are written as DOT attribute lists so the file round-trips through parseAsDot() without data loss.

Only the currently active relation is exported. If the graph has multiple relations the caller is responsible for warning the user before invoking this method.

Parameters

fileName

Destination file path.

Returns

true on success, false if the file could not be opened.

saveToEdgeListSimpleFormat()

bool Graph::saveToEdgeListSimpleFormat

(

const QString &

fileName

)

Saves the active relation to a simple (unweighted) edge list file.

Format: one edge per line — "source target" — using the node label as identifier (spaces in labels replaced with underscores). A comment header line describes the columns.

Only the currently active relation is exported.

Parameters

fileName

Destination file path.

Returns

true on success, false if the file cannot be opened.

saveToEdgeListWeightedFormat()

bool Graph::saveToEdgeListWeightedFormat

(

const QString &

fileName

)

Saves the active relation to a weighted edge list file.

Format: one edge per line — "source target weight" — using the node label as identifier (spaces in labels replaced with underscores). A comment header line describes the columns.

Only the currently active relation is exported.

Parameters

fileName

Destination file path.

Returns

true on success, false if the file cannot be opened.

saveToFile()

void Graph::saveToFile	(	const QString &	fileName,
		const int &	fileType,
		const bool &	saveEdgeWeights = true,
		const bool &	saveZeroWeightEdges = false )

Saves the current graph to a file.

Checks the requested file type and calls the corresponding saveGraphTo...() method

Parameters

fileName
fileType

saveToGraphMLFormat()

bool Graph::saveToGraphMLFormat	(	const QString &	fileName,
		const bool &	saveZeroWeightEdges = false,
		QString	networkName = "",
		int	maxWidth = 0,
		int	maxHeight = 0 )

Saves the current graph to a GraphML-formatted file.

Parameters

fileName
networkName
maxWidth
maxHeight

Returns

bool

saveToPajekFormat()

bool Graph::saveToPajekFormat	(	const QString &	fileName,
		QString	networkName = "",
		int	maxWidth = 0,
		int	maxHeight = 0 )

Save the current graph to a Pajek (.paj) file.

Writes:

• *Network <name>
• *Vertices N with labels, colors, positions, and shapes (as available)
• Edges either as:
  • single-relation: *Arcs + *Edges sections (legacy behavior), or
  • multi-relation: one *Matrix block per relation.

Relation header export (multi-relation):

• Canonical matrix header syntax:
  • *Matrix :k
  • *Matrix :k "Label"
• If the relation label is empty/unlabeled, no quotes are emitted.
• Labels are normalized to avoid polluted forms like 9 'star':
  • strips wrapping quotes
  • strips a leading index token (e.g. 9, 9:)
  • escapes internal quotes per Pajek convention (" -> "")

Notes:

• Numeric output is written using the C locale for stable decimal formatting.
• maxWidth/maxHeight are used to scale vertex coordinates to Pajek's normalized coordinate space. If 0, current canvas dimensions are used.
• The active relation is restored before returning.
• On success, updates the graph’s file name and file format metadata.

Parameters

fileName	Output file path (should end in .paj).
networkName	Optional exported network name. If empty, uses graph name. If "unnamed", derives from file name without extension.
maxWidth	Canvas width used to normalize X coordinates (0 => use current).
maxHeight	Canvas height used to normalize Y coordinates (0 => use current).

Returns

true on success, false if the file cannot be opened/written.

Returns a canonical label for Pajek *Matrix headers (no surrounding quotes).

Pajek *Matrix Header Export Contract

EXPORT IS STRICT AND CANONICAL.

Multirelational headers MUST be written in exactly one of these forms:

*Matrix :k *Matrix :k "Label"

Where:

• k is the 1-based relation index.
• No leading index is embedded in the label.
• No empty quotes are emitted.
• No alternate forms like "*Matrix k:", "*Matrix k: """, etc.

Relation labels are normalized before export:

• Strip wrapping quotes.
• Strip leading index pollution (e.g. "9 'star'" → "star").
• Treat empty or index-only labels as unlabeled.
• Escape internal double quotes as "" (Pajek rule).

IMPORT remains tolerant. EXPORT remains canonical.

saveToUCINETDLFormat()

bool Graph::saveToUCINETDLFormat

(

const QString &

fileName

)

Saves the current graph to a UCINET DL file (FULLMATRIX format).

Emits one header block followed by all relation matrices back-to-back in the DATA section. The parser advances its relation counter whenever the source row counter exceeds N, so no separator is needed between matrices.

Single relation: N=<n>, no NM or LEVEL LABELS. Multi-relation: N=<n> NM=<m>, LEVEL LABELS block with relation names.

Node labels are always written in ROW LABELS / COLUMN LABELS sections. Edge weights are written as-is (0 for absent edges).

Parameters

fileName

Destination file path.

Returns

true on success, false if the file cannot be opened.

setAverageDistanceCached()

void Graph::setAverageDistanceCached

(

qreal

v

)

setConnectedCached()

void Graph::setConnectedCached

(

bool

v

)

setDiameterCached()

void Graph::setDiameterCached

(

int

v

)

setDirected()

void Graph::setDirected	(	const bool &	toggle = true,
		const bool &	signalMW = true )

Toggles the graph directed or undirected.

Parameters

toggle
signalMW

setEdgeColor

void Graph::setEdgeColor ( const int & v1, const int & v2, const QString & color )

signal

setEdgeLabel

void Graph::setEdgeLabel ( const int & v1, const int & v2, const QString & label )

signal

setEdgeWeight

void Graph::setEdgeWeight ( const int & v1, const int & v2, const qreal & weight )

signal

setFileFormat()

void Graph::setFileFormat

(

const int &

fileFormat

)

setFileName()

void Graph::setFileName

(

const QString &

fileName

)

Sets the file name of the current graph.

Parameters

fileName

setModStatus()

void Graph::setModStatus	(	const int &	graphNewStatus,
		const bool &	signalMW = true )

Sets the graph modification status.

If there are major changes or new network, it signals to MW to update the UI.

Parameters

int	graphNewStatus
bool	signalMW

setName()

void Graph::setName

(

const QString &

graphName

)

Sets the name of the current graph.

Parameters

graphName

setNodeColor

void Graph::setNodeColor ( const int & v, const QString & color )

signal

setNodeLabel

void Graph::setNodeLabel ( const int & v, const QString & label )

signal

setNodeLabelColor

void Graph::setNodeLabelColor ( const int & v, const QString & color )

signal

setNodeLabelDistance

void Graph::setNodeLabelDistance ( const int & v, const int & distance )

signal

setNodeLabelSize

void Graph::setNodeLabelSize ( const int & v, const int & size )

signal

setNodeNumberColor

void Graph::setNodeNumberColor ( const int & v, const QString & color )

signal

setNodeNumberDistance

void Graph::setNodeNumberDistance ( const int & v, const int & distance )

signal

setNodeNumberSize

void Graph::setNodeNumberSize ( const int & v, const int & size )

signal

setNodePos

void Graph::setNodePos ( const int & , const qreal & , const qreal &  )

signal

setNodeShape

void Graph::setNodeShape ( const int & v, const QString & shape, const QString & iconPath = QString() )

signal

setNodeSize

void Graph::setNodeSize ( const int & v, const int & size )

signal

setReportsChartType()

void Graph::setReportsChartType

(

const int &

type

)

Sets the chart type in reports.

Parameters

type

setReportsDataDir()

void Graph::setReportsDataDir

(

const QString &

dir

)

Sets the directory where reports are saved This is used when exporting prominence distribution images to be used in HTML reports.

Parameters

dir

setReportsLabelLength()

void Graph::setReportsLabelLength

(

const int &

length

)

Sets the length of labels in reports.

Parameters

length

setReportsOutputFormat()

void Graph::setReportsOutputFormat

(

const int &

format

)

Sets the output format (HTML or CSV) new reports are written in.

Parameters

format

setReportsRealNumberPrecision()

void Graph::setReportsRealNumberPrecision

(

const int &

precision

)

Sets the precision (number of fraction digits) the app will use when writing real numbers in reports.

Parameters

precision

setSelectionChanged

void Graph::setSelectionChanged ( const QList< int > selectedVertices, const QList< SelectedEdge > selectedEdges )

slot

Sets the user-selected vertices and edges.

Usually called from GW, it emits selection counts to MW

Parameters

selectedVertices
selectedEdges

setSymmetric()

void Graph::setSymmetric

(

)

Transforms the graph to symmetric (all edges reciprocal).

setSymmetricCached()

void Graph::setSymmetricCached

(

bool

v

)

setUndirected()

void Graph::setUndirected	(	const bool &	toggle = true,
		const bool &	signalMW = true )

Makes the graph undirected or directed.

Parameters

toggle
signalMW

setVertexVisibility

void Graph::setVertexVisibility ( const int & number, const bool & toggle )

signal

setWeighted()

void Graph::setWeighted

(

const bool &

toggle = true

)

Sets the graph to be weighted ( valued edges ).

Parameters

toggle

showZeroWeightEdgesSet()

void Graph::showZeroWeightEdgesSet

(

const bool &

toggle

)

sign()

int Graph::sign

(

const qreal &

D

)

Graph::sign returns the sign of number D as integer (1 or -1).

