VS Code & Anaconda Setup

TL;DR

• Editor = VS Code; environments = conda (one isolated env per project).
• Track dependencies in requirements.txt so the setup is reproducible.

A short, practical section: get the local environment ready so the rest of the course runs cleanly.

What it covers

• VS Code as the editor, with the Python extension.
• Anaconda / conda to create isolated environments so each project has its own dependencies and Python version — no global clutter, no version clashes.

The setup, in commands

# create an isolated environment for the course
conda create -n rag python=3.11 -y
conda activate rag

# open the project folder in VS Code
code .

Why isolate environments

Every RAG/agent library pins specific versions. A dedicated environment means one project's langchain version can't break another's, and you can delete the whole environment to start clean. Keep a requirements.txt so the setup is reproducible.

Cheat sheet

Step	Command
Create env	conda create -n rag python=3.11 -y
Activate	conda activate rag
Open editor	code .
Freeze deps	pip freeze > requirements.txt

⚠ Common mistakes

• Installing everything into the base/global environment — versions clash across projects. Always activate a project env first.
• Committing your .env / API keys to git. Keep secrets out of the repo.

Editor = VS Code; environments = conda. One isolated environment per project, tracked in requirements.txt, keeps dependencies from fighting each other.

Quick self-check

Why use a separate conda environment per project?

So one project's library versions can't break another's — and you can delete the env to start clean. requirements.txt makes it reproducible.

Related: Glossary · Next: Data Ingestion & Parsing →