To schedule a Python script or Python function in Airflow, we use PythonOperator.

Note: For Airflow 2.0+, consider using the TaskFlow API for a more modern and cleaner approach to writing Python tasks.

from airflow import DAG
from airflow.operators.python_operator import PythonOperator

def my_python_function():
    # your code goes here
    print('Hello')

def my_python_function_with_context(**context):
    # For more detail about "context" object,
    # please refer to https://blog.duyet.net/2019/08/airflow-context.html
    ds = context['ds']
    print(f'Dag run at {ds}')

dag = DAG('dag_id')
PythonOperator(dag=dag,
               task_id='my_python_function',
               python_callable=my_python_function)

PythonOperator(dag=dag,
               task_id='my_python_function_with_context',
               provide_context=True,
               python_callable=my_python_function_with_context)

Passing in arguments

Use the op_args and op_kwargs arguments to pass additional arguments to the Python callable.

from airflow import DAG
from airflow.operators.python_operator import PythonOperator

def my_python_function(ds, lucky_number, **kwargs):
    print(f'Dag run at {ds}')
    print(f'Random number is {lucky_number}')

dag = DAG('dag_id')
run_this = PythonOperator(dag=dag,
               task_id='my_python_function',
               provide_context=True,
               python_callable=my_python_function,
               op_kwargs={
                   'lucky_number': 99
               })

References

• http://airflow.apache.org/docs/stable/howto/operator/python.html
• Airflow - "context" dictionary