Parameters

D

Returns

signalDatasetDescription

void Graph::signalDatasetDescription ( QString )

signal

signalDrawEdge

void Graph::signalDrawEdge ( const int & v1, const int & v2, const qreal & weight, const QString & label = "", const QString & color = "black", const int & type = 0, const bool & drawArrows = true, const bool & bezier = false, const bool & weightNumbers = false )

signal

signalDrawNode

void Graph::signalDrawNode ( const QPointF & p, const int & num, const int & size, const QString & nodeShape, const QString & nodeIconPath, const QString & nodeColor, const QString & numberColor, const int & numSize, const int & numDistance, const QString & label, const QString & labelColor, const int & labelSize, const int & labelDistance )

signal

Signals to GraphicsWidget

signalEdgeClicked

void Graph::signalEdgeClicked ( const MyEdge & edge = MyEdge(), const bool & openMenu = false )

signal

signalEdgeType

void Graph::signalEdgeType ( const int & v1, const int & v2, const int & type )

signal

signalGraphDirectedChanged

void Graph::signalGraphDirectedChanged ( const bool & directed )

signal

signalGraphLoaded

void Graph::signalGraphLoaded ( const int & fileType, const QString & fileName = QString(), const QString & netName = QString(), const int & totalNodes = 0, const int & totalLinks = 0, const qreal & density = 0, const qint64 & elapsedTime = 0, const QString & message = QString() )

signal

signalGraphModified

void Graph::signalGraphModified ( const bool & undirected, const int & vertices, const int & edges, const qreal & density, const bool & notSaved = true )

signal

signalGraphSavedStatus

void Graph::signalGraphSavedStatus ( const int & status )

signal

signalLayoutFinished

void Graph::signalLayoutFinished ( )

signal

signalNetworkManagerRequest

void Graph::signalNetworkManagerRequest ( const QUrl & currentUrl, const NetworkRequestType & type )

signal

Signals to MainWindow

signalNodeClickedInfo

void Graph::signalNodeClickedInfo ( const int & number = 0, const QPointF & p = QPointF(), const QString & label = QString(), const int & inDegree = 0, const int & outDegree = 0 )

signal

signalNodesFound

void Graph::signalNodesFound ( const QList< int > foundList )

signal

signalPromininenceDistributionChartUpdate

void Graph::signalPromininenceDistributionChartUpdate ( QAbstractSeries * series, QAbstractAxis * axisX = Q_NULLPTR, const qreal & min = 0, const qreal & max = 0, QAbstractAxis * axisY = Q_NULLPTR, const qreal & minF = 0, const qreal & maxF = 0 )

signal

signalRelationAddToMW

void Graph::signalRelationAddToMW ( const QString & newRelation )

signal

signalRelationChangedToGW

void Graph::signalRelationChangedToGW ( int )

signal

signalRelationChangedToMW

void Graph::signalRelationChangedToMW ( const int & relIndex = RAND_MAX )

signal

signalRelationRenamedToMW

void Graph::signalRelationRenamedToMW ( const QString & newRelName )

signal

signalRelationsClear

void Graph::signalRelationsClear ( )

signal

signalRemoveEdge

void Graph::signalRemoveEdge ( const int & v1, const int & v2, const bool & removeReverse )

signal

signalRemoveNode

void Graph::signalRemoveNode ( int )

signal

signalSelectionChanged

void Graph::signalSelectionChanged ( const int & selectedVertices, const int & selectedEdges )

signal

signalSetEdgesVisibilityBatch

void Graph::signalSetEdgesVisibilityBatch ( const QList< EdgeVisibilityChange > & changes )

signal

signalSetEdgeVisibility

void Graph::signalSetEdgeVisibility ( const int & relation, const int & source, const int & target, const bool & toggle, const bool & preserveReverseEdge = false, const int & edgeWeight = 1, const int & reverseEdgeWeight = 1 )

signal

signalWebCrawlParse

void Graph::signalWebCrawlParse ( QNetworkReply * reply )

signal

slotCancelComputation

void Graph::slotCancelComputation ( )

slot

Slot called by MainWindow when the user clicks Cancel in the progress dialog.

slotHandleCrawlerRequestReply

void Graph::slotHandleCrawlerRequestReply ( )

slot

Gets the reply of a MW network request made by Web Crawler, and emits that reply as is to the Web Crawler.

startWebCrawler

void Graph::startWebCrawler ( const QUrl & startUrl, const QStringList & urlPatternsIncluded, const QStringList & urlPatternsExcluded, const QStringList & linkClasses, const int & maxNodes, const int & maxLinksPerPage, const bool & intLinks, const bool & childLinks, const bool & parentLinks, const bool & selfLinks, const bool & extLinksIncluded, const bool & extLinksCrawl, const bool & socialLinks, const bool & delayedRequests )

slot

Creates a new WebCrawler, that will parse the downloaded HTML code of each webpage we download. Moves the WebCrawler to a new thread and starts the thread. Then creates the fist node (initial url), and starts the web spider to download the first HTML page. Called by MW with user options.

Parameters

startUrl
urlPatternsIncluded
urlPatternsExcluded
linkClasses
maxNodes
maxLinksPerPage
intLinks
childLinks
parentLinks
selfLinks
extLinksIncluded
extLinksCrawl
socialLinks
delayedRequests

statusMessage

void Graph::statusMessage ( const QString & message )

signal

subgraphExtract

Graph * Graph::subgraphExtract ( const QString & name, const bool & includeCustomAttributes = true )

slot

Extracts currently visible (non-filtered) nodes and their inter-edges into an independent Graph copy.

Collects all enabled vertices, then delegates to subgraphFromVertexList().

Parameters

name	Name assigned to the new graph.
includeCustomAttributes	When true, node and edge custom attribute maps are copied verbatim.

Returns

Heap-allocated Graph* containing the subgraph, or nullptr if no visible vertices exist.

subgraphExtractFromSelection

Graph * Graph::subgraphExtractFromSelection ( const QString & name, const bool & includeCustomAttributes = true )

slot

Extracts currently selected nodes and their inter-edges into an independent Graph copy.

Uses the current canvas selection (getSelectedVertices()), then delegates to subgraphFromVertexList(). Only edges between selected nodes are included.

Parameters

name	Name assigned to the new graph.
includeCustomAttributes	When true, node and edge custom attribute maps are copied verbatim.

Returns

Heap-allocated Graph* containing the subgraph, or nullptr if no nodes are selected.

subgraphFromVertexList()

Graph * Graph::subgraphFromVertexList ( const QList< int > & vertexNums, const QString & name, const bool & includeCustomAttributes = true )

private

Core helper: builds an independent Graph copy from an explicit list of vertex numbers.

private member functions

Shared by subgraphExtract() (visible nodes) and subgraphExtractFromSelection() (selected nodes). Vertices are renumbered from 1. For every relation, enabled edges whose both endpoints are in vertexNums are copied. All visual properties and, when includeCustomAttributes is true, node and edge custom attribute maps are preserved verbatim.

The caller takes ownership of the returned Graph object.

Parameters

vertexNums	Source vertex numbers to include.
name	Name assigned to the new graph.
includeCustomAttributes	When true, node and edge custom attribute maps are copied verbatim.

Returns

Heap-allocated Graph* containing the subgraph, or nullptr if vertexNums is empty.

symmetricCached()

bool Graph::symmetricCached

(

)

const

triadType_examine_MAN_label()

void Graph::triadType_examine_MAN_label	(	int	mut,
		int	asy,
		int	nul,
		GraphVertex *	vert1,
		GraphVertex *	vert2,
		GraphVertex *	vert3 )

Examines the triad type (in Mutual-Asymmetric-Null label format) and increases by one the proper frequency element inside QList::triadTypeFreqs

uiProminenceDistributionArea()

void Graph::uiProminenceDistributionArea ( const QVector< QPair< qreal, qreal > > & points, const qreal min, const qreal max, const qreal minF, const qreal maxF, const QString & name, const QString & distImageFileName )

protected

uiProminenceDistributionBars()

void Graph::uiProminenceDistributionBars ( const QStringList & categories, const QVector< qreal > & frequencies, const qreal min, const qreal max, const qreal minF, const qreal maxF, const QString & name, const QString & distImageFileName )

protected

uiProminenceDistributionSpline()

void Graph::uiProminenceDistributionSpline ( const QVector< QPair< qreal, qreal > > & points, qreal min, qreal max, qreal minF, qreal maxF, const QString & seriesName, const QString & distImageFileName )

protected

vertexAtIndex() [1/2]

GraphVertex * Graph::vertexAtIndex

(

int

idx

)

Returns the vertex at a given index.

Parameters

idx

Returns

GraphVertex*

vertexAtIndex() [2/2]

const GraphVertex * Graph::vertexAtIndex

(

int

idx

)

const

Returns the vertex at a given index.

Parameters

idx

Returns

GraphVertex*

vertexAttributesImport()

int Graph::vertexAttributesImport	(	const QStringList &	headers,
		const QVector< QStringList > &	rows,
		int	idColumn,
		bool	matchByLabel )

Imports custom attributes from a parsed table into existing vertices.

Each row in rows is matched to a vertex either by node number (matchByLabel == false) or by label (matchByLabel == true) using the value in column idColumn. All other columns become custom attributes on the matched vertex. Rows that do not match any vertex are silently skipped.

