In this example, we show you how to import and use dataflows from the Apache Hamilton Dataflow Hub. You can use them either directly or pull and edit a local copy.
For the purpose of this example, we will create a virtual environment with hamilton, the hamilton contrib module, and the requirements for the dataflow we'll import.
python -m venv ./venv. venv/bin/activate (on MacOS / Linux) or . venv/bin/Scripts (Windows)pip install -r requirements.txtOr run it in Google Colab: 
There are 3 main ways to use community dataflows: static installation, dynamic installation, and local copy (see documentation). We present each of them in this example:
The script run.py uses the direct import from hamilton.contrib.user.zilto import xgboost_optuna. It's as simple as that! (but first pip install sf-hamilton-contrib --upgrade)
The first part of the notebook notebook.ipynb imports the same dataflow via xgboost_optuna = hamilton.dataflows.import_module("xgboost_optuna", "zilto"). This will download and cache the module in your local directory {USER_PATH}/.hamilton.
After completing the dynamic installation, the second part of the notebook includes hamilton.dataflows.copy(xgboost_optuna, destination_path="./my_local_path") will create a local copy at the desire location. Then, you'll be able to do from my_local_path import xgboost_optuna.
You can find more information on how to contribute in the contrib module's README