oncokb-variant-recommendation

Set up python environment

Running Locally

1. First Time Setup

   1. Create a python virtual environment the .venv folder for python to use

      python3 -m venv .venv

   2. Create a .env

      cp env.example .env

2. Activating Virtual Environment

   1. Make python download and use open source code inside your .venv folder

      source .venv/bin/activate

   2. Check if python using .venv folder

      which python3

3. Install the latest packages for the project

   pip install -r requirements.txt

Running Tests

pytest

• Note that all test files must end in _test.py

Create/Update requirements file

pip freeze > requirements.txt

Set Airflow

Install Airflow

pip install 'apache-airflow==2.9.1' \
 --constraint "https://raw.githubusercontent.com/apache/airflow/constraints-2.9.1/constraints-3.8.txt"

• Change constraints-3.8.txt to your python version
• More details in Airflow document

Initialize Airflow

1. Create an airflow folder under your object, and create a dags folder under airflow.

2. To configure Airflow to recognize your DAGs directory, you need to set the AIRFLOW_HOME environment variable. replacing /path/to/dags/folder/parent/folder with the actual path to your desired directory:

export AIRFLOW_HOME= path/to/dags/flod/parent/flod

3. Run Airflow Standalone, and get the default username and password.

airflow standalone

The command initializes the database, creates a user, and starts all components.

• If you prefer to run individual components of Airflow manually, or if you need personalized user information, instead of using the all-in-one standalone command, you can run the following:

airflow db init

airflow users create \
    --username admin \
    --firstname Peter \
    --lastname Parker \
    --role Admin \
    --email [email protected]

airflow webserver --port 8080

airflow scheduler

airflow triggerer

• More details in Airflow document

4. Access the Airflow UI: Visit localhost:8080 in your browser.

5. Connect to AWS S3: Choose connections under Admin, create a new connection. Input oncokb_s3 in Connection Id, choose Connection Type as Amazon Web Services , and input AWS Access Key ID and AWS Secret Access Key.

6. Connect to MySQL: Choose connections under Admin, create a new connection. Input oncokb_mysql in Connection Id, choose Connection Type as MySQL , and inputHost,schema,login,password and port.

Note:

• If you want to close all Airflow DAG or connectionsexamples on Airflow webserver. Open airflow.cfg and change load_examples = False or load_default_connections = False.

Test Airflow

With the Airflow CLI, run to test your dag, you can check the result and logs at Airflow UI.

airflow dags test <dag_id>

You can use CLI to list all dags you have.

airflow dags list