Example — CSV input with idColumn=0, matchByLabel=false. Native columns are routed to their proper setters (Label, Size, Color, Shape); only Visible is silently skipped (it is a filter state, not a persistent attribute):

#,Label,Size,Color,type,year_founded
1,Alice,12,#ff0000,investor,2010   ← Label/Size/Color updated; type/year_founded → custom attrs
2,Bob,10,#0000ff,founder,2018

Returns

Number of vertices that received at least one attribute update.

vertexClicked()

int Graph::vertexClicked

(

)

const

Returns the number of the clicked vertex.

Returns

int

vertexClickedSet

void Graph::vertexClickedSet ( const int & v1, const QPointF & p )

slot

Sets the clicked vertex.

Signals to MW to show node info on the status bar.

Parameters

v1
p

vertexColor()

QColor Graph::vertexColor

(

const int &

v1

)

const

Graph::vertexColor.

Parameters

v1

Returns

vertexColorInit()

void Graph::vertexColorInit

(

const QString &

color

)

Graph::vertexColorInit default vertex color initialization.

Parameters

color

vertexColorSet()

void Graph::vertexColorSet	(	const int &	v1,
		const QString &	color )

Changes the color of vertex v1.

Parameters

v1
color

vertexComponentId()

const QHash< int, int > & Graph::vertexComponentId ( ) const

inline

vertexCreate

void Graph::vertexCreate ( const int & number, const int & size, const QString & color, const QString & numColor, const int & numSize, const QString & label, const QString & labelColor, const int & labelSize, const QPointF & p, const QString & shape, const QString & iconPath = QString(), const bool & signalMW = false, const QHash< QString, QString > & customAttributes = QHash<QString, QString>() )

slot

Creates a new vertex.

Slots to signals from Parser

Main vertex creation slot, associated with homonymous signal from Parser. Adds a vertex to the Graph and signals drawNode to GW The new vertex has number num and specific color, label, label color, shape and position p.

Parameters

num
size
nodeColor
numColor
numSize
label
lColor
lSize
p
nodeShape
signalMW

vertexCreateAtPos

void Graph::vertexCreateAtPos ( const QPointF & p )

slot

Creates a new vertex in the given position.

Called from GW, with i and p as parameters. Calls the main creation slot with init node values.

Parameters

QPointF

The clicked pos of the new node.

vertexCreateAtPosRandom

void Graph::vertexCreateAtPosRandom ( const bool & signalMW = false )

slot

Creates a new randomly positioned vertex with default values.

Computes a random position p inside the useable canvas area Then calls the main creation slot with init node values.

Parameters

bool

vertexCreateAtPosRandomWithLabel

void Graph::vertexCreateAtPosRandomWithLabel ( const int & i, const QString & label, const bool & signalMW = false )

slot

Creates a new randomly positioned vertex with specific number and label. All other values are from the defaults.

Called from WebCrawler and Parser with parameters label and i. Computes a random position p the useable canvas area Then calls the main creation slot with init node values.

Parameters

i
label
signalMW

vertexCustomAttributeRemove()

void Graph::vertexCustomAttributeRemove	(	const int &	v1,
		const QString &	key )

Removes a single custom attribute key from vertex v1. Does nothing if the key does not exist.

Parameters

v1	Vertex number.
key	Attribute key to remove.

vertexCustomAttributes()

QHash< QString, QString > Graph::vertexCustomAttributes

(

const int &

v1

)

const

Retrieves the vertex at the position specified by the index v1 from the vpos map and calls its customAttributes method.

Parameters

v1	The index of the vertex whose custom attributes are to be accessed.

vertexCustomAttributeSet()

void Graph::vertexCustomAttributeSet	(	const int &	v1,
		const QString &	key,
		const QString &	value )

Sets a single custom attribute key/value on vertex v1, leaving all other attributes untouched.

Parameters

v1	Vertex number.
key	Attribute key.
value	Attribute value.

vertexCustomAttributesSet()

void Graph::vertexCustomAttributesSet	(	const int &	v1,
		const QHash< QString, QString > &	customAttributes )

Sets custom attributes for a specified vertex.

This function assigns a set of custom attributes to a vertex identified by its index. It also updates the modification status to indicate that vertex metadata has been changed.

Parameters

v1	The index of the vertex for which custom attributes are being set.
customAttributes	A QHash containing the custom attributes to be set for the vertex. The keys and values of the QHash are both QStrings.

vertexDegreeIn()

int Graph::vertexDegreeIn

(

int

v1

)

Returns the inDegree (sum of inbound edge weights) of vertex v1.

Parameters

v1

Returns

vertexDegreeOut()

int Graph::vertexDegreeOut

(

int

v1

)

Returns the outDegree (sum of outbound edge weights) of vertex v1.

Parameters

v1

Returns

vertexEdgesInbound()

int Graph::vertexEdgesInbound

(

int

v1

)

Returns the number of inbound edges (arcs) to vertex v1.

Parameters

v1

Returns

int

vertexEdgesOutbound()

int Graph::vertexEdgesOutbound

(

int

v1

)

Returns the number of outbound edges (arcs) from vertex v1.

Parameters

v1

Returns

vertexExists() [1/2]

bool Graph::vertexExists

(

const int &

v1

)

Checks if the given vertex exists in the graph.

Parameters

vertex

number

Returns

true if it exists

vertexExists() [2/2]

bool Graph::vertexExists

(

const QString &

label

)

Checks if there is a vertex with a specific label in the graph.

Parameters

label

Returns

true if it exists

vertexFilterByAttribute

void Graph::vertexFilterByAttribute ( const FilterCondition & cond )

slot

Shows only vertices whose custom attribute satisfies cond; all other vertices are hidden.

Non-destructive: pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes. Call vertexFilterRestoreAll() to undo.

Returns early with a status message if the resulting visible set is empty.

Parameters

cond	The filter condition (key, operator, value).

vertexFilterByCentrality

void Graph::vertexFilterByCentrality ( const float threshold, const bool overThreshold, const IndexType centralityIndex )

slot

Filters vertices by their score on the given centrality or prestige index.

Pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes, so vertexFilterRestoreAll() can undo this filter.

Performs two passes over the graph:

• Pass 1: Enables or disables each vertex depending on whether its score satisfies the threshold condition. Emits setVertexVisibility for every vertex whose state changes.
• Pass 2: Updates all edges. An edge is made visible only when both its source and target vertices are enabled after pass 1. Emits signalSetEdgeVisibility for every out-edge of every vertex.

The two-pass design is required so that edge visibility can be determined correctly: during pass 2 all vertices are already in their final enabled state, so both endpoints of any edge can be queried reliably.

Returns early with a status message if the requested index has not been computed yet — the user should run the corresponding analysis first.

Parameters

threshold	Score threshold to compare against.
overThreshold	If true, disable vertices with score >= threshold; if false, disable vertices with score <= threshold.
centralityIndex	The centrality or prestige index to use, as defined by the IndexType enum in global.h.

See also

isCentralityIndexComputed()

vertexFilterRestoreAll()

IndexType

vertexFilterByEgoNetwork

void Graph::vertexFilterByEgoNetwork ( const int v1, const int depth = 1 )

slot

Saves current visibility state and shows only the ego network of vertex v1 at the given depth.

Non-destructive: pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes. Call vertexFilterRestoreAll() to undo.

Only out-edges in the current relation are used to determine neighbors. Works correctly for both directed and undirected graphs.

Parameters

v1	The ego vertex (center of the neighborhood).
depth	Neighborhood depth (currently only depth=1 supported).

vertexFilterByQuery

void Graph::vertexFilterByQuery ( const GraphQuery & query )

slot

Shows only vertices whose custom attributes satisfy ALL conditions in query.

AND logic: a vertex must pass every condition to remain visible. Non-destructive: pushes a GraphVisibilitySnapshot with FilterSpec::Type::Query before making any changes. Undo via vertexFilterRemoveAt() or vertexFilterRestoreAll().

Follows the same two-pass pattern as vertexFilterByAttribute() so edge visibility is resolved after all vertex states are final.

Parameters

query

Compound query; all conditions should have Scope::Nodes.

vertexFilterBySelection

void Graph::vertexFilterBySelection ( const QList< int > & selectedVertices )

slot

Saves current visibility state and shows only the selected vertices and the edges between them.

Non-destructive: pushes a GraphVisibilitySnapshot onto m_visibilityHistory before making any changes. Call vertexFilterRestoreAll() to undo.

Implements the "Focus on Selection" mode (#210): all nodes not in selectedVertices are hidden, and only edges whose both endpoints are in the selection remain visible.

Parameters

selectedVertices

List of vertex numbers that should remain visible.

vertexFilterRemoveAt

void Graph::vertexFilterRemoveAt ( int stackIndex )

slot

Removes the filter at stackIndex (0 = oldest) and replays the rest.

Drains m_visibilityHistory to a list, restores to the pre-first-filter base state (the oldest snapshot), then re-applies every spec except the one at stackIndex. Each replay call pushes a fresh snapshot, leaving the stack correct for subsequent single-step restores.

Centrality replay requires the relevant index to still be computed; if it is not, vertexFilterByCentrality() returns early and that step is skipped.

vertexFilterReplaySpec()

void Graph::vertexFilterReplaySpec ( const FilterSpec & spec )

private

Re-applies one filter from its stored replay parameters.

Used by vertexFilterRemoveAt() during stack reconstruction.

vertexFilterRestoreAll

void Graph::vertexFilterRestoreAll ( )

slot

Restores vertex and edge visibility from the top snapshot on the history stack.

