Querying a MongoDB Database

Introduction to MongoDB in Python

Filip Schouwenaars

Machine Learning Researcher

Databases and collections

Database
- Related collections
- film
Collection
- Documents with flexible schema
- movies

visual

Databases and collections in Python

from pymongo import MongoClient 
client = MongoClient()


# List available databases
client.list_database_names()

['admin', 'config', 'film', 'local']

# List collections on film db
client.film.list_collection_names()

['movies']

Getting all documents

# Set up client
from pymongo import MongoClient 
client = MongoClient()


# Return all documents
client.film.movies.find()


# Avoid dots
mov = client.film.movies

mov.find()

From cursor to Python list

# Set up client
from pymongo import MongoClient 
client = MongoClient()
mov = client.film.movies