environments

Environments

Recommended is venv because rest don't really work, but uv is on the scene

Update Outdated Packages

pip3 list --outdated --format=freeze | grep -v '^\-e' | cut -d = -f 1 | xargs -n1 pip3 install -U

UV

  • Package manager written in Rust combining pip, venv, penv

  • 10-100x faster then pip

  • Uses pyproject.toml to manage dependencies, .python-version to manage py version,

  • First time you run uv run, uv sync, or uv lock, creates venv and uv.lock which has exact dependencies

# Initialized project in current dir with pyproject
uv init
# To migrate from requirements.txt
uv add -r requirements.txt -c constraints.txt
# To sync environment with listed dependencies
uv sync

uv add 'requests==2.31.0'
uv add ruff
uv run ruff check

uv remove requests

Creating Specific Envs

uv python install 3.10 3.11 3.12
uv venv --python 3.12.0

Groups

uv add --group dev ruff #Dev dependencies
# Once groups are defined, the --all-groups, --no-default-groups, --group, --only-group, and --no-group options can be used to include or exclude their dependencies.

uv sync --all-groups
uv sync --dev #equivalent to --group dev

Other

uv run -- flask run -p 3000 # runs flask or arbitrary commands

# Can do many pip commands thought its an implementation not pass through
uv pip install
uv pip freeze:

uvx pycowsay 'hello world!' #runs pip packages like pipx

Venv

python3 -m venv venv
source venv/bin/activate
#....

deactivate

Conda/Mamba

Mamba is a faster drop in replacement for conda (100% faster, takes half the time)

mamba create -n makeittalk_env python=3.6

Conda

Creation:

conda create --name myenv [python=verisonnumber] [package names]
conda create --name myenv python=2.7 scipy

Usage

source activate [envname]

Installation

pip install -r requirements.txt

Inspection:

List all packages

conda list
conda info --envs

List all environments

Removal:

conda remove --name flowers --al

EXPORT:

conda env export > environment.yml

More:

conda env --help
conda create --help

And Jupyter notebook

conda install ipykernel

Poetry

  • Looks like Poetry is a better/fast Pipenv

  • Not recommended sometimes it just doesn't work, but venv does like with dlib

    • ERROR: Could not build wheels for dlib, which is required to install pyproject.toml-based projects

Differences of Poetry vs Pipenv

  • pyproject.toml(PEP spec) vs Pipfile to store more deps info

  • Pipenv has python version management

  • Faster

Installation

curl -sSL https://install.python-poetry.org | python3 -

curl -sSL https://install.python-poetry.org | asdf exec python3 - #if you are using asdf to manage python versions to avoid default pythons installed on MacOS

Setup

Use asdf to manage python versions

asdf global python 3.11.2 #seems to need global setting to desired version
poetry init #setup in project creating pyproject.toml
#may or may not need to set python version in the setup wizard
poetry env use python #might be needed to set version

Pull from existing

cat requirements.txt | xargs poetry add #can then delete requirements.txt

Usage

poetry add pendulum #add new package
poetry add pendulum@~3.1.4 #for specific minor(3.1) version
poetry add pendulum@^3.1.4 #for specific major(3) version

poetry install #install all packages

poetry run python your_script.py #run

poetry shell #to enter the shell
poetry update #equivalent to deleting poetry.lock and rerunning

Make sure to put poetry.lock in VC

poetry env info #show env info

Pipenv

Pipenv vs virtualenv

Benefits of pipenv

  • deterministic install with pipfile.lock

  • Manages environment and pip simuletously

  • Using new commands mean you dont have to remember to pip freeze

  • can have dev and prod envs

  • Easy way to enforce others use the best practice of environments, environments allow different global versions of packages

  • Actually better syntax then virtual environments and pip freeze and requirements.txt

Disadvantages

  • New intrusive commands for installation and running

  • another layer of abstraction using virtualenvs underneath

  • https://github.com/pypa/pipenv/issues/796, if you mv the file you have to reinstall the pipenv :O [fix by having ]

Setup

pip install pipenv

pip install --user pipenv

pipenv installcreate env if one doesn't exist or install everything

pipenv --python 3 to specify version

Usage

pipenv install requests instead of pip install

pipenv install pymongo==3.4.0 for specific version

pipenv run to run stuff in virtual env

pipenv run python instead of python

Deploying

pipenv install --deploy —system

PreviousSecretsNextImporting

Last updated