If the stack is empty (no filter is active), this is a no-op. Pops the snapshot after restoring, so repeated calls walk back through the filter history one step at a time.

vertexFindByIndexScore()

bool Graph::vertexFindByIndexScore	(	const int &	index,
		const QStringList &	thresholds,
		const bool &	considerWeights,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Finds vertices by their index score.

Parameters

QStringList

Returns

vertexFindByLabel()

bool Graph::vertexFindByLabel

(

const QStringList &

labelList

)

Finds vertices by their label.

Parameters

QStringList

Returns

vertexFindByNumber()

bool Graph::vertexFindByNumber

(

const QStringList &

numList

)

Finds vertices in strList by their number.

Parameters

QStringList

Returns

vertexIndexByNumber()

int Graph::vertexIndexByNumber

(

int

v

)

const

Returns the index of a vertex by its number.

Returns the vpos or -1

Complexity: O(logN) for vpos retrieval

Parameters

vertex

number

Returns

vertex pos or -1

vertexIndexIfExists() [1/2]

int Graph::vertexIndexIfExists

(

const int &

v1

)

Returns the internal position of the given vertex in m_graph, if it exists.

Complexity: O(logN) for vpos retrieval

Parameters

vertex

number

Returns

vertex pos or -1

vertexIndexIfExists() [2/2]

int Graph::vertexIndexIfExists

(

const QString &

label

)

Returns the internal position of the vertex with the given label, if it exists.

Complexity: O(N)

Parameters

label

Returns

vertex pos or -1

vertexIsolated()

bool Graph::vertexIsolated

(

const int &

v1

)

const

Checks if vertex is isolated.

Parameters

v1

Returns

vertexIsolatedAllToggle

void Graph::vertexIsolatedAllToggle ( const bool & toggle )

slot

Toggles the status of all isolated vertices (those without links).

For each isolate vertex in the Graph, emits the setVertexVisibility signal

Parameters

toggle

vertexLabel()

QString Graph::vertexLabel

(

const int &

v

)

const

Returns the label of a vertex v1.

Parameters

v1

Returns

vertexLabelColorInit()

void Graph::vertexLabelColorInit

(

QString

color

)

Graph::vertexLabelColorInit Changes the default vertex label color.

Parameters

color

vertexLabelColorSet()

void Graph::vertexLabelColorSet	(	const int &	v1,
		const QString &	color )

Changes the label color of vertex v1 or all vertices if v1 = 0.

Parameters

v1
color

vertexLabelDistanceAllSet()

void Graph::vertexLabelDistanceAllSet

(

const int &

newDistance

)

Changes the distance of all vertex labels from their vertices.

Parameters

size

vertexLabelDistanceInit()

void Graph::vertexLabelDistanceInit

(

const int &

distance

)

Changes the default distance of vertex labels.

Parameters

distance

vertexLabelDistanceSet()

void Graph::vertexLabelDistanceSet	(	const int &	v,
		const int &	newDistance )

Changes the distance.of vertex v label from the vertex.

Parameters

v
size

vertexLabelSet()

void Graph::vertexLabelSet	(	const int &	v1,
		const QString &	label )

Changes the label of a vertex v1.

Parameters

v1
label

vertexLabelSizeInit()

void Graph::vertexLabelSizeInit

(

int

newSize

)

Graph::vertexLabelSizeInit Changes the default size of vertex labels.

Parameters

newSize

vertexLabelSizeSet()

void Graph::vertexLabelSizeSet	(	const int &	v1,
		const int &	labelSize )

Changes the label size of vertex v1 or all vertices if v1=0.

Parameters

v1
size

vertexLabelsVisibilitySet()

void Graph::vertexLabelsVisibilitySet

(

bool

toggle

)

vertexNumberColorInit()

void Graph::vertexNumberColorInit

(

const QString &

color

)

Changes the initial color of the vertex numbers.

Parameters

color

vertexNumberColorSet()

void Graph::vertexNumberColorSet	(	const int &	v1 = 0,
		const QString &	color = "#000000" )

Graph::vertexColorSet Changes the color of vertex v1.

Parameters

v1
color

vertexNumberDistanceInit()

void Graph::vertexNumberDistanceInit

(

const int &

distance

)

Changes the initial distance of vertex numbers.

Parameters

distance

vertexNumberDistanceSet()

void Graph::vertexNumberDistanceSet	(	const int &	v,
		const int &	newDistance )

Changes the distance.of vertex v number from the vertex.

Parameters

v
size

vertexNumberMax()

int Graph::vertexNumberMax

(

)

Returns the number of the last vertex in the graph.

Returns

int

vertexNumberMin()

int Graph::vertexNumberMin

(

)

Returns the number of the first vertex in the graph.

Returns

int

vertexNumberSizeInit()

void Graph::vertexNumberSizeInit

(

const int &

size

)

Changes the initial size of vertex numbers.

Parameters

size

vertexNumberSizeSet()

void Graph::vertexNumberSizeSet	(	const int &	v,
		const int &	size )

Changes the size of vertex v number.

Parameters

v
size

vertexOutNeighborsSet()

QSet< int > Graph::vertexOutNeighborsSet	(	const int &	v1,
		const bool	includeInEdges = false )

Returns the set of all 1-hop neighbors of vertex v1 in the current relation.

By default returns only out-neighbors (vertices reachable via enabled out-edges). When includeInEdges is true, also includes in-neighbors (vertices that have an enabled out-edge to v1), giving the full set of adjacent nodes regardless of edge direction.

Only edges in the current relation that are enabled are considered.

Parameters

v1	The vertex number to query.
includeInEdges	If true, union of out- and in-neighbors is returned. Default: false (out-neighbors only).

Returns

QSet<int> of neighbor vertex numbers.

vertexPos()

QPointF Graph::vertexPos

(

const int &

v1

)

const

Returns the position of the given vertex.

Parameters

v1

Returns

vertexPosSet()

void Graph::vertexPosSet	(	const int &	v1,
		const int &	x,
		const int &	y )

Changes the position of the given vertex.

Called from MW/GW when node moves to update its position

Parameters

v1
x
y

vertexPtr()

GraphVertex * Graph::vertexPtr

(

const int

v

)

vertexReciprocalNeighborsList()

QList< int > Graph::vertexReciprocalNeighborsList

(

const int &

v1

)

Returns a list of all vertices reciprocally connected to vertex v1 in the current relation.

A vertex is included only if there is an enabled, reciprocal edge between v1 and that vertex in the current relation. For general 1-hop neighbors (out-edges only), iterate m_outEdges directly.

Parameters

v1	The vertex number to query.

Returns

QList<int> of reciprocally connected vertex numbers.

vertexReciprocalNeighborsSet()

QSet< int > Graph::vertexReciprocalNeighborsSet

(

const int &

v1

)

Returns the set of all vertices reciprocally connected to vertex v1 in the current relation.

A vertex is included only if there is an enabled, reciprocal edge between v1 and that vertex in the current relation. For general 1-hop neighbors (out-edges only), iterate m_outEdges directly.

Parameters

v1	The vertex number to query.

Returns

QSet<int> of reciprocally connected vertex numbers.

vertexRemove()

void Graph::vertexRemove

(

const int &

v1

)

Removes the vertex v1 from the graph First, it removes all edges to doomed from other vertices Then it changes the vpos of all subsequent vertices inside m_graph Finally, it removes the vertex.

Parameters

int

v1

vertexRemoveDummyNode

void Graph::vertexRemoveDummyNode ( int i )

slot

Deletes a dummy node.

This is called from Parser (as pajek) to delete any redundant (dummy) nodes.

Parameters

int	i number of node

vertexShape()

QString Graph::vertexShape

(

const int &

v1

)

Returns the shape of this vertex.

Parameters

v1

Returns

vertexShapeIconPath()

QString Graph::vertexShapeIconPath

(

const int &

v1

)

Returns the IconPath of vertex v1.

Parameters

v1

Returns

vertexShapeSet()

void Graph::vertexShapeSet	(	const int &	v1,
		const QString &	shape,
		const QString &	iconPath = QString() )

Changes the shape and iconPath of vertex v1, or all vertices if v1=-1.

Parameters

v1
shape
iconPath

vertexShapeSetDefault()

void Graph::vertexShapeSetDefault	(	const QString	shape,
		const QString &	iconPath = QString() )

Sets the default vertex shape and iconPath.

Parameters

shape
iconPath

vertexSize()

int Graph::vertexSize

(

const int &

v

)

const

Returns the size of vertex v.

Parameters

v

Returns

int

vertexSizeInit()

void Graph::vertexSizeInit

(

const int

size

)

Sets the initial vertex size.

Parameters

size

vertexSizeSet()

void Graph::vertexSizeSet	(	const int &	v,
		const int &	size )

Changes the size of a vertex v or all vertices if v=0.

Called from MW (i.e. user changing node properties)

Parameters

v
size

vertices()

int Graph::vertices	(	const bool &	dropIsolates = false,
		const bool &	countAll = false,
		const bool &	recount = false )

Gets the number of vertices in the graph.

If countAll = true, returns |V| where V the set of all (enabled or not) vertices If countAll = false, it skips disabled vertices If countAll = false and dropIsolates = true, it skips both disabled and isolated vertices

The cache is keyed on (dropIsolates, countAll): a cached count computed under one combination of flags must never be handed back to a caller asking with a different combination, since callers size fixed allocations off this return value (see #274).

