HAPI.Node Class Module | Startpage |
HAPI | HAPI.Node |
Alpha(lIndex As Long) [[As Double]] (RW)
- The alpha value for one configuration of the discrete parents (continuous nodes only).
Belief(lStateIndex As Long) [[As Single]] (R)
- Returns the calculated probability of a state of a discrete chance node (use Propagate to update probabilities).
Beta(ndCGParent As Node, lIndex As Long) [[As Double]] (RW)
- The beta value for one configuration of the discrete parents (continuous nodes only).
CaseIsSet(lIndex As Long) [[As Boolean]] (R)
- Test whether the value of this node in case index currently is set.
CaseState(lIndex As Long) [[As Long]] (RW)
- Retrieve the state value of this node associated with the case index.
CaseValue(lIndex As Long) [[As Double]] (RW)
- Retrieve the value of this node assoociated with case lIndex.
Category() [[As hCategories]] (R)
- Holds the "category" (hCategoryChance, hCategoryDecision, or hCategoryUtility) of a node.
Children() [[As Collection]] (R)
- Holds a collection of all the children nodes of a node.
Distribution() [[As Table]] (R)
- Returns the Table object representing the distribution function of a continuous node.
Domain() [[As Domain]] (R)
- Holds the Domain object in which this node is created.
Entropy() [[As Double]] (R)
- Returns the entropy of this node
EvidenceIsEntered() [[As Boolean]] (R)
- True, if non-trivial evidence has been entered. Otherwise, False.
EvidenceIsPropagated() [[As Boolean]] (R)
- True, if evidence has been entered and propagated for the node. Otherwise, False.
EvidenceToPropagate() [[As Boolean]] (R)
- True, if evidence has been entered and has not been propagated or if no evidence has been entered. Otherwise, False.
ExpectedUtility(lStateIndex As Long) [[As Single]] (R)
- Returns the calculated expected utility of a state of a decision node (use Propagate to update utilities).
ExperienceTable() [[As Table]] (R)
- Return the experience table of the node.
FadingTable() [[As Table]] (R)
- Return the fading table of the node.
Finding(lStateIndex As Long) [[As Single]] (RW)
- Holds entered evidence for one state of a discrete node.
Gamma(lIndex As Long) [[As Double]] (RW)
- The gamma value for one configuration of the discrete parents (continuous nodes only).
GetSampledValue() [[As Double]] (R)
- Returns the value of the current object within the configuration generated by the most recent call to Domain.simulate().
GetSamplesState() [[As Integer]] (R)
- Returns the state index of the discrete chance node within the configuration generated by the most recent call to Domain.simulate().
HasExperienceTable() [[As Boolean]] (R)
- Checks whether or not the node has an experience table.
HasFadingTable() [[As Boolean]] (R)
- Checks whether or not the node has an experience table.
HasUserAttribute(strKey As String) [[As Boolean]] (R)
- Determines whether a user attribute is set on the Node object (see UserAttribute).
HomeClass() [[As Class]] (R)
- The class in which this node is placed.
Instance() [[As Node]] (R)
- Returns the instance which this output clone originates from.
InstanceClass() [[As Class]] (R)
- The class from which this instance node is derived.
JunctionTree() [[As JunctionTree]] (R)
- The JunctionTree object that this node is part of (if the corresponding Domain object has been compiled).
Kind() [[As hKinds]] (R)
- Holds the "kind" (hKindContinuous or hKindDiscrete) of a node as specified in the lKind argument of GetNewNode.
Label() [[As String]] (RW)
- Allows an application to access a more readable label than Name. This label can contain any string.
LikelihoodIsEntered() [[As Boolean]] (R)
- True, if non-deterministic findings have been entered (Finding(i)=x where 0<x<1 for some state i). Otherwise, False.
LikelihoodIsPropagated() [[As Boolean]] (R)
- True, if non-deterministic findings (likelihood) have been entered and propagated. Otherwise, False.
Master() [[As Node]] (R)
- Returns the "master" of this (cloned) output Node of an instance node.
Mean() [[As Double]] (R)
- Returns the calculated mean of the distribution function of a continuous node (use Propagate to update distributions).
MutualInformation(nd As Node) [[As Double]] (R)
- Returns the mutual information between this node and the specified node.
NetworkModel() [[As NetworkModel]] (R)
- Return the NetworkModel which this Node belongs to.
NumberOfStates() [[As Long]] (RW)
- Holds the number of states of a discrete chance node or a decision node.
PropagatedFinding(lStateIndex As Long) [[As Single]] (R)
- The finding value registered on a state of a discrete node at the last call of Propagate.
PropagatedValue() [[As Double]] (R)
- Returns the value that has been propagated for a continuous node. If no value has been propagated, an error is raised.
Selection() [[As Long]] (R)
- Get the state index randomly selected during the last call of Simulate (discrete nodes only).
StateIndexFromLabel(strState As String) [[As Long]] (R)
- The StateIndexFromLabel property gets the state index of a node state (discrete nodes only) with the given label. If the label is not found -1 is returned
StateIndexFromValue(dVal As Double) [[As Long]] (R)
- The StateIndexFromLabel property gets the state index of a node state (discrete nodes only) equal to or containing the given value. If the value is not found -1 is returned
StateLabel(lStateIndex As Integer) [[As String]] (RW)
- The StateLabel property sets/gets the state label of a node state (discrete nodes only).
StateValue(lState As Long) [[As Double]] (RW)
- Associate a value with a state of this node (which must be anumeric node). It is required that the state values form an increasing sequence (this is checked when the table is generated for this node).
