Data Schema

Vertex:

• person(id: string, name: string, is_confirmed: bool, confirmed_at: datetime)
• address(id: string, name: string, street_id: string)
• street(id: string, name: string, city_id: string)
• city(id: string, name: string, province_id: string)
• province(id: string, name: string)
• district(id: string, name: string, city_id: string)
• town(id: string, name: string, district_id: string)

Edge:

• person_live_with_person(sid, did, start_time, end_time)
• person_visit_address(sid, did, start_time, end_time)
• belong_to(sid, did)

  • sid(source ID) could be {address, district, street, town, city} id, did(destination ID) could be {distict, street, town, city, province} id

Schema DDL:

-- Create Vertex:

CREATE TAG IF NOT EXISTS 人(name string NOT NULL, is_confirmed bool NOT NULL);
CREATE TAG IF NOT EXISTS 联系方式(num int NOT NULL);
CREATE TAG IF NOT EXISTS 地址(name string NOT NULL);
CREATE TAG IF NOT EXISTS 街道(name string NOT NULL);
CREATE TAG IF NOT EXISTS 村镇(name string NOT NULL);
CREATE TAG IF NOT EXISTS 行政区(name string NOT NULL);
CREATE TAG IF NOT EXISTS 城市(name string NOT NULL);
CREATE TAG IF NOT EXISTS 省份(name string NOT NULL);

-- Create Edge:

CREATE EDGE 同住(start_time float, end_time float);
CREATE EDGE 到访(start_time float, end_time float);
CREATE EDGE 属于();
CREATE EDGE 住址();

-- Create Index: 

CREATE TAG INDEX IF NOT EXISTS person_index_0 on 人(name(10), is_confirmed);
CREATE TAG INDEX IF NOT EXISTS addr_index_0 on 地址(name(10));
CREATE TAG INDEX IF NOT EXISTS addr_index_1 on 地址(risk_level(10));
CREATE TAG INDEX IF NOT EXISTS street_index_0 on 街道(name(10));
CREATE TAG INDEX IF NOT EXISTS town_index_0 on 村镇(name(10));
CREATE TAG INDEX IF NOT EXISTS dist_index_0 on 行政区(name(10));
CREATE TAG INDEX IF NOT EXISTS city_index_0 on 城市(name(10));
CREATE TAG INDEX IF NOT EXISTS prov_index_0 on 省份(name(10));

CREATE EDGE INDEX IF NOT EXISTS live_index_0 on 同住(start_time);
CREATE EDGE INDEX IF NOT EXISTS live_index_1 on 同住(end_time);
CREATE EDGE INDEX IF NOT EXISTS visit_index_0 on 到访(start_time);
CREATE EDGE INDEX IF NOT EXISTS visit_index_1 on 到访(end_time); 

How to Generate Data

# install dependency
python3 -m pip install -r requirements.txt

# data generation
python3 data_generator.py

# Check files
$ tree data
data
├── address.csv
├── city.csv
├── district.csv
├── person.csv
├── person_livewith.csv
├── person_visit.csv
├── province.csv
├── street.csv
└── town.csv

0 directories, 9 files

Data birdview

address.csv

a_0,翔安王街x座 420514,s_160127
a_1,房山张街V座 641042,s_605928
a_2,高坪淮安路B座 280940,s_131728
a_3,东丽太原路Y座 678891,s_292193
a_4,南溪永安街C座 960428,s_307763

city.csv

city_id,city_name,province_id
1101, 市辖区, 11
1201, 市辖区, 12
1305, 邢台市, 13
1304, 邯郸市, 13

district.csv

district_id,district_name,city_id
110116, 怀柔区, 1101
110117, 平谷区, 1101
110114, 昌平区, 1101
110115, 大兴区, 1101

person.csv

p_0,焦明,False,
p_1,陈凤英,False,
p_2,金丹,False,
p_3,李婷,False,
p_4,葛娜,False,
p_5,曾璐,False,
p_6,朱秀荣,False,

