nuget
Created by freemso on 17-3-14.
Graph
01 Syntax
Microsoft.VisualBasic.MachineLearning.Bootstrapping.node2vec.Graph
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| preprocess | 1 | pre-processing of transition probabilities for guiding the random walks |
| computeAliasEdge | 1 | to compute the alias method for an edge |
| walk | 1 | random walk in the graph starting from a node |
| simulateWalks | 1 | simulation of a bunch of walks |
| sortedNeighborList | 1 | get a node's neighbors in a sorted list the set of the neighbors of node is defined as {x|node-->x} sort the nodes according to its ids |
| hasEdge | 1 | check whether there is an edge between two nodes note that all the edges in the graph are directive |
| getEdge | 1 | get the edge between two nodes |
| addEdge | 1 | add a new edge to the graph if such an edge already exists, update the weight note that all the edges in the graph are directed if the graph is not directed, we just simply add… |
| addNode | 1 | add a node with the id to the graph if such a node already exists, return it and do nothing if not, create a new node, add it to the graph and return it |
03 Members
preprocess
pre-processing of transition probabilities for guiding the random walks
computeAliasEdge(
VertexEdge)to compute the alias method for an edge
Parameters
| Name | Type | Description |
|---|---|---|
edge | VertexEdge | the edge to compute |
Returns
the node2vec.AliasMethod object that store distribution information
walk(
Int32, Vertex)random walk in the graph starting from a node
Parameters
| Name | Type | Description |
|---|---|---|
walkLength | Int32 | the steps of this walk |
startNode | Vertex | the start node of this walk |
Returns
the path that we pass, expressed as a Node List
simulateWalks(
Int32, Int32)simulation of a bunch of walks
Parameters
| Name | Type | Description |
|---|---|---|
numWalks | Int32 | iteration times |
walkLength | Int32 | steps of every walk |
Returns
the list of paths that we've walked
sortedNeighborList(
Vertex)get a node's neighbors in a sorted list the set of the neighbors of node is defined as {x|node-->x} sort the nodes according to its ids
Parameters
| Name | Type | Description |
|---|---|---|
node | Vertex | the node |
Returns
a sorted list of nodes
hasEdge(
Vertex, Vertex)check whether there is an edge between two nodes note that all the edges in the graph are directive
Parameters
| Name | Type | Description |
|---|---|---|
src | Vertex | node1 |
dst | Vertex | node2 |
Returns
true is there is an edge
getEdge(
Vertex, Vertex)get the edge between two nodes
Parameters
| Name | Type | Description |
|---|---|---|
src | Vertex | node1 |
dst | Vertex | node2 |
Returns
the edge, null is not exist such an edge
addEdge(
Vertex, Vertex, Double)add a new edge to the graph if such an edge already exists, update the weight note that all the edges in the graph are directed if the graph is not directed, we just simply add two directed edges with the opposite directions that connect two nodes
Parameters
| Name | Type | Description |
|---|---|---|
src | Vertex | first node of the edge |
dst | Vertex | second node of the edge |
weight | Double | of the edge |
addNode(
Int32, String)add a node with the id to the graph if such a node already exists, return it and do nothing if not, create a new node, add it to the graph and return it
Parameters
| Name | Type | Description |
|---|---|---|
id | Int32 | the id of the node |
Returns
the node found