Parameters

dropIsolates
countAll

Returns

verticesBegin()

VList::const_iterator Graph::verticesBegin

(

)

const

iterator helpers

verticesCreateSubgraph()

void Graph::verticesCreateSubgraph	(	QList< int >	vList,
		const int &	type = SUBGRAPH_CLIQUE,
		const int &	center = 0 )

Creates a subgraph (clique, star, cycle, line) with vertices in vList Iff vList is empty, then fallbacks to the m_verticesSelected.

Parameters

vList

verticesEnd()

VList::const_iterator Graph::verticesEnd

(

)

const

verticesList()

QList< int > Graph::verticesList

(

)

Returns a list of all vertices numbers inside the graph.

Returns

QList<int>

verticesListIsolated()

QList< int > Graph::verticesListIsolated

(

)

Returns a list of all isolated vertices inside the graph.

Returns

QList<int>

verticesSet()

QSet< int > Graph::verticesSet

(

)

Returns a QSet of all vertices numbers inside the graph.

Returns

verticesWithInboundEdges()

int Graph::verticesWithInboundEdges

(

)

Returns the sum of vertices having edgesInbound.

Returns

verticesWithOutboundEdges()

int Graph::verticesWithOutboundEdges

(

)

Returns the sum of vertices having edgesOutbound.

Returns

verticesWithReciprocalEdges()

int Graph::verticesWithReciprocalEdges

(

)

Returns the sum of vertices having reciprocal edges.

Returns

visibilityHistoryEmpty

bool Graph::visibilityHistoryEmpty ( ) const

slot

Returns true if the visibility history stack is empty.

Used by the UI to determine whether a "Restore All" action should be enabled. The stack holds one entry per non-destructive filter operation (e.g. ego network focus). Each call to vertexFilterRestoreAll() pops one entry; when the stack is empty, there is nothing left to restore.

Returns

true if no filter snapshots are pending, false otherwise.

See also

vertexFilterByEgoNetwork()

vertexFilterRestoreAll()

walksBetween()

int Graph::walksBetween	(	int	v1,
		int	v2,
		int	length )

Calculates and returns the number of walks of a given length between v1 and v2.

Parameters

v1
v2
length

Returns

webCrawlTerminateThreads()

void Graph::webCrawlTerminateThreads

(

QString

reason

)

called from Graph, when closing network, to terminate all crawler processes Also called indirectly when wc_spider finishes

Parameters

reason

webSpider

void Graph::webSpider ( )

slot

A loop, that takes urls awaiting in front of the urlQueue, and signals to the MW to make the network request.

writeCentralityBetweenness()

bool Graph::writeCentralityBetweenness	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes Betweenness centralities to file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityBonacich()

bool Graph::writeCentralityBonacich	(	const QString	fileName,
		const qreal &	alpha,
		const qreal &	beta,
		const bool &	weigths = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false,
		const int &	format = ReportFormat::Html )

writeCentralityCloseness()

bool Graph::writeCentralityCloseness	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the closeness centralities to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityClosenessInfluenceRange()

bool Graph::writeCentralityClosenessInfluenceRange	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the "improved" closeness centrality indices to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityDegree()

bool Graph::writeCentralityDegree	(	const QString	fileName,
		const bool	considerWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the Degree Centrality to a file.

Parameters

fileName
considerWeights
dropIsolates

writeCentralityEccentricity()

bool Graph::writeCentralityEccentricity	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the Eccentricity centralities (aka Harary Graph Centrality) to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityEigenvector()

bool Graph::writeCentralityEigenvector	(	const QString	fileName,
		const bool &	considerWeights = true,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false,
		const int &	format = ReportFormat::Html )

Writes the eigenvector centralities to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityInformation()

bool Graph::writeCentralityInformation	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const int &	format = ReportFormat::Html )

Writes the information centralities to file.

Parameters

fileName
considerWeights
inverseWeights

writeCentralityKatz()

bool Graph::writeCentralityKatz	(	const QString	fileName,
		const qreal &	alpha,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false,
		const int &	format = ReportFormat::Html )

Writes the Katz Centrality report to a file, then displays it.

Parameters

fileName
alpha
considerWeights
inverseWeights
dropIsolates
format

writeCentralityPower()

bool Graph::writeCentralityPower	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes Power Centralities to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCentralityStress()

bool Graph::writeCentralityStress	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the Stress centralities to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeCliqueCensus()

bool Graph::writeCliqueCensus	(	const QString &	fileName,
		const bool	considerWeights )

Calls graphCliques() to compute all cliques (maximal connected subgraphs) of the network. Then writes the results into a file, along with the Actor by clique analysis, the Co-membership matrix and the Hierarchical clustering of overlap matrix.

Parameters

fileName
considerWeights

Write the actor by clique analysis matrix. For each actor-clique pair, we compute the proportion of clique members adjacent

Write the actor by actor analysis matrix. For each pair, we compute their clique co-membership

Write the Hierarchical clustering of overlap matrix

writeClusteringCoefficient()

bool Graph::writeClusteringCoefficient	(	const QString	fileName,
		const bool	considerWeights,
		const int &	format = ReportFormat::Html )

Writes the clustering coefficients to a file.

Parameters

fileName
considerWeights

writeClusteringHierarchical()

bool Graph::writeClusteringHierarchical	(	const QString &	fileName,
		const QString &	varLocation,
		const QString &	matrix = "Adjacency",
		const QString &	metric = "Manhattan",
		const QString &	method = "Complete",
		const bool &	diagonal = false,
		const bool &	dendrogram = false,
		const bool &	considerWeights = true,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false )

Performs Hierarchical Cluster Analysis (HCA) and writes the results to an HTML report file.

The analysis proceeds in three phases:

1. Build the structural equivalence matrix (adjacency or geodesic distances).
2. Run the hierarchical clustering algorithm on the equivalence matrix.
3. Write the HTML report including the equivalence matrix and dendrogram.

Parameters

fileName	Path to the output HTML report file.
varLocation	Whether variables are in rows or columns ("rows"/"cols").
matrix	Input matrix type: "adjacency" or "distances".
metric	Distance/dissimilarity metric (e.g., "euclidean", "manhattan").
method	Clustering linkage method (e.g., "single", "complete", "average").
diagonal	If true, include the diagonal of the matrix in computations.
dendrogram	If true, include dendrogram output in the report.
considerWeights	If true, edge weights are used in distance computations.
inverseWeights	If true, edge weights are inverted before use.
dropIsolates	If true, isolate nodes are excluded from the analysis.

Returns

true on success, false if the computation was cancelled or an error occurred.

writeClusteringHierarchicalResultsToStream()

void Graph::writeClusteringHierarchicalResultsToStream	(	QTextStream &	outText,
		const int	N,
		const bool &	dendrogram = false )

Writes Hierarchical Clustering results to given output stream Before running this method, the method Graph::graphClusteringHierarchical() must execute and return true. Otherwise, the result is unpredictable...

Parameters

outText
N
dendrogram

writeDataSetToFile()

void Graph::writeDataSetToFile	(	const QString	dir,
		const QString	fileName )

Writes a "famous" dataset to the given file Datasets are hardcoded! They are exported in the given fileName...

TODO: Move all these datasets to a separate class

Parameters

fileName

writeEccentricity()

bool Graph::writeEccentricity	(	const QString	fileName,
		const bool	considerWeights = false,
		const bool	inverseWeights = false,
		const bool	dropIsolates = false,
		const int &	format = ReportFormat::Html )

Writes the Eccentricity report to file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeGeodesicDistribution()

bool Graph::writeGeodesicDistribution	(	const QString &	fileName,
		const bool &	considerWeights = false,
		const bool &	inverseWeights = false )

Writes a geodesic distance distribution report to an HTML file.

Computes the distribution via graphGeodesicDistanceDistribution() (cache-aware), then writes a sortable HTML table (distance | pair count | % | cumulative %) to fileName.

Parameters

fileName	Full path of the output HTML file.
considerWeights	Passed through to graphGeodesicDistanceDistribution().
inverseWeights	Passed through to graphGeodesicDistanceDistribution().

Returns

true on success; false if the file could not be opened.

writeMatrix()

bool Graph::writeMatrix	(	const QString &	fn,
		const int &	matrix = MATRIX_ADJACENCY,
		const bool &	considerWeights = true,
		const bool &	inverseWeights = false,
		const bool &	dropIsolates = false,
		const QString &	varLocation = "Rows",
		const bool &	simpler = false,
		const int &	format = ReportFormat::Html )

Computes and writes the specified matrix of the social network to an HTML report file.

Supported matrix types: adjacency, laplacian, degree, geodesic distances, shortest paths (geodesics), inverse adjacency, reachability, transpose, cocitation, and tie-profile distance matrices (Euclidean, Hamming, Jaccard, Manhattan, Chebyshev).

Parameters

fn	Path to the output HTML report file.
matrix	Matrix type constant (e.g., MATRIX_ADJACENCY, MATRIX_DISTANCES).
considerWeights	If true, edge weights are used in distance computations.
inverseWeights	If true, edge weights are inverted before use.
dropIsolates	If true, isolate nodes are excluded from the analysis.
varLocation	Whether variables are in rows or columns ("rows"/"cols"), used for tie-profile distance matrices.
simpler	Reserved for future use.

