Environment

Environment#

A Python environment is an isolated workspace that contains a specific Python version along with libraries, dependencies, and tools. Managing environments is essential for:

  • Reproducibility of projects

  • Preventing version conflicts

  • Allowing multiple projects with different dependencies

  • Providing a clean and controlled space for development

There are several environment managers and distributions available.
The most popular ones are:

  • Conda

  • Miniconda

  • Mamba

  • Anaconda

  • pip + venv

  • Other Python distributions such as WinPython or ActivePython

1. Conda (Anaconda / Miniconda)#

Conda is one of the most widely used environment and package managers, especially in data science.

🔹 Key Features#

Feature

Conda

Manages Python environments

✔

Installs non-Python packages (C/C++, CUDA, R…)

✔

Works offline once packages are downloaded

✔

Huge scientific stack (NumPy, Pandas, Jupyter…)

✔

🔹 Example Commands#

conda create --name myenv python=3.11
conda activate myenv
conda install numpy pandas
conda list
conda remove --name myenv --all  # delete environment

Conda is recommended for beginners and heavy scientific workloads.

2. Anaconda vs Miniconda#

Feature

Anaconda

Miniconda

Preloaded packages

✔ Many (>150)

✘ Minimal installation only

Size

Large (~3GB installed)

Small (~400MB)

Best for

Beginners, ready-to-use environment

Custom lightweight environments

If you want fast install and flexibility → choose Miniconda
If you want everything pre-installed → choose Anaconda

3. Mamba (fast Conda alternative)#

Mamba is a drop-in replacement for conda, built in C++ and optimized for speed.
It solves dependencies much faster and is ideal for large environments or automated workflows.

Example#

mamba create -n myenv python=3.10
mamba install scipy jupyter

Good for:

Large environment creation

Continuous integration (CI) pipelines

Faster dependency solver than conda

4. pip + venv (lightweight Python standard tool)#

For pure-Python packages or smaller development environments, pip + venv is often sufficient. It comes built into Python, making it an excellent lightweight alternative when full package management like Conda is unnecessary.

Create and Activate an Environment#

python -m venv myenv
source myenv/bin/activate   # Linux/Mac
myenv\Scripts\activate      # Windows
pip install numpy
pip freeze > requirements.txt

Advantages#

  • Built-in, no extra installation needed

  • Simple and lightweight

  • Works well for web development and small projects

Disadvantages#

  • Limited when installing packages that need compiled dependencies

  • Less suited for heavy scientific libraries (TensorFlow, GDAL, CUDA-based tools)

5. Environment Comparison Overview#

Choosing the right environment tool depends on workflow, package complexity, and whether scientific / compiled libraries are required.
Below is a comparison of the most commonly used Python environment managers.

Environment Tool

Primary Use Case

Strengths

Drawbacks

Conda

Data science, ML, scientific computing

Manages compiled packages, strong isolation, cross-platform

Dependency solving can be slow

Mamba

High-performance Conda replacement

Very fast installs and environment solving, conda-compatible

Smaller community and ecosystem size

Anaconda Distribution

All-in-one preconfigured environment

Comes with Jupyter, ML libraries, plotting packages preinstalled

Large installation size (~3GB)

Miniconda

Minimal conda setup for customization

Lightweight, flexible base for building custom environments

Requires users to install everything manually

pip + venv

General Python development

Built-in, lightweight, easy to use

Not ideal for heavy compiled scientific libraries (e.g., TensorFlow, CUDA)
Contents