GRAKN_REFINERY2

Setup Docker

1. Download docker desktop and install it.
2. In run folowing command:

echo [wsl2] >>~/.wslconfig
echo memory=6GB >>~/.wslconfig

Update to match your system. roughly 50% of available memory is recommended.

Run in docker

docker compose up

Open the browser and navigate to http://localhost:8080/login to view the site. First time use register to create a new account. First time data load will take a few minutes to pupulate typedb, but you can continue with your user registration.

Open the browser and navigate to http://localhost:8081/ to view the mongo database.

Installation

First pull repo into a directory

Install Pipenv

pip install pipenv Flask asdf

Then setup python virtual env

pipenv shell

then install dependencies with

pipenv install

then after everything installed

export FLASK_APP=refinery.py
flask run --host=0.0.0.0

then access localhost\login to access the webpage.

Testing TypeDB Connection and Basic Dataset

Make sure TypeDB is running on localhost, with pm_4 logs loaded

The input data for this test is input_test1_connection.json

To run the test, run python examples\z_test1_tdb_query.py

The test will produce an output json that contains the basic dataset. Note all algorithms independently produce output, and sometimes input jsons

Testing TypeDB Grouping and Grouped Dataset

Make sure TypeDB is running on localhost, with pm_4 logs loaded

The input data for this test is input_test2_colaGraph_sample.json, which contains the sample Basic dataset, and input_test2_definition.json which contains the group definitions for 3 nested groups

To run the test, run python examples\z_test2_Grouping.py

The test will produce an output json. Note all algorithms independently produce output, and sometimes input jsons

Testing TypeDB Group Collapse and Collapsed Dataset

Make sure TypeDB is running on localhost, with pm_4 logs loaded

The input data for this test is input_test3_Grouped.json, which contains the sample Grouped dataset, and a string that contains one of the group names, in this case "session"

To run the test, run python examples\z_test3_Collapsing.py

The test will produce an output json. Note all algorithms independently produce output, and sometimes input jsons