1
0
Fork 0
arangodb/Documentation/Books/Users/General-Graphs/GeneralGraphFunctions.mdpp

383 lines
10 KiB
Plaintext

!CHAPTER Graph Module
The graph module provides functions dealing with graph structures.
!SECTION First Steps with Graphs
A Graph consists of *vertices* and *edges*. Edges are stored as documents in *edge
collections*. A vertex can be a document of a *document collection* or of an edge
collection (so edges can be used as vertices). Wich collections are used within
a graph is defined via *edge definitions*. A graph can contain more than one edge
definition, at least one is needed.
!SUBSECTION Create a graph
The creation of a graph requires the name of the graph and a definition of its edges.
For every type of edge definition a convenience method exists that can be used to create a graph.
```js
> var graph = require("org/arangodb/graph");
> var g = graph._create(graphName, edgeDefinitions);
```
There are different types of edge defintions:
!SUBSECTION Edge Definitions
The edge definitions for a graph is an Array containing arbitrary many directed and/or undirected relations as defined below.
The edge definitons array is initialised with the following call:
```js
> var edgeDefinitions = graph._edgeDefinitions(edgeDefinition1,......edgeDefinitionN);
```
To add further edge definitions to the array one must call:
```js
> graph._extendEdgeDefinitions(edgeDefinitions, edgeDefinition1,......edgeDefinitionN);
```
!SUBSUBSECTION Undirected Relation
<!-- @startDocuBlock JSF_general_graph_undirectedRelationDefinition -->
!SUBSUBSECTION Directed Relation
<!-- @startDocuBlock JSF_general_graph_undirectedRelationDefinition -->
!SUBSUBSECTION Complete Example to create a graph
Example Call:
```js
> var graph = require("org/arangodb/graph");
> var edgeDefinitions = graph._edgeDefinitions();
> graph._extendEdgeDefinitions(edgeDefinitions, graph._undirectedRelationDefinition("friend_of", ["Customer"]));
> graph._extendEdgeDefinitions(edgeDefinitions, graph._directedRelationDefinition("has_bought", ["Customer", "Company"], ["Groceries", "Electronics"]));
> graph._create("myStore", edgeDefinitions);
{
_id: "_graphs/123",
_rev: "123",
_key: "123"
}
```
alternative call:
```js
> var graph = require("org/arangodb/graph");
> var edgeDefinitions = graph._edgeDefinitions(graph._undirectedRelationDefinition("friend_of", ["Customer"]), graph._directedRelationDefinition("has_bought", ["Customer", "Company"], ["Groceries", "Electronics"]));
> graph._create("myStore", edgeDefinitions);
{
_id: "_graphs/123",
_rev: "123",
_key: "123"
};
```
!SUBSECTION Orphan Collections
Each graph has an orphan collection. It consists of arbitrary many vertex collection (type *document*), that are not
used in an edge definition of the graph. If the graph is extended with an edge definition, which is part of the orphan
collection, it will be removed from the orphan collection automatically.
!SUBSUBSECTION Add
<!-- @startDocuBlock JSF_general_graph__addOrphanCollection -->
!SUBSUBSECTION Read
<!-- @startDocuBlock JSF_general_graph__getOrphanCollections -->
!SUBSUBSECTION Remove
<!-- @startDocuBlock JSF_general_graph__removeOrphanCollection -->
!SUBSECTION Read a graph
```js
> var graph = require("org/arangodb/graph");
> var g = graph._graph("myStore");
```
- - -
!SUBSECTION Remove a graph
Removes a graph from the collection *\_graphs*.
```js
> graph._drop(graphId, dropCollections);
true
```
graphId: string - id of the graph to be removed
dropCollections: bool - optional. *true* all collections of the graph will be deleted.