For numbered nodes, dValue indicates the specific number to be associate with the specified state.
For interval nodes, the values specified for states i and i+1 are the left and right endpoints of the interval denoted by state i. To indicate the endpoint of the right-most interval, specify lState equal to the number of states of this node.
SubType() [[As hSubtypes]] (RW)
- Set the subtype of this node (which must be a discrete chance or decision node).
If stSubType is hSubtypeBoolean, the node must have exactly two states. Moreover, when the node has subtype 'boolean', the number of states in the node cannot be changed.
The default subtype for a node is hSubtypeLabel.
Table() [[As Table]] (R)
- Holds a Table object representing the conditional probability function of chance nodes and the utility function of utility nodes.
UserAttribute(strKey As String) [[As String]] (RW)
- Holds the user attribute value for a given key.
UserData() [[As Object]] (RW)
- Store the user controlled object reference associated with this node.
The HUGIN API provides as slot within the Node object exclusively for the use of the user application. This can be used for associating arbitrary user data (such as references to windows, tables, databases, etc.), and will not be handled by HUGIN. In particular, this means that the user is responsible for storing the needed (or desired) information - HUGIN will not attempt to do so!
Value() [[As Double]] (RW)
- Holds the entered evidence of a continuous node.
Variance() [[As Double]] (R)
- Returns the calculated variance of the distribution function of a continuous node (use Propagate to update distributions).
getEdgeConstraint(B As Node) [[As hEdgeConstraints]] (R)
- Retrieve the learning constraints specified for the edge a - b.
name() [[As String]] (RW)
- Holds the unique name of a Node object.
parents() [[As Collection]] (R)
- Holds a collection of all the parent nodes of a node.
AddParent(ndParent As Node)
- Adds the specified node as a parent of this node.
AddToInputs()
- Adds this Node to the list of input nodes in the Class/Domain of this node.
AddToOutputs()
- Adds this Node to the list of output nodes in the Class/Domain of this node.
ComputeSensitivityData(lIndex As Long)
- Computes the constants of the sensitivity functions for the specified output probability and all CPT parameters in the network.
Delete()
- Delete a Node from the current Domain object.
GenerateTable()
- Generates the conditional probability table of this from its model
GetPosition(lX As Long, lY As Long)
- Returns position of the Node object through reference parameters.
GetSensitivityConstants(lIndex As Long, lA as Single, lB As Single, lC As Single, lD As Single)
- Computes the four constants of the specified sensitivity function.
GetSensitivityConstantsByOutput(lIndex As Long, lOutput As Long, lA as Single, lB As Single, lC As Single, lD As Single)
- Computes the four constants of the specified sensitivity function.
RemoveFromInputs()
- Removes this Node from the list of input nodes in the Class/Domain of this node.
RemoveFromOutputs()
- Removes this Node from the list of output nodes in the Class/Domain of this node.
RemoveParent(ndParent As Node)
- Remove a parent node.
RemoveUserAttribute(strKey As String)
- Removes the specified user attribute.
RetractFindings()
- "Unenter" all entered findings on this node.
ReverseEdge(nd As Node)
SelectState(lStateIndex As Long)
- Enters evidence that a node is in a certain state.
SetInput(hInput As Node, hNode As Node)
- setInput binds a Node in a Domain to an input node in an instance of a Class in the same Domain.
SetPosition(lX As Long, lY As Long)
- Set the position of the node.
SubstituteClass(newClass As Class)
- Substitute the class of this node with the new class given as argument.
SwitchParent(ndOldParent As Node, ndNewParent As Node)
- Substitutes the first node specified with the second node specified as a parent of this node. Substitute a new parent for an old parent. The new parent must be compatible with the old parent: This implies that it must be of the same class, have the same number of states, etc.
UnsetCase(lIndex As Long)
- Specify that the value of this Node for case lIndex is unknown.
UnsetInput(hInput As Node)
- unsetInput removes a binding from a Node in a Domain to the input node.
setEdgeConstraint(B As Node, C As hEdgeConstraints)
- Specify constraint as the learning constraint for the edge a-b.
Clone () As Node
- Returns a clone of this node
GetAttributes() As Collection
- Returns the list of attributes associated with this.
GetInput(hNode As Node) As Node
- GetInput returns the input node, which has previously been set by calling SetInput
GetModel() As Model
- Retrieve the model for the table of this node.
GetOutput(hNode As Node) As Node
- GetOutput returns the output node, which has previously been set by calling SetOutput
GetSource() As NodeList
- Returns a NodeList of Class nodes that identifies this Domain node.
InitRelationships() As Boolean
- Initializes parental relationships of the Node object when a Domain object is being loaded.
NewModel(colNodes As Collection) As Model
- Create and return a new model for this node (which must be a utility or discrete chance node) using colNodes (comprising a subset of the parents of this node) to define the configurations of the model.
If the node already have a model associated with it, the existing model will be deleted before the new model is installed.
When a model exists for a node, it overrides the normal table associated with the node. Instead, the table data will be generated automatically using the model when the domain to which the node belongs is compiled.
Discrete nodes belongs to exactly one of four hSubtypes, namely Labelled, Numbered, Boolean, and Interval.
A labelled node is the "normal" discrete node where each state is associated with a name. A boolean node consists of exactly two labeled states, True and False. A numbered node has a real value associated with each state, and this value can be used in numeric expressions. The interval node contains states that each represent an interval og real values. The intervals constitute a continuous interval from the lower limit of the first state to the upper limit of the last state.
![]() |
HUGIN Expert A/S, 2008 - comments to activex@hugin.com |