Skip to content

MongoDB Learning Labs — Jupyter Notebooks (Python/pymongo)

This folder contains Jupyter notebooks for hands-on MongoDB learning using Python and the pymongo driver. Each notebook corresponds to a lab exercise and includes:

  • 📖 Markdown theory cells with links to documentation
  • 🔧 Executable Python/pymongo code cells
  • 🔀 Dual connection support: Docker (local) or MongoDB Atlas (cloud)

Quick Start

1. Install dependencies

pip install -r ../requirements.txt

2. Start MongoDB

Option A: Docker (local 3-node replica set) 

cd ../docker
docker compose up -d
mongosh "mongodb://127.0.0.1:27017/mongo_labs?directConnection=true"

Option B: MongoDB Atlas (cloud) - Get your Atlas connection string - In the first notebook cell, set USE_ATLAS = True and fill in ATLAS_URI

3. Run a notebook

jupyter notebook 01_database_basics.ipynb

Or launch Jupyter Lab: 

jupyter lab

Lab Structure

Lab File Topics
01 01_database_basics.ipynb CRUD, projections, operators, bulk writes
02 02_document_modeling.ipynb Embedding vs referencing, one-to-many, many-to-many
03 03_indexes.ipynb Single-field, compound, unique, multikey, text, TTL, explain()
04+ To be generated Aggregation, transactions, TTL/capped, advanced patterns, replica sets, security, monitoring

Connection Setup

Each notebook's first cell includes automatic connection logic:

USE_ATLAS = False  # Toggle this

# If False: connects to Docker replica set at localhost:27017
# If True: connects to Atlas (set ATLAS_URI first)

Key Differences: JS vs Python

Concept mongosh (JS) pymongo (Python)
Insert db.coll.insertOne({...}) db.coll.insert_one({...})
Find db.coll.findOne(...) db.coll.find_one(...)
Update db.coll.updateOne(filter, {$set:...}) db.coll.update_one(filter, {"$set":...})
Aggregation db.coll.aggregate([...]) db.coll.aggregate([...])
Transactions session.startTransaction() session.start_transaction()
Indexes db.coll.createIndex({field:1}) db.coll.create_index([("field",1)])

Tips

  • Run cells in order — each lab builds on seeded data from previous cells
  • Reset collections — each lab starts with db.collection.drop() before inserting test data
  • Pretty printing — use the pp object (PrettyPrinter) for readable output
  • Error handling — cells include try/except for operations that may fail on second run
  • Connection toggle — change USE_ATLAS and re-run the connection cell to switch environments

Next Steps

  1. Start with Lab 01 (Database Basics) to learn CRUD operations
  2. Progress to Lab 02 (Document Modeling) to understand embedding vs referencing
  3. Master Lab 03 (Indexes) for query performance optimization
  4. Advanced labs (04+) cover pipelines, transactions, sharding, and production patterns

See also: - Theory modules — Comprehensive explanations with diagrams - mongosh labs — Original JavaScript versions - Docker setup — Replica set initialization