Are you having trouble installing local packages in Deepnote? We got you covered.
Poetry is a robust dependency manager for Python projects, simplifying the management of dependencies and virtual environments. Unlike traditional package managers, it ensures consistency across different environments, maintaining a stable development workflow.
We will guide you through the steps of installing local packages in a Deepnote virtual environment. This process may seem daunting, but with our help, you will be able to navigate it with ease.
File tree
├── app
│ └── main.ipynb
├── poetry.lock
├── pyproject.toml
├── mypackage
│ ├── module1.py
│ └── module2.py
├── init.ipynb
└── .deepnote
Outline your Pyproject.toml accurately
Your pyproject.toml must contain your local package like so:
[tool.poetry]
name = "MyPackage
version = "0.1.0"
packages = [
{ include = "mypackage" },
]
[tool.poetry.dependencies]
python = "~3.9"
statsmodels = "^0.13.0"
ipykernel = { version = "^6.4.2", optional = true }
PuLP = "^2.5.1"
[tool.poetry.extras]
notebook = ["ipykernel"]
[build-system]
requires = ["poetry-core>=1.0.0"]
build-backend = "poetry.core.masonry.api"
Let Poetry handle your dependencies
Use the command poetry install
to install your dependencies.
Configure Deepnote's environment variable
Fine-tune Deepnote's environment variable, PYTHONPATH, to locate your package:
- In Deepnote, proceed to your project settings.
- Under environment variables, add a new variable with the name PYTHONPATH and value /work.
Confirm the Installation
Open your Jupyter notebook (e.g., ./app/main.ipynb) and attempt to import your package:
import mypackage
If everything is set up correctly, you should be able to import and use your package without issues.
Remember, following these simple steps will allow you to easily install and use your local packages in a Deepnote virtual environment, enhancing your productivity and innovation.
If you encounter further issues, please get in touch with our support. Happy hacking in Deepnote!