图形数据库之-neo4j(非关系型数据库)

2013-10-25

图形数据库之---neo4j(非关系型数据库)Neo4j是一个高性能的,NOSQL图形数据库，它将结构化数据存储在网络上

图形数据库之---neo4j(非关系型数据库)

Neo4j是一个高性能的,NOSQL图形数据库，它将结构化数据存储在网络上而不是表中。Neo4j也可以被看作是一个高性能的图引擎，该引擎具有成熟数据库的所有特性。程序员工作在一个面向对象的、灵活的网络结构下而不是严格、静态的表中——但是他们可以享受到具备完全的事务特性、企业级的数据库的所有好处。

Neo4j因其嵌入式、高性能、轻量级等优势，越来越受到关注。

图形数据结构
在一个图中包含两种基本的数据类型：Nodes（节点）和 Relationships（关系）。Nodes 和 Relationships 包含key/value形式的属性。Nodes通过Relationships所定义的关系相连起来，形成关系型网络结构。

neo4j是一种图数据库，同时它也是一种嵌入式数据库。它对图数据是以节点和边(关系)模式进行存储。每个节点可以包含一系列信息，通过Node类里面的setProperty()方法对节点信息进行存储，Node也可以使用createRelationshipTo()方法实现个节点和其他节点的联系，并且该方法返回的是一个Relationship对象，我们也可以对Relationship设置属性，也就是节点和节点之间的关系属性。什么叫关系属性?例如：person1àperson2,person1和person2的关系可以是朋友也可以是同学还可以是亲人，这里的朋友、同学、亲人就是这里的Relationship的属性。那么关系属性就是描叙两个节点之间的关系类型。这就方便在对节点进行查找的时候对节点进行过滤。