*false* no collection of the graph will be deleted. Default: *true*
!SECTION Edge
!SUBSECTION Save
<!-- @startDocuBlock JSF_general_graph_vertex_collection_save -->
!SUBSECTION Replace
<!-- @startDocuBlock JSF_general_graph_vertex_collection_replace -->
!SUBSECTION Update
<!-- @startDocuBlock JSF_general_graph_vertex_collection_update -->
!SUBSECTION Remove
<!-- @startDocuBlock JSF_general_graph_vertex_collection_remove -->
!SECTION Edge
!SUBSECTION Save
<!-- @startDocuBlock JSF_general_graph_edge_collection_save -->
!SUBSECTION Replace
<!-- @startDocuBlock JSF_general_graph_edge_collection_replace -->
!SUBSECTION Update
<!-- @startDocuBlock JSF_general_graph_edge_collection_update -->
!SUBSECTION Remove
<!-- @startDocuBlock JSF_general_graph_edge_collection_remove -->
!SECTION Vertices
!SUBSECTION Get vertex *from* of an edge
<!-- @startDocuBlock JSF_general_graph_getFromVertex -->
!SUBSECTION Get vertex *to* of an edge
<!-- @startDocuBlock JSF_general_graph_getToVertex -->
!SECTION Some Methods
```javascript
graph.listCommonNeighbors(vertex1, vertex2, options)
```
vertex1: string - vertex id
vertex2: string - vertex id
options:
* see getNeighbors
```javascript
graph.amountCommonNeighbors(vertex1, vertex2, options)
```
vertex1: string - vertex id
vertex2: string - vertex id
options:
* see getNeighbors
```javascript
graph.listCommonProperties((vertex1, vertex2)
```
vertex1: string - vertex id
vertex2: string - vertex id
```javascript
graph.amountCommonProperties((vertex1, vertex2)
```
vertex1: string - vertex id
vertex2: string - vertex id
```javascript
graph.pathTo(vertex1, vertex2, options)
```
vertex1: string - vertex id
vertex2: string - vertex id
options: see determinePredecessors
```javascript
graph.distanceTo(vertex1, vertex2, options)
```
vertex1: string - vertex id
vertex2: string - vertex id
options: see determinePredecessors
```javascript
graph.determinePredecessors(vertex1, source, options)
```
vertex1: string - vertex id
source: ???
options:
* cached: Boolean -> If true a cached version will be used
```javascript
graph.pathesForTree(vertex1, tree, path_to_here)
```
vertex1: string - vertex id
tree: ???
path_to_here: Internal Array, should initially be undefined or an empty array
```javascript
graph.getNeighbors(vertex1, options)
```
vertex1: string - vertex id
options:
* direction:
"inbound" -> consider only inbound edges
"outbount" -> consider only outbound edges
"any"(default) -> consider both directions
* weight: attribute-name -> use this attribute to determine edgeweight
* weight_function: function -> use this function to calculate the weight
* default-weight -> us this value if weight could not be calculated otherwise, default is Infinity
* only: function -> will be invoked on any edge, neighbors will only be included if this returns true or is not defined.
