图数据库neo4j学习笔记

1.Neo4是非关系型数据库，所以为已存在的节点增加属性可直接通过UPDATE属性来实现，如原节点：

{

  "name": "迪丽热巴",

  "age": 30,

  "sex": "女"

}

Update语句为:
match (u:userinfo)

        where id(u) = 12

        set u.name = '迪丽热巴',u.age = 32,u.sex = '女',u.personid='222402196201193174';

这样就增加了personid这个属性了，而后通过update语句：
match (u:userinfo) where id(u) = 5 set u.personid='411282197509280661'

就能更新属性了。

2.将数据导入到Neo4j步骤:

将数据导出成cvs文件，例如:

 

 (因无法上传EXCEL,所以将列表贴出来了)

放在某个可以访问的服务里，例如：

 

通过http://localhost:8095/person.csv  能下载到这个文件。

执行

LOAD CSV WITH HEADERS FROM "http://localhost:8095/person.csv" AS row

CREATE (n:person)

SET n = row

导入数据

 

 

3.CQL编写:

a.两节点建立关系：

match (a:person),(b:bankcard) where a.personid=b.personid create (a)-[l:帐户]->(b) RETURN a,b

b.删除所有节点的关系：

match (a:person)-[l:帐户]->(b:bankcard) delete l

c.按姓名查询两节点信息：

match (a:person),(b:bankcard) where a.personid=b.personid and a.personname='孟君浩' RETURN a,b

导入案件:

LOAD CSV WITH HEADERS FROM "http://localhost:8080/case.csv" AS row

CREATE (n:case)

SET n = row

导入案件、人员关系：

LOAD CSV WITH HEADERS FROM "http://localhost:8080/case_person.csv" AS row

CREATE (n:case_person)

SET n = row

案件嫌疑人关系:

match (a:case),(b:case_person) where a.caseid=b.caseid create (a)-[l:嫌疑]->(b) RETURN a,b

 

match (a:person),(b:case_person) where a.personid=b.personid create (a)-[l:涉案]->(b) RETURN a,b

 

match (a:person),(b:case_person),(c:case) where a.personid=b.personid and b.caseid=c.caseid  create (c)-[l:嫌疑人]->(a) return a,c

b.删除所有节点的关系：

match (a: case)-[l: 嫌疑人]->(b: person) delete l

c.查询人员：

match (a:person),(b:case_person),(c:case) where a.personid=b.personid and b.caseid=c.caseid and a.personid='410225197206141869' RETURN  a,c

d.导入转帐数据：

LOAD CSV WITH HEADERS FROM "http://localhost:8080/transfer.csv" AS row CREATE (n:transfer) SET n = row

e.建立关系:

match (fp:person),(fb:bankcard),(e:transfer),(tp:person),(tb:bankcard) where fp.personid=fb.personid and fb.cardid=e.from_cardid and e.to_cardid=tb.cardid and tb.personid=tp.personid create (fp)-[l:转帐]->(tp) RETURN fp,tp

f.按身份证号查询所有关系

match (u:person{personid:'654226195912244342'})- [l] -> (u1)

return  u

union

match (u)- [l] -> (u1:person{personid:'654226195912244342'})

return u

 

match (p)- [l] -> (p1:person{personid:'430426199308077079'})

return p,type(l) as relation,p1

 

g.设置人员标签:

MATCH (a:person) set a:老板:行政违法人员:高危地区:线索移送:反洗钱可疑交易主体:战支嫌疑主体

MATCH (a:bankcard) set a:财务人员:已开户:高危地区:非法集资:反洗钱可疑交易:偷税:前科:漏税

 

企业数据导入：

LOAD CSV WITH HEADERS FROM "http://localhost:8080/company.csv" AS row CREATE (n:company) SET n = row

 

LOAD CSV WITH HEADERS FROM "http://localhost:8080/company_person.csv" AS row CREATE (n:company_person) SET n = row

生成关系：

match (c:company),(cp:company_person),(p:person) where cp.personid=p.personid and cp.comid=c.comid and cp.relationship='普通员工' create (p)-[l:普通员工]->(c) RETURN c,p

 

match (c:company),(cp:company_person),(p:person) where cp.personid=p.personid and cp.comid=c.comid and cp.relationship='法人' create (p)-[l:法人]->(c) RETURN c,p

 

1.生成关系：

match(b1:bankcard),(t:transfer),(b2:bankcard)

where b1.cardid=t.from_cardid and t.to_cardid=b2.cardid

create (b1)-[:转帐{trandate:t.trandate,charge:t.charge,remark:b1.cardname+'转帐'+b2.cardname}]->(b2)

2. 关系全文搜索
a.创建关系索引
CALL db.index.fulltext.createRelationshipIndex("relationIndex",["转帐","员工"],["charge","trandate","remark"])
"转帐","员工"  是关系名称
"charge","trandate","remark" 是关系的属性
b.关系全文搜索

              CALL db.index.fulltext.queryRelationships("relationIndex", '${key}') YIELD relationship 

              with startNode(relationship) as startnode,type(relationship) as r,endNode(relationship) as endnode

              return startnode,labels(startnode) as startlabel,r,endnode,labels(endnode) as endlabel

              skip 0

              limit 20
key是关健字
c.删除索引
call db.index.fulltext.drop("relationIndex")

3.节点全文搜索
a.创建节点索引
CALL db.index.fulltext.createNodeIndex("personIndex",["person","case","company"],["personid","personname","mobile","address","work","education","caseid","casename","comid","comname","remark"])
b.关系节点搜索
CALL db.index.fulltext.queryNodes("personIndex", "思") YIELD node return node
c.删除索引
call db.index.fulltext.drop("personIndex")

4.关系数据统计：
       示例:按身份证号查询人员出入帐统计

       match(b1:bankcard{personid:'${personid}'})-[r:转帐]->(b2:bankcard)

               return 'out' as farward,substring(r.trandate,0,7) as month,sum(r.charge) as money,count(1) as count

               union

               match(b1:bankcard)-[r:转帐]->(b2:bankcard{personid:'${personid}'})

               return 'in' as farward,substring(r.trandate,0,7) as month,sum(r.charge) as money,count(1) as count
5.教训：
   a.关系是有自己属性的，关系名1:关系名2:关系名3{属性1:值1,属性2:'值2'......}
   b.全文关系搜索依赖于关系的“属性集合”进行搜索。
   c.关系属性是可以统计的

 

人员转帐关系：

match (p1:person),(b1:bankcard)-[r:转帐]->(b2:bankcard),(p2:person)

where (p1.personid=b1.personid) and (b2.personid=p2.personid)

create (p1)-[l:资金往来{inout:'转出'}]->(p2)

 

6.导入内存不足解决

There is not enough memory to perform the current task. Please try increasing 'dbms.memory.heap.max_size' in the neo4j configuration (normally in 'conf/neo4j.conf' or, if you are using Neo4j Desktop, found through the user interface) or if you are running an embedded installation increase the heap by using '-Xmx' command line flag, and then restart the database.

下面以桌面版为示例：

桌面版->(左侧)数据库->你建的数据库->Settings->

dbms.memory.heap.initial_size=2G

dbms.memory.heap.max_size=4G

dbms.memory.pagecache.size=2G

都改为原来的4位内存

然后点击:apply，重启数据库就行了

 

LOAD CSV WITH HEADERS FROM "http://localhost:8080/commission.csv" AS row

CREATE (n:commission)

SET n = row

 

 

 导入333754节点合计5516 ms

后面我会发布springboot+mysql操作的代码，请期待。