person_livewith.csv

p_8717,p_5397,1644980628.0,1647218816.0
p_8752,p_4142,1640384466.0,1643261933.0
p_9877,p_1924,1642333831.0,1645039968.0
p_8762,p_5514,1640413494.0,1642081958.0
p_1650,p_2292,1646517804.0,1647573515.0

person_visit.csv

p_8429,a_142,1644193437.0,1647634538.0
p_1341,a_933,1645185598.0,1646532593.0
p_4358,a_387,1644457470.0,1644783704.0
p_9153,a_341,1642268021.0,1644437771.0

province.csv

province_id, province_name
11, 北京市
12, 天津市
13, 河北省

street.csv

incremental_id,street_id, street_name, town_id
s_1,110112117236, 应寺村委会,110112117
s_2,110112117237, 熬硝营村委会,110112117
s_3,110112117238, 临沟屯村委会,110112117
s_4,110112105250, 苍上村委会,110112105

town.csv

town_id, town_name, district_id
110114013, 史各庄街道办事处, 110114
110114011, 回龙观街道办事处, 110114
110114012, 龙泽园街道办事处, 110114
110114010, 霍营街道办事处, 110114

Import to Nebula Graph

We could leverage nebula-importer to import the dataset into NebulaGraph, the reference configuration file is here

For example, if you run NebulaGraph with Docker-Compose in docker network named nebula-net, you could do this to load the data:

1. Create NebulaGraph with Nebula-UP

curl -fsSL nebula-up.siwei.io/install.sh | bash

2. Create Graph Space and Schema:

Access the NebulaGraph console:

~/.nebula-up/console.sh

Create Graph Space and Schema:

CREATE SPACE IF NOT EXISTS covid_trace (partition_num=5, replica_factor=1, vid_type=FIXED_STRING(32));
:sleep 20
USE covid_trace;

CREATE TAG IF NOT EXISTS 人(name string NOT NULL, is_confirmed bool NOT NULL);
CREATE TAG IF NOT EXISTS 联系方式(num int NOT NULL);
CREATE TAG IF NOT EXISTS 地址(name string NOT NULL);
CREATE TAG IF NOT EXISTS 街道(name string NOT NULL);
CREATE TAG IF NOT EXISTS 村镇(name string NOT NULL);
CREATE TAG IF NOT EXISTS 行政区(name string NOT NULL);
CREATE TAG IF NOT EXISTS 城市(name string NOT NULL);
CREATE TAG IF NOT EXISTS 省份(name string NOT NULL);

CREATE EDGE 同住(start_time float, end_time float);
CREATE EDGE 到访(start_time float, end_time float);
CREATE EDGE 属于();
CREATE EDGE 住址();

:sleep 20
CREATE TAG INDEX IF NOT EXISTS person_index_0 on 人(name(10), is_confirmed);
CREATE TAG INDEX IF NOT EXISTS addr_index_0 on 地址(name(10));
CREATE TAG INDEX IF NOT EXISTS addr_index_1 on 地址(risk_level(10));
CREATE TAG INDEX IF NOT EXISTS street_index_0 on 街道(name(10));
CREATE TAG INDEX IF NOT EXISTS town_index_0 on 村镇(name(10));
CREATE TAG INDEX IF NOT EXISTS dist_index_0 on 行政区(name(10));
CREATE TAG INDEX IF NOT EXISTS city_index_0 on 城市(name(10));
CREATE TAG INDEX IF NOT EXISTS prov_index_0 on 省份(name(10));

CREATE EDGE INDEX IF NOT EXISTS live_index_0 on 同住(start_time);
CREATE EDGE INDEX IF NOT EXISTS live_index_1 on 同住(end_time);
CREATE EDGE INDEX IF NOT EXISTS visit_index_0 on 到访(start_time);
CREATE EDGE INDEX IF NOT EXISTS visit_index_1 on 到访(end_time);

3. Import data

git clone https://github.com/wey-gu/covid-track-graph-datagen && cd covid-track-graph-datagen

docker run --rm -ti \
    --network=nebula-net \
    -v ${PWD}/:/root \
    vesoft/nebula-importer:v3.1.0 \
    --config /root/nebula-importer-config.yaml