neo4j数据库中的节点类似关系型数据库中的表。neo4j节点之间的关系类似关系型数据库中的2表之间的第三张表，neo4j数据库的节点之间的关系上也可以保存一些2者之间关系的属性键值对。同样，可以为每种关系创建一个名字，把关系归类管理。图形数据库比关系型性能好很多，操作性很好，可以灵活的控制，根据neo4j中Node的不同，我们可以赋予不同的属性
http://docs.neo4j.org/refcard/2.0/
START`STARTn=node()STARTn=node({ids})STARTn=node({id1}), m=node({id2})STARTn=node:nodeIndexName(key={value})`node_auto_indexfor the automatic index.
MATCH`MATCH(n:Person)-[:KNOWS]->(m:Person)WHEREn.name="Alice"`START clause is required then.
`MATCH(n)-->(m)`MATCH except the ones containing property maps.
`MATCHp = (n)-->(m)`p.
WHERE`WHEREn.property <> {value}`Predicates`n.property <> {value}has(n.property)n.number >= 1` `ANDn.number <= 10n:PersonidentifierISNULL`NULL.
`NOThas(n.property)ORn.property = {value}`TRUE.
`n.property = {value}`NULL, which is only equal to NULL.
`n.property =~` `"Tob."(n)-[:KNOWS]->(m)NOT(n)-[:KNOWS]->(m)`(n)-[:KNOWS]->(m) from the result.
`n.propertyIN[{value1}, {value2}]`Predicate Functions`all(xINcollection` `WHEREhas(x.property))`true if the predicate is TRUE for all elements of the collection.
`any(xINcollection` `WHEREhas(x.property))`true if the predicate is TRUE for at least one element of the collection.
`none(xINcollection` `WHEREhas(x.property))`TRUE if the predicate is FALSE for all elements of the collection.
`single(xINcollection` `WHEREhas(x.property))`TRUE if the predicate is TRUE for exactly one element in the collection.
Scalar Functions`length(collection)type(a_relationship)startNode(a_relationship)endNode(a_relationship)coalesce(n.property, {defaultValue})`NULL expression.
`head(collection)last(collection)timestamp()id(node_or_relationship)`String Functions`str({expression})replace({original}, {search}, {replacement})`search with replacement. All arguments are be expressions.
`substring({original}, {begin}, {sub_length})`sub_length argument is optional.
`left({original}, {sub_length}),` `right({original}, {sub_length})trim({original}),ltrim({original}),` `rtrim({original})upper({original}),lower({original})`RETURN`RETURNRETURNn` `AScolumnNameRETURNDISTINCT` `nORDERBY` `n.propertyORDERBY` `n.property` `DESCSKIP{skip_number}LIMIT{limit_number}SKIP{skip_number}` `LIMIT{limit_number}`WITH`MATCH(user)-[:FRIEND]-(friend)WHEREuser.name = {name}WITHuser,` `count(friend)ASfriendsWHEREfriends > 10RETURNuser`WITH syntax is similar to RETURN. It separates query parts explicitly, allowing you to declare which identifiers to carry over to the next part.
`MATCH(user)-[:FRIEND]-(friend)WITHuser,` `count(friend)ASfriendsORDERBY` `friends` `DESCSKIP1` `LIMIT3RETURNuser`ORDER BY, SKIP, LIMIT with WITH.
CASE`CASEn.eyes` `WHEN'blue'` `THEN` `1` `WHEN'brown'` `THEN` `2` `ELSE3END`THEN value from the matching WHEN value. TheELSE value is optional, and substituted for NULL if missing.
`CASE` `WHENn.eyes =` `'blue'THEN` `1` `WHENn.age < 40` `THEN2` `ELSE3END`THEN value from the first WHEN predicate evaluating to TRUE. Predicates are evaluated in order.
Collection Functions`labels(n)nodes(path)relationships(path)extract(xINcoll | x.prop)filter(xINcoll` `WHEREx.prop <> {value})`TRUE.
`tail(coll)range({first_num}, {last_num}, {step})`step argument is optional.
`reduce(s =` `"", n` `INcoll | s + n.prop)FOREACH(n` `INcoll |` `SETn.marked =` `TRUE)`Aggregation`count()count(identifier)`NULL values.
`count(DISTINCTidentifier)`DISTINCTmodifier, which removes duplicates from the values.
`sum(n.property)avg(n.property)max(n.property)min(n.property)collect(n.property)`NULL.
`percentile_disc(n.property, {percentile})`percentile argument is from 0.0to 1.0.
`percentile_cont(n.property, {percentile})stdev(n.property)stdevp(n.property)(CREATE[UNIQUE] |` `MERGE)[SET|DELETE|FOREACH][RETURN[ORDERBY] [SKIP] [LIMIT]][START][MATCH][WHERE][WITH[ORDERBY] [SKIP] [LIMIT]][CREATE[UNIQUE]|MERGE][SET|DELETE|FOREACH][RETURN[ORDERBY] [SKIP] [LIMIT]]`CREATE`CREATE(n {name: {value}})CREATE(n {map})CREATE(n {collectionOfMaps})CREATE(n)-[r:KNOWS]->(m)CREATE(n)-[:LOVES {since: {value}}]->(m)`CREATE UNIQUE`CREATEUNIQUE` `(n)-[:KNOWS]->(m {property: {value}})`MERGE`MERGE(n:Person {property: {value}})ONCREATE` `n` `SETn.created =` `timestamp()ONMATCH` `n` `SETn.access = n.access+1`ON CREATE n and ON MATCH n for conditional updates.
SET`SETn.property={value}SETn={map}SETn:Person`Person to a node.
DELETE`DELETEn, rREMOVEn:Person`n.
`REMOVEn.property`INDEX`CREATEINDEX` `ON` `:Person(name)`Person and property name.
`DROPINDEX` `ON` `:Person(name)`Person and property name.
`MATCH(n:Person)` `WHEREn.name = {value}`lower(n.name) = {value} will not use an index.
Mathematical Functions`abs({expr})rand()round({expr})floor({expr})ceil({expr})sqrt({expr})sign({expr})`0 if zero, -1 if negative, 1 if positive.
`sin({expr})`cos, tan, cot, asin, acos,atan, atan2({y-expr}, x-expr}).
`degrees({expr}),radians({expr}),pi()`radians for the reverse. pi for π.
`log10({expr}),log({expr}),exp({expr}),e()`e to the power of the parameter. Value of e.

热点排行