Returns

true on success, false if the computation was cancelled or an error occurred.

writeMatrixAdjacency()

bool Graph::writeMatrixAdjacency	(	const QString	fn,
		const bool &	markDiag = true,
		const int &	format = ReportFormat::Html )

Writes the adjacency matrix of G to a specified file fn

writeMatrixAdjacencyPlot()

void Graph::writeMatrixAdjacencyPlot	(	const QString	fn,
		const bool &	simpler = false )

Writes a visual representation of the adjacency matrix of the graph to the specified file.

The resulting matrix HAS NO spaces between elements.

Parameters

fn
simpler

writeMatrixAdjacencyTo()

void Graph::writeMatrixAdjacencyTo	(	QTextStream &	os,
		const bool &	saveEdgeWeights = true )

Exports the adjacency matrix to a given textstream

writeMatrixCSVTable()

void Graph::writeMatrixCSVTable	(	QTextStream &	outText,
		Matrix &	M,
		const bool &	printInfinity = true,
		const bool &	dropIsolates = false )

Writes the matrix M as a comma-separated table to the specified text stream outText.

CSV sibling of writeMatrixHTMLTable() (same vertex filtering, same RAND_MAX/large-magnitude/ precision value-formatting rules), but lean: header row of vertex numbers, one data row per vertex, no markup, no min/max summary paragraph - a CSV consumer (spreadsheet app) has no use for prose. Row/column headers are always plain vertex numbers, never labels, so no CSV quoting/escaping is needed here (unlike the per-node score tables planned for centrality/ prestige reports, where free-text labels will need it).

Parameters

outText
M
printInfinity
dropIsolates

writeMatrixDissimilarities()

bool Graph::writeMatrixDissimilarities	(	const QString	fileName,
		const QString &	metricStr,
		const QString &	varLocation,
		const bool &	diagonal,
		const bool &	considerWeights,
		const int &	format = ReportFormat::Html )

Writes dissimilarity matrix based on a metric/measure to given html file.

Parameters

fileName
measure
varLocation
diagonal
considerWeights

writeMatrixHTMLTable()

void Graph::writeMatrixHTMLTable	(	QTextStream &	outText,
		Matrix &	M,
		const bool &	markDiag = true,
		const bool &	plain = false,
		const bool &	printInfinity = true,
		const bool &	dropIsolates = false )

Writes the matrix M as HTML.

specified text stream outText It is the same as Matrix::printHTMLTable except that this method omits disabled vertices, thus the table header is correct

Parameters

outText
M
markDiag
plain
printInfinity

writeMatrixSimilarityMatching()

bool Graph::writeMatrixSimilarityMatching	(	const QString	fileName,
		const QString &	measure = "Simple",
		const QString &	matrix = "adjacency",
		const QString &	varLocation = "rows",
		const bool &	diagonal = false,
		const bool &	considerWeights = true,
		const int &	format = ReportFormat::Html )

Writes similarity matrix based on a matching measure to given html file.

Parameters

fileName
measure
matrix
varLocation
diagonal
considerWeights

writeMatrixSimilarityPearson()

bool Graph::writeMatrixSimilarityPearson	(	const QString	fileName,
		const bool	considerWeights,
		const QString &	matrix = "adjacency",
		const QString &	varLocation = "rows",
		const bool &	diagonal = false,
		const int &	format = ReportFormat::Html )

Calls Graph::createMatrixSimilarityPearson() and writes Pearson Correlation Coefficients to given file.

Parameters

fileName
considerWeights

writeMatrixWalks()

bool Graph::writeMatrixWalks	(	const QString &	fn,
		const int &	length = 0,
		const bool &	simpler = false,
		const int &	format = ReportFormat::Html )

Writes the walks of given length matrix to a file in HTML. If length = 0, it writes the Total Walks matrix.

Parameters

fn
length
simpler

writePrestigeDegree()

bool Graph::writePrestigeDegree	(	const QString	fileName,
		const bool	considerWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the Degree Prestige of each node to a file.

Parameters

fileName
considerWeights
dropIsolates

writePrestigePageRank()

bool Graph::writePrestigePageRank	(	const QString	fileName,
		const bool	dropIsolates = false,
		const int &	format = ReportFormat::Html )

Writes the PageRank scores of vertices to a file.

Parameters

fileName
dropIsolates

writePrestigeProximity()

bool Graph::writePrestigeProximity	(	const QString	fileName,
		const bool	considerWeights,
		const bool	inverseWeights,
		const bool	dropIsolates,
		const int &	format = ReportFormat::Html )

Writes the proximity prestige indices to a file.

Parameters

fileName
considerWeights
inverseWeights
dropIsolates

writeReciprocity()

bool Graph::writeReciprocity	(	const QString	fileName,
		const bool	considerWeights = false,
		const int &	format = ReportFormat::Html )

Writes reciprocity report to filename.

Parameters

fileName
considerWeights

writeScoreTableCSV()

void Graph::writeScoreTableCSV	(	QTextStream &	outText,
		const QStringList &	dataColumnHeaders,
		const std::function< QVector< qreal >(GraphVertex *)> &	rowValues,
		const std::function< bool(GraphVertex *)> &	isBlanked = nullptr,
		const std::function< bool(GraphVertex *)> &	isSkipped = nullptr )

CSV sibling of writeScoreTableHTML() - same column shape and semantics, comma-delimited, no HTML markup, same RAND_MAX-as-infinity convention. Reuses TableExport::csvQuote() for the Label column, the only free-text field (Node numbers and scores are always numeric, so they never need escaping).

Parameters

outText
dataColumnHeaders
rowValues
isBlanked
isSkipped

writeScoreTableHTML()

void Graph::writeScoreTableHTML	(	QTextStream &	outText,
		const QStringList &	dataColumnHeaders,
		const std::function< QVector< qreal >(GraphVertex *)> &	rowValues,
		const std::function< bool(GraphVertex *)> &	isBlanked = nullptr,
		const std::function< bool(GraphVertex *)> &	isSkipped = nullptr )

Renders a per-node score table as a sortable HTML.

shared by every centrality/prestige report writer (WS16 Step 2) plus Reciprocity/Clustering Coefficient/Eccentricity (WS16 Step 3). "Node" and "Label" are always the first two columns; dataColumnHeaders supplies the rest, in order, and rowValues must return that many values per row, in the same order - a value equal to RAND_MAX renders as the infinity glyph, matching writeMatrixHTMLTable()'s sentinel convention. isBlanked (optional) lets each report keep its own existing isolate-handling: rows for which it returns true get "--" placeholders instead of rowValues's output.

Parameters

outText
dataColumnHeaders
rowValues
isBlanked

writeTriadCensus()

bool Graph::writeTriadCensus	(	const QString	fileName,
		const bool	considerWeights,
		const int &	format = ReportFormat::Html )

Friends And Related Symbol Documentation

DistanceEngine

friend class DistanceEngine

friend

Member Data Documentation

actualDateTime

QDateTime Graph::actualDateTime

private

AM

Matrix Graph::AM

private

averageCLC

qreal Graph::averageCLC

private

calculatedAdjacencyMatrix

bool Graph::calculatedAdjacencyMatrix

private

calculatedBPC

bool Graph::calculatedBPC

private

calculatedCentralities

bool Graph::calculatedCentralities

private

calculatedDC

bool Graph::calculatedDC

private

calculatedDistances

bool Graph::calculatedDistances

private

calculatedDP

bool Graph::calculatedDP

private

calculatedEdges

bool Graph::calculatedEdges

private

calculatedEVC

bool Graph::calculatedEVC

private

calculatedGraphDensity

bool Graph::calculatedGraphDensity

private

calculatedGraphReciprocity

bool Graph::calculatedGraphReciprocity

private

calculatedGraphSymmetry

bool Graph::calculatedGraphSymmetry

private

calculatedGraphWeighted

bool Graph::calculatedGraphWeighted

private

calculatedIC

bool Graph::calculatedIC

private

calculatedIRCC

bool Graph::calculatedIRCC

private

calculatedIsolates

bool Graph::calculatedIsolates

private

calculatedKC

bool Graph::calculatedKC

private

calculatedPP

bool Graph::calculatedPP

private

calculatedPRP

bool Graph::calculatedPRP

private

calculatedTriad

bool Graph::calculatedTriad

private

calculatedVertices

bool Graph::calculatedVertices

private

calculatedVerticesList

bool Graph::calculatedVerticesList

private

calculatedVerticesSet

bool Graph::calculatedVerticesSet

private

canvasHeight

qreal Graph::canvasHeight

private

canvasWidth

qreal Graph::canvasWidth

private

classesBPC

int Graph::classesBPC

private

classesCLC

int Graph::classesCLC

private

classesEC

int Graph::classesEC

private

classesEccentricity

int Graph::classesEccentricity

private

classesEVC

int Graph::classesEVC

private

classesIC

int Graph::classesIC

private

classesIRCC

int Graph::classesIRCC

private

classesKC

int Graph::classesKC

private

classesPP

int Graph::classesPP

private

classesPRP

int Graph::classesPRP

private

classesSBC

int Graph::classesSBC

private

classesSCC

int Graph::classesSCC

private

classesSDC

int Graph::classesSDC

private

classesSDP

int Graph::classesSDP

private

classesSPC

int Graph::classesSPC

private

classesSSC

int Graph::classesSSC

private

CLQM

Matrix Graph::CLQM

private

csRecDepth

int Graph::csRecDepth

private

d_factor

qreal Graph::d_factor

private

denomEC

qreal Graph::denomEC

private

denomEVC

qreal Graph::denomEVC

private

denomIC

qreal Graph::denomIC

private

denomIRCC

qreal Graph::denomIRCC

private

denomPP

qreal Graph::denomPP

private

denomPRC

qreal Graph::denomPRC

private

denomSBC

qreal Graph::denomSBC

private

denomSCC

qreal Graph::denomSCC

private

denomSPC

qreal Graph::denomSPC

private

discreteBCs

H_StrToInt Graph::discreteBCs

private

discreteBPCs

H_StrToInt Graph::discreteBPCs

private

discreteCCs

H_StrToInt Graph::discreteCCs

private

discreteCLCs

H_StrToInt Graph::discreteCLCs

private

discreteDPs

H_StrToInt Graph::discreteDPs

private

used in resolveClasses and graphDistancesGeodesic()

