What is Conda

Conda is an open-source cross-platform package, dependency and environment manager. Born for Python data science, it now manages binary dependencies of any language (e.g. CUDA, MKL, compilers) — the key difference from pure-Python toolchains.

Why Conda

In Python development, "works on my machine" is equally painful: different projects need different library versions, even different Python interpreters; some scientific libraries (e.g. NumPy) depend on specific C/Fortran runtimes. Conda solves this with isolated environments + binary packages:

• Isolation: each environment has its own Python interpreter and package directory,互不干扰.
• Binary distribution: prebuilt binaries avoid local compilation — faster and more reliable installs.

Anaconda vs Miniconda

Both run the same Conda engine; they differ in preinstalled content:

Item	Anaconda	Miniconda
Size	~3–5 GB	~100 MB
Preinstalled	Conda + 1500+ scientific packages (NumPy, Pandas, Jupyter…)	Only Conda + Python
Best for	Beginners, out-of-the-box use	Lightweight, on-demand, CI / containers
Startup	Slower (many packages)	Fast

💡 Recommendation: prefer Miniconda for daily dev and servers, installing on demand via conda install; choose Anaconda for teaching or a quick trial.

Core components

Conda uses a local-repo + channel architecture:

• Conda CLI: the conda commands you run.
• Environment: isolated workspaces, each with its own interpreter and packages.
• Channel: the download source for packages, defaulting to defaults on Anaconda.org; the community favours conda-forge.

tip

Typical workflow

Install Miniconda → conda create new env → conda activate
                                          ↓
                       conda install pkg / pip install PyPI pkg

Next, read Core Concepts to understand environments, channels and the relation with pip.