ww-jermaine / data_cord19

Transfers CORD-19 data into a neo4j graph

CORD-19 Data loader

This python script helps to transform the data set from the COVID-19 Open Research Dataset Challenge into a neo4j graph

Maintainer: Tim

Version: 0.1.0

Python version: Python3

ThanksTo: https://pypi.org/project/cord-19-tools/

Usage

Docker

Run

docker run -it --rm --network host --name cord19 -e CONFIGS_NEO4J='{"host":"localhost"}' covidgraph/data-cord19

NOTE: For details on the -e CONFIGS_NEO4Jenv variable see https://github.com/covidgraph/motherlode/blob/master/README.md#the-neo4j-connection-string

Build local image

From the root directorie of this repo run:

docker build -t data-cord19 .

Run local image

docker run -it --rm --network host --name data-cord19 -e CONFIGS_NEO4J='{"host":"localhost"}' data-cord19

Examples (neo4j runs on the docker linux host machine)

docker run -it --rm --network host --name data-cord19 -v ${PWD}/dataset:/app/dataset -e CONFIGS_NEO4J='{"host":"localhost"}' data-cord19

docker run -it --rm --name data-cord19 -e CONFIGS_NEO4J='{"host":"localhost"}' data-cord19

My common dev docker call looks like this:

docker run -it --rm --network host --name data-cord19 -v ${PWD}/dataset:/app/dataset -v ${PWD}/dataloader:/app/dataloader -e CONFIGS_NEO4J='{"host":"localhost"}' data-cord19

Envs

The most important Env variables are:

ENV: will be PROD or DEV

CONFIGS_NEO4J: defaults to {"host":"localhost"}. The connections details for the database. For details see https://github.com/covidgraph/motherlode/blob/master/README.md#the-neo4j-connection-string

besides that you can set all variables in dataloader/config.py via env variable with a CONFIGS_ prefix. See https://git.connect.dzd-ev.de/dzdtools/pythonmodules/-/tree/master/Configs for more details on how to manipulate the parameters

Volumes

/app/dataset

Here is the downloaded data set located. You can mount this path with -v /mylocal/path:/app/dataset to prevent redownloading of the dataset.

/app/dataloader

Here is the python source code located. You can mount this for development or tinkering

Local

Copy dataloader/env/DEFAULT.env to dataloader/env/DEVELOPMENT.env:

cp dataloader/env/DEFAULT.env dataloader/env/DEVELOPMENT.env

Enter your neo4j connection string at dataloader/env/DEVELOPMENT.env into the variable CONFIGS_NEO4J:

CONFIGS_NEO4J={"host":"localhost"}

Install the requirements with

pip3 install -r requirement.txt

run the main.py

python3 main.py

Data

Scheme

Exmaple