discreteEccentricities

H_StrToInt Graph::discreteEccentricities

private

discreteECs

H_StrToInt Graph::discreteECs

private

discreteEVCs

H_StrToInt Graph::discreteEVCs

private

discreteICs

H_StrToInt Graph::discreteICs

private

discreteIRCCs

H_StrToInt Graph::discreteIRCCs

private

discreteKCs

H_StrToInt Graph::discreteKCs

private

discretePCs

H_StrToInt Graph::discretePCs

private

discretePPs

H_StrToInt Graph::discretePPs

private

discretePRPs

H_StrToInt Graph::discretePRPs

private

discreteSCs

H_StrToInt Graph::discreteSCs

private

discreteSDCs

H_StrToInt Graph::discreteSDCs

private

DM

Matrix Graph::DM

private

edgeReverseWeightTemp

qreal Graph::edgeReverseWeightTemp

private

edgeWeightTemp

qreal Graph::edgeWeightTemp

private

file_parser

Parser* Graph::file_parser

private

file_parserThread

QThread Graph::file_parserThread

private

groupCC

qreal Graph::groupCC

private

groupDC

qreal Graph::groupDC

private

groupDP

qreal Graph::groupDP

private

groupEC

qreal Graph::groupEC

private

groupEVC

qreal Graph::groupEVC

private

groupIRCC

qreal Graph::groupIRCC

private

groupPP

qreal Graph::groupPP

private

groupSBC

qreal Graph::groupSBC

private

groupSC

qreal Graph::groupSC

private

groupSPC

qreal Graph::groupSPC

private

htmlEnd

QString Graph::htmlEnd

private

htmlHead

QString Graph::htmlHead

private

htmlHeadLight

QString Graph::htmlHeadLight

private

inboundEdgesVert

int Graph::inboundEdgesVert

private

initEdgeColor

QString Graph::initEdgeColor

private

initEdgeColorZero

QString Graph::initEdgeColorZero

private

initEdgeLabels

bool Graph::initEdgeLabels

private

initEdgeWeightNumbers

bool Graph::initEdgeWeightNumbers

private

initShowZeroWeightEdges

bool Graph::initShowZeroWeightEdges

private

initVertexColor

QString Graph::initVertexColor

private

initVertexIconPath

QString Graph::initVertexIconPath

private

initVertexLabelColor

QString Graph::initVertexLabelColor

private

initVertexLabelDistance

int Graph::initVertexLabelDistance

private

initVertexLabelSize

int Graph::initVertexLabelSize

private

initVertexNumberColor

QString Graph::initVertexNumberColor

private

initVertexNumberDistance

int Graph::initVertexNumberDistance

private

initVertexNumberSize

int Graph::initVertexNumberSize

private

initVertexShape

QString Graph::initVertexShape

private

initVertexSize

int Graph::initVertexSize

private

invAM

Matrix Graph::invAM

private

invM

Matrix Graph::invM

private

m_apspDist

QHash<int, Matrix> Graph::m_apspDist

private

m_apspSigma

QHash<int, Matrix> Graph::m_apspSigma

private

m_clickedEdge

MyEdge Graph::m_clickedEdge

private

m_cliques

QMultiMap<int, L_int> Graph::m_cliques

private

m_clusteringLevel

QList<qreal> Graph::m_clusteringLevel

private

m_clusterPairNamesPerSeq

QMap<int, V_str> Graph::m_clusterPairNamesPerSeq

private

m_clustersByName

QMap<QString, V_int> Graph::m_clustersByName

private

m_clustersPerSequence

QMap<int, V_int> Graph::m_clustersPerSequence

private

m_crawler_max_urls

int Graph::m_crawler_max_urls

private

m_crawler_visited_urls

int Graph::m_crawler_visited_urls

private

m_curRelation

int Graph::m_curRelation

private

m_fieldWidth

int Graph::m_fieldWidth

private

m_fileFormat

int Graph::m_fileFormat

private

m_fileName

QString Graph::m_fileName

private

m_graph

VList Graph::m_graph

private

m_graphAverageDistance

qreal Graph::m_graphAverageDistance

private

m_graphDensity

qreal Graph::m_graphDensity

private

m_graphDiameter

int Graph::m_graphDiameter

private

m_graphFileFormatExportSupported

QList<int> Graph::m_graphFileFormatExportSupported

private

m_graphGeodesicsCount

qreal Graph::m_graphGeodesicsCount

private

m_graphHasVertexCustomIcons

bool Graph::m_graphHasVertexCustomIcons

private

m_graphIsConnected

bool Graph::m_graphIsConnected

private

m_graphIsDirected

bool Graph::m_graphIsDirected

private

m_graphIsSymmetric

bool Graph::m_graphIsSymmetric

private

m_graphIsWeighted

bool Graph::m_graphIsWeighted

private

m_graphModStatus

int Graph::m_graphModStatus

private

General & initialisation variables

m_graphName

QString Graph::m_graphName

private

m_graphReciprocityArc

qreal Graph::m_graphReciprocityArc

private

m_graphReciprocityDyad

qreal Graph::m_graphReciprocityDyad

private

m_graphReciprocityPairsReciprocated

int Graph::m_graphReciprocityPairsReciprocated

private

m_graphReciprocityPairsTotal

int Graph::m_graphReciprocityPairsTotal

private

m_graphReciprocityTiesNonSymmetric

int Graph::m_graphReciprocityTiesNonSymmetric

private

m_graphReciprocityTiesReciprocated

int Graph::m_graphReciprocityTiesReciprocated

private

m_graphReciprocityTiesTotal

int Graph::m_graphReciprocityTiesTotal

private

m_graphStronglyConnectedComponents

int Graph::m_graphStronglyConnectedComponents

private

m_graphSumDistance

qreal Graph::m_graphSumDistance

private

m_graphWeaklyConnectedComponents

int Graph::m_graphWeaklyConnectedComponents

private

m_lastBonacichAlpha

qreal Graph::m_lastBonacichAlpha = -1

private

Same caching purpose as m_lastKatzAlpha (WS11, #39); -1 means "never computed this session". Alpha is kept positive-only by the dialog so this sentinel is safe - beta (which can be negative) has no sentinel role.

m_lastBonacichBeta

qreal Graph::m_lastBonacichBeta = 0

private

m_lastKatzAlpha

qreal Graph::m_lastKatzAlpha = -1

private

Cache of the last alpha used to compute Katz Centrality, read by layoutByProminenceIndex() (WS11, #10) since the generic layout-by-prominence dispatch has no parameter slot. -1 means "never computed this session".