```javascript
graph.measurement(vertex1, measurement)
```
vertex1: string - vertex id
measurement: String
* "eccentricity": Calculates the eccentricity of the vertex
* "betweenness": Calculates the betweenness of the vertex
* "closeness": Calculates the closeness of the vertex
!SECTION Using Graphs in AQL {#JSModuleGraphAQL}
Complete Documentation can be copied from normal AQL documentation, with:
* replace VertexCollection/EdgeCollection by Graph
!SUBSECTION PATHS
* **BUILD ON** `ahuacatl.js`: 4090 `GRAPH_PATHS` -> uses `COLLECTION` on second arg, has to use `COLLECTION` or `GRAPH` accordingly. Has to pass the graph to traverser
Paths returns a handle for all paths included in the graph:
`GRAPH_PATHS(graphname, direction, followcycles)`
* `graphname` defines the graph
* `direction` defines the direction
* `followcycles` defines if cyclic paths should be followed
Example calls:
```javascript
FOR p in PATHS(shop, "outbound")
FILTER p.source._id == "123456/123456" && LENGTH(p.edges) == 2
RETURN p.vertices[*].name
```
!SUBSECTION TRAVERSAL
`GRAPH_TRAVERSAL(graphname, startVertex, direction, options)}`
* **BUILD ON** `ahuacatl.js`: 4243 `TRAVERSAL_FUNC` -> uses `COLLECTION` on first and second arg, has to use `COLLECTION` or `GRAPH` accordingly. Has to pass the graph to traverser
* **TO CHANGE** `common/modules/org/arangodb/graph/traversal.js`: 106 `collectionDatasourceFactory` should be able to work on Graphs
Traverses the graph described by the `graphname`,
starting at the vertex identified by id `startVertex`. Vertex connectivity is
specified by the `direction` parameter:
- `"outbound"`: Vertices are connected in `_from` to `_to` order
- `"inbound"`: Vertices are connected in `_to` to `_from` order
- `"any"`: Vertices are connected in both `_to` to `_from` and in
`_from` to `_to` order
All this is defined already for TRAVERSAL, no changes should be applied here
```javascript
TRAVERSAL(shop, "products/arangodb", "outbound", {
strategy: "depthfirst",
order: "postorder",
itemOrder: "backward",
maxDepth: 6,
paths: true
})
```
!SUBSECTION TRAVERSAL_TREES
`GRAPH_TRAVERSAL_TREE(graphname, startVertex, direction, connectName, options)`
* **BUILD ON** `ahuacatl.js`: 4243 `TRAVERSAL_FUNC` -> uses `COLLECTION` on first and second arg, has to use `COLLECTION` or `GRAPH` accordingly. Has to pass the graph to traverser
* **TO CHANGE** `common/modules/org/arangodb/graph/traversal.js`: 106 `collectionDatasourceFactory` should be able to work on Graphs
```javascript
GRAPH_TRAVERSAL_TREE(shop, "products/arangodb", "inbound", "sold", {
itemOrder: "forward"
})
```
Makes internal use of TRAVERSAL, modyfing that is sufficient.
!SUBSECTION SHORTEST_PATHS
* **BUILD ON** `ahuacatl.js`: 4243 `TRAVERSAL_FUNC` -> uses `COLLECTION` on first and second arg, has to use `COLLECTION` or `GRAPH` accordingly. Has to pass the graph to traverser
* **TO CHANGE** `common/modules/org/arangodb/graph/traversal.js`: 106 `collectionDatasourceFactory` should be able to work on Graphs
`GRAPH_SHORTEST_PATH(graphname, startVertex, endVertex, direction, options)`:
Equal to functionality of `SHORTEST_PATH`.
Makes internal use of TRAVERSAL, modyfing that is sufficient.
!SUBSECTION EDGES
* **BUILD ON** `ahuacatl.js`: 4479 `GRAPH_EDGES` -> uses `COLLECTION` on first argument, has to use `COLLECTION` or `GRAPH` accordingly.
`GRAPH_EDGES(graphname, startvertex, direction, edgeexamples, collectionRestrictions)`
Same as original, but with optional `collectionRestrictions`to define which edge collections have to be included. Default is all.
!SUBSECTION NEIGHBORS
* **BUILD ON** `ahuacatl.js`: 4508 `GRAPH_NEIGHBORS` -> uses `COLLECTION` on first, has to use `COLLECTION` or `GRAPH` accordingly.
`GRAPH_NEIGHBORS(graphname, startvertex, direction, edgeexamples)`
* Each of the graph functions in AQL (`PATHS`, `TRAVERSAL`, `TRAVERSAL_TREES`, `SHORTEST_PATHS`, `EDGES`, `NEIGHBORS`) will take the graph as its first argument (which parts of the other arguments will be pushed to be defined in FILTER and not in the signature of the function was discussed, but postponed because it is a detail).