Neo4j-数据导入

# cust.csv
id:ID,certno,name,label,cust_certtype:string,cust_namespell:string,birthday:int,sex:string,address:string,mob_phone:string
# tran.csv
:START_ID,:END_ID,Type,tran_date:int,amount:int,count:int

neo4j-import适应场景
neo4j-import参数  (bin/neo4j-import help)
    - 指定最大进程数 --processors <max processor count>
    - 跳过重复节点 --skip-duplicate-nodes <true/false>
    - 跳过异常关系 --skip-bad-relationships <true/false>
    - 最大跳过数量 --bad-tolerance <max number of bad entries, or true for unlimited>
    - 跳过异常行(例如行列数异常) --ignore-extra-columns <true/false>


命令样例:

bin/neo4j-import --bad-tolerance=1000000 --skip-duplicate-nodes=true --skip-bad-relationships=true --into data/databases/graph.db --id-type string --nodes:cust import/c.csv  --relationships:tran import/t.csv

# cust.csv
id:string:id_index,certno:string:id_index,name,Label:label,cust_certtype:string,cust_namespell:string,birthday:int,sex:string,address:string,mob_phone:string
874018718864465,874018718864465,鞠瑜,cust,其他证件,juyu,19801010,男,河北省岩市龙潭东莞街d座 188848,15767524738
# tran.csv
id:string:id_index,certno:string:id_index,Type,tran_date:int,amount:int,count:int
874018718864465,411224195908138440,tran,20180108,20,2

# batch-import命令
sh import.sh /home/fdm/neo4j_test/neo4j-community-3.0.4/data/databases/fdm.db /home/fdm/import_data/c.csv /home/fdm/import_data/t.csv

# batch-import3.0已经编译好的软件包
https://github.com/mo9527/batch-import-tool

# batch.properties配置文件
dump_configuration=false
cache_type=none
use_memory_mapped_buffers=true
neostore.propertystore.db.index.keys.mapped_memory=1000M
neostore.propertystore.db.index.mapped_memory=10M
neostore.nodestore.db.mapped_memory=10240M
neostore.relationshipstore.db.mapped_memory=10240M
neostore.propertystore.db.mapped_memory=5120M
neostore.propertystore.db.strings.mapped_memory=2000M
#batch_import.csv.quotes=true
#batch_import.csv.delim=,workInfoId
#contactRecordId deviceId workInfoId
#batch_array_separator=,

batch_import.csv.quotes=true
batch_import.csv.delim=,
batch_import.keep_db=true
batch_import.node_index.id_index=exact
batch_import.node_index.id_index2=exact
batch_import.node_index.id_index3=exact

load csv 导入格式要求

# cust.csv
id,certno,name,label,cust_certtype,cust_namespell,birthday,sex,address,mob_phone
# tran.csv
start_id,end_id,type,tran_date,amount,count

# create 节点
LOAD CSV WITH HEADERS FROM "file:///cust.csv" AS line create
(p:cust{id:line.id,certno:line.certno,name:line.name,label:line.label,cust_certtype:line.cust_certtype,cust_namespell:line.cust_namespell,birthday:toInteger(line.birthday),sex:line.sex,address:line.address,mob_phone:line.mob_phone})

# 创建索引
CREATE INDEX ON :cust(id)

# create 边
LOAD CSV WITH HEADERS FROM "file:///tran.csv" AS line match
(from:cust{id:line.start_id}),(to:cust{id:line.end_id}) create (from)-[r:tran{ type:line. type,tran_date:line.tran_date,amount:line.amount,count:line.count}]->(to)