m_progressCanceled

std::atomic<bool> Graph::m_progressCanceled

private

m_relationsDirected

QList<bool> Graph::m_relationsDirected

private

m_relationsList

QList<QString> Graph::m_relationsList

private

m_reportsChartType

ChartType Graph::m_reportsChartType

private

m_reportsDataDir

QString Graph::m_reportsDataDir

private

m_reportsLabelLength

int Graph::m_reportsLabelLength

private

m_reportsOutputFormat

ReportFormat Graph::m_reportsOutputFormat

private

m_reportsRealPrecision

int Graph::m_reportsRealPrecision

private

m_reserveEdgesPerVertexSize

int Graph::m_reserveEdgesPerVertexSize

private

m_selectedEdges

QList<SelectedEdge> Graph::m_selectedEdges

private

m_totalEdges

int Graph::m_totalEdges

private

m_totalVertices

int Graph::m_totalVertices

private

m_vertexClicked

int Graph::m_vertexClicked

private

m_vertexComponentId

QHash<int,int> Graph::m_vertexComponentId

private

m_vertexPairsNotConnected

QMultiHash<int, int> Graph::m_vertexPairsNotConnected

private

m_vertexPairsUnilaterallyConnected

QHash<int, int> Graph::m_vertexPairsUnilaterallyConnected

private

m_verticesCacheCountAll

bool Graph::m_verticesCacheCountAll = false

private

m_verticesCacheDropIsolates

bool Graph::m_verticesCacheDropIsolates = false

private

m_verticesIsolatedList

QList<int> Graph::m_verticesIsolatedList

private

m_verticesList

QList<int> Graph::m_verticesList

private

m_verticesSelected

QList<int> Graph::m_verticesSelected

private

m_verticesSet

QSet<int> Graph::m_verticesSet

private

m_visibilityHistory

QStack<GraphVisibilitySnapshot> Graph::m_visibilityHistory

private

maxBPC

qreal Graph::maxBPC

private

maxCLC

qreal Graph::maxCLC

private

maxEC

qreal Graph::maxEC

private

maxEccentricity

qreal Graph::maxEccentricity

private

maxEVC

qreal Graph::maxEVC

private

maxIC

qreal Graph::maxIC

private

maxIndexBC

qreal Graph::maxIndexBC

private

maxIndexCC

qreal Graph::maxIndexCC

private

maxIndexEC

qreal Graph::maxIndexEC

private

maxIndexIC

qreal Graph::maxIndexIC

private

maxIndexPC

qreal Graph::maxIndexPC

private

maxIndexSC

qreal Graph::maxIndexSC

private

maxIRCC

qreal Graph::maxIRCC

private

maxKC

qreal Graph::maxKC

private

maxNodeBPC

int Graph::maxNodeBPC

private

maxNodeCLC

int Graph::maxNodeCLC

private

maxNodeDP

int Graph::maxNodeDP

private

maxNodeEC

int Graph::maxNodeEC

private

maxNodeEccentricity

int Graph::maxNodeEccentricity

private

maxNodeEVC

int Graph::maxNodeEVC

private

maxNodeIC

int Graph::maxNodeIC

private

maxNodeIRCC

int Graph::maxNodeIRCC

private

maxNodeKC

int Graph::maxNodeKC

private

maxNodePP

int Graph::maxNodePP

private

maxNodePRP

int Graph::maxNodePRP

private

maxNodeSBC

int Graph::maxNodeSBC

private

maxNodeSCC

int Graph::maxNodeSCC

private

maxNodeSDC

int Graph::maxNodeSDC

private

maxNodeSPC

int Graph::maxNodeSPC

private

maxNodeSSC

int Graph::maxNodeSSC

private

maxPP

qreal Graph::maxPP

private

maxPRP

qreal Graph::maxPRP

private

maxSBC

qreal Graph::maxSBC

private

maxSCC

qreal Graph::maxSCC

private

maxSDC

qreal Graph::maxSDC

private

maxSDP

qreal Graph::maxSDP

private

maxSPC

qreal Graph::maxSPC

private

maxSSC

qreal Graph::maxSSC

private

meanBPC

qreal Graph::meanBPC

private

meanEC

qreal Graph::meanEC

private

meanEVC

qreal Graph::meanEVC

private

meanIC

qreal Graph::meanIC

private

meanIRCC

qreal Graph::meanIRCC

private

meanKC

qreal Graph::meanKC

private

meanPP

qreal Graph::meanPP

private

meanPRP

qreal Graph::meanPRP

private

meanSBC

qreal Graph::meanSBC

private

meanSCC

qreal Graph::meanSCC

private

meanSDC

qreal Graph::meanSDC

private

meanSDP

qreal Graph::meanSDP

private

meanSPC

qreal Graph::meanSPC

private

meanSSC

qreal Graph::meanSSC

private

minBPC

qreal Graph::minBPC

private

minCLC

qreal Graph::minCLC

private

minEC

qreal Graph::minEC

private

minEccentricity

qreal Graph::minEccentricity

private

minEVC

qreal Graph::minEVC

private

minIC

qreal Graph::minIC

private

minIRCC

qreal Graph::minIRCC

private

minKC

qreal Graph::minKC

private

minNodeBPC

int Graph::minNodeBPC

private

minNodeCLC

int Graph::minNodeCLC

private

minNodeDP

int Graph::minNodeDP

private

minNodeEC

int Graph::minNodeEC

private

minNodeEccentricity

int Graph::minNodeEccentricity

private

minNodeEVC

int Graph::minNodeEVC

private

minNodeIC

int Graph::minNodeIC

private

minNodeIRCC

int Graph::minNodeIRCC

private

minNodeKC

int Graph::minNodeKC

private

minNodePP

int Graph::minNodePP

private

minNodePRP

int Graph::minNodePRP

private

minNodeSBC

int Graph::minNodeSBC

private

minNodeSCC

int Graph::minNodeSCC

private

minNodeSDC

int Graph::minNodeSDC

private

minNodeSPC

int Graph::minNodeSPC

private

minNodeSSC

int Graph::minNodeSSC

private

minPP

qreal Graph::minPP

private

minPRP

qreal Graph::minPRP

private

minSBC

qreal Graph::minSBC

private

minSCC

qreal Graph::minSCC

private

minSDC

qreal Graph::minSDC

private

minSDP

qreal Graph::minSDP

private

minSPC

qreal Graph::minSPC

private

minSSC

qreal Graph::minSSC

private

neighboursHash

QHash<int, QSet<int> > Graph::neighboursHash

private

nomEC

qreal Graph::nomEC

private

nomEVC

qreal Graph::nomEVC

private

nomIC

qreal Graph::nomIC

private

nomIRCC

qreal Graph::nomIRCC

private

nomPP

qreal Graph::nomPP

private

nomPRC

qreal Graph::nomPRC

private

nomSBC

qreal Graph::nomSBC

private

nomSCC

qreal Graph::nomSCC

private

nomSPC

qreal Graph::nomSPC

private

order

bool Graph::order

private

outboundEdgesVert

int Graph::outboundEdgesVert

private

reciprocalEdgesVert

int Graph::reciprocalEdgesVert

private

SIGMA

Matrix Graph::SIGMA

private

sumBC

qreal Graph::sumBC

private

sumBPC

qreal Graph::sumBPC

private

sumCC

qreal Graph::sumCC

private

sumDC

qreal Graph::sumDC

private

sumDP

qreal Graph::sumDP

private

sumEC

qreal Graph::sumEC

private

sumEVC

qreal Graph::sumEVC

private

sumIC

qreal Graph::sumIC

private

sumIRCC

qreal Graph::sumIRCC

private

sumKC

qreal Graph::sumKC

private

sumPC

qreal Graph::sumPC

private

sumPP

qreal Graph::sumPP

private

sumPRP

qreal Graph::sumPRP

private

sumSBC

qreal Graph::sumSBC

private

sumSC

qreal Graph::sumSC

private

sumSCC

qreal Graph::sumSCC

private

sumSDC

qreal Graph::sumSDC

private

sumSDP

qreal Graph::sumSDP

private

sumSEVC

qreal Graph::sumSEVC

private

sumSPC

qreal Graph::sumSPC

private

sumSSC

qreal Graph::sumSSC

private

t_sumIC

qreal Graph::t_sumIC

private

t_sumPRP

qreal Graph::t_sumPRP

private

triadTypeFreqs

QList<int> Graph::triadTypeFreqs

private

urlQueue

QQueue<QUrl>* Graph::urlQueue

private

varianceBPC

qreal Graph::varianceBPC

private

varianceCLC

qreal Graph::varianceCLC

private

varianceEC

qreal Graph::varianceEC

private

varianceEVC

qreal Graph::varianceEVC

private

varianceIC

qreal Graph::varianceIC

private

varianceIRCC

qreal Graph::varianceIRCC

private

varianceKC

qreal Graph::varianceKC

private

variancePP

qreal Graph::variancePP

private

variancePRP

qreal Graph::variancePRP

private

varianceSBC

qreal Graph::varianceSBC

private

varianceSCC

qreal Graph::varianceSCC

private

varianceSDC

qreal Graph::varianceSDC

private

varianceSDP

qreal Graph::varianceSDP

private

varianceSPC

qreal Graph::varianceSPC

private

varianceSSC

qreal Graph::varianceSSC

private

vpos

H_Int Graph::vpos

vpos stores the real position of each vertex inside m_graph. It starts at zero (0). We need to know the place of a vertex inside m_graph after adding or removing many vertices

web_crawler

WebCrawler* Graph::web_crawler

private

webcrawlerThread

QThread Graph::webcrawlerThread

private

WM

Matrix Graph::WM

private

XM

Matrix Graph::XM

private

XRM

Matrix Graph::XRM

private

XSM

Matrix Graph::XSM

private

---

The documentation for this class was generated from the following files:

• graph.h
• graph.cpp
• graph_centrality.cpp
• graph_centrality_bonacich.cpp
• graph_centrality_katz.cpp
• graph_prestige.cpp
• graph_clustering_coefficients.cpp
• graph_clustering_hierarchical.cpp
• graph_triad_census.cpp
• graph_cliques.cpp
• graph_connectivity.cpp
• graph_metadata.cpp
• graph_state_flags.cpp
• graph_structure_metrics.cpp
• graph_crawler.cpp
• graph_distance_cache.cpp
• graph_distance_facade.cpp
• graph_edge_filters.cpp
• graph_node_filters.cpp
• graph_query_filters.cpp
• graph_random_networks.cpp
• graph_io.cpp
• graph_layouts_basic.cpp
• graph_layouts_force.cpp
• graph_matrix_adjacency.cpp
• graph_matrix_distances.cpp
• graph_matrix_reachability.cpp
• graph_prominence_distribution.cpp
• graph_reachability_walks.cpp
• graph_relations.cpp
• graph_reports.cpp
• graph_reports_settings.cpp
• graph_similarity_matrices.cpp
• graph_edges.cpp
• graph_subgraphs.cpp
• graph_vertices.cpp
• graph_canvas.cpp
• graph_edge_style.cpp
• graph_selection.cpp
• graph_ui_facade.cpp
• graph_ui_prominence_distribution.cpp
• graph_vertex_style.cpp
• graph_type_strings.cpp