Info

Test the live project at Site If you are using Arc Browser, you can use the dark mode boost Boost

Datasets info

You can deeply investigate all the infos about datasets at Repository

RUNNING THE project

requirements

• php
• composer
• env file (use .env.example as a template). the following variables need to be set:
  • DB_DATABASE
  • DB_USERNAME
  • DB_PASSWORD
  • GIT_USERNAME
  • GIT_TOKEN

To run locally

composer install
php artisan migrate:fresh
cd website/storage/app/public && git clone https://github.com/HypergraphRepository/datasets && git clone https://github.com/geon0325/MoCHy.git && git config credential.helper store 
php artisan storage:link
php artisan serve
php artisan icons:cache

On first run, if you open the website, you will need to generate the api key through the button on landing page. You can populate the database using the python script.

python dependencies

pip install -r requirements.txt

you can generate a venv inside scripts folder and install the dependencies there:

python3.10 -m venv venv
. venv/bin/activate
pip3 install matplotlib python-dotenv requests mysql-connector-python
python3 -m pip install julia

# julia dependencies to call julia from python, you need to install the julia python package:

python3 -m pip install --user julia

Then you need to start python3 shell and run:

import julia
julia.install()

add Suppressor and SimpleHypergraph packages to julia through package manager

At the end you can check if you populate the database correctly by running inside scripts folder:

python3 checkRepo.py

You can pull update repository of dataset by running:

bash gitPull.sh

if you have setup correctly the repository, you can run the scheduler locally to activate the cron jobs:

php artisan schedule:work

Look at the cron job section to set up a cron job on your machine.

Docker build

Change hgraph.conf file to point nginx to localhost

server {
    listen       80;
    server_name  localhost;

    root /var/www/public;
    index index.php index.html;
    
    error_log  /var/log/nginx/error.log;
    access_log /var/log/nginx/access.log;
    
    location ~ \.php$ {
        try_files $uri =404;
        fastcgi_split_path_info ^(.+\.php)(/.+)$;
        fastcgi_pass app:9000;
        fastcgi_index index.php;
        include fastcgi_params;
        fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
        fastcgi_param PATH_INFO $fastcgi_path_info;
    }
    location / {
        try_files $uri $uri/ /index.php?$query_string;
        gzip_static on;
    }
}

Depending on the architecture of your machine, you have to modify the env file to build the sys.so image

DOCKER_PLATFORM="linux/amd64"

To start or stop the docker compose with:

docker compose up -d (start in background)
docker compose down (-v to remove volumes)

After the dockers are up, you can run the migration to initialize the database and start a shell in the container

docker exec -it hgraph php artisan migrate:fresh
docker exec -it hgraph bash

Then, while in the hgraph docker (in the folder var/www), you have to copy (or move) the custom julia image to the scripts folder then

cp ../../../sysimage/sys.so scripts/
python3 scripts/updateDB.py

If you want to rebuild images after a change, run:

docker compose up --build 
docker-compose up -d --build <service_name> --no-cache

If you want to rebuild images without cache, run:

docker compose build --no-cache

If you want to see the logs, run:

docker compose logs -f -t

To use the julia script, compile a custom system image for PyJulia, run

python3 -m julia.sysimage sys.so

To build a single service

docker compose build name_service

To run a command inside the docker

sudo docker exec -it hgraph php artisan <command>

Enter in postgres container

docker exec -it website-postgres-1 psql <namedb> <username>

cron job

To run the scheduler with the scripts, is good practice to set up a cron job this way:

crontab -e

and insert the following line:

* * * * * cd /path-to-your-project && php artisan schedule:run >> /dev/null 2>&1

To renew certificate with certbot, run:

docker compose run --rm certbot renew

You can also add a cron job to renew the certificate automatically:

@monthly docker compose -f /path-to-your-project/docker-compose.yml run --rm certbot renew

Jupyter Notebook

To add your Notebook files simply place them in the following project folder:

/storage/app/public/notebooks/

Also, it is important to know that Jupyter Notebook files cannot be edited or created through the web page.