conda and Anaconda¶

conda is a cross-platform package and environment manager. It can install Python packages as well as non-Python dependencies (C libraries, compilers, etc.).

Anaconda vs Miniconda¶

Distribution	Size	Includes
Anaconda	~3 GB	Python + 250+ packages (NumPy, Pandas, Jupyter, etc.)
Miniconda	~50 MB	Python + conda only

Recommendation: Start with Miniconda and install only what you need.

Installation¶

Miniconda¶

Download from: https://docs.conda.io/en/latest/miniconda.html

# Linux
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

# macOS
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh
bash Miniconda3-latest-MacOSX-x86_64.sh

Verify Installation¶

conda --version
conda info

Environment Management¶

Create Environment¶

conda create -n myenv python=3.11
conda create -n dataenv python=3.11 numpy pandas matplotlib

Activate / Deactivate¶

conda activate myenv
# ... do work ...
conda deactivate

List Environments¶

conda env list
conda info --envs

Remove Environment¶

conda env remove -n myenv

Clone Environment¶

conda create -n newenv --clone oldenv

Package Management¶

Install Packages¶

conda install numpy
conda install numpy pandas scikit-learn
conda install numpy=1.24.0              # Specific version

Update Packages¶

conda update numpy
conda update --all                      # Update all packages

Remove Packages¶

conda remove numpy

Search Packages¶

conda search numpy
conda search "numpy>=1.20"

List Installed Packages¶

conda list
conda list numpy                        # Filter by name

Channels¶

Channels are repositories where conda looks for packages.

Default Channel¶

Anaconda's default channel (defaults) is managed by Anaconda Inc.

Add Channel¶

conda config --add channels conda-forge
conda config --set channel_priority strict

Install from Specific Channel¶

conda install -c conda-forge numpy
conda install -c pytorch pytorch

View Channels¶

conda config --show channels

Environment Files¶

Export Environment¶

conda env export > environment.yml
conda env export --no-builds > environment.yml  # More portable

Example environment.yml:

name: myenv
channels:
  - conda-forge
  - defaults
dependencies:
  - python=3.11
  - numpy=1.24
  - pandas=2.0
  - pip
  - pip:
    - some-pip-only-package

Create from File¶

conda env create -f environment.yml

Update from File¶

conda env update -f environment.yml

conda vs pip¶

Feature	conda	pip
Package source	Anaconda repo / conda-forge	PyPI
Non-Python deps	✅ Yes	❌ No
Environment mgmt	✅ Built-in	❌ Needs venv
Dependency solver	SAT solver	Basic resolver
Binary packages	Pre-compiled	Wheels or source
Speed	Slower	Faster

Using pip with conda¶

You can use pip inside a conda environment:

conda activate myenv
conda install pip
pip install some-package-not-on-conda

Best practice: Install conda packages first, then pip packages.

# environment.yml
dependencies:
  - numpy
  - pandas
  - pip
  - pip:
    - package-only-on-pypi

Anaconda Licensing¶

⚠️ Important: Since 2020, Anaconda's terms of service require a paid license for commercial use (organizations with 200+ employees).

Free alternatives: - conda-forge (community channel) - Miniforge (conda-forge default installer) - Mambaforge (Miniforge + mamba)

See the conda-forge and Miniforge page for details.

Common Commands¶

Command	Description
`conda create -n env python=3.11`	Create environment
`conda activate env`	Activate environment
`conda deactivate`	Deactivate environment
`conda install pkg`	Install package
`conda update pkg`	Update package
`conda remove pkg`	Remove package
`conda list`	List packages
`conda env list`	List environments
`conda env export > env.yml`	Export environment
`conda env create -f env.yml`	Create from file
`conda clean --all`	Remove unused packages/caches

Configuration¶

View Config¶

conda config --show

Common Settings¶

# Always activate base environment
conda config --set auto_activate_base true

# Don't activate base by default
conda config --set auto_activate_base false

# Set channel priority
conda config --set channel_priority strict

# Add conda-forge
conda config --add channels conda-forge

.condarc File¶

# ~/.condarc
channels:
  - conda-forge
  - defaults
auto_activate_base: false
channel_priority: strict

Key Takeaways¶

Miniconda is lighter than Anaconda
conda manages both Python packages and system dependencies
Use environments to isolate projects
Channels are package repositories (defaults, conda-forge)
Can use pip inside conda environments (conda packages first)
Check licensing for commercial use — consider conda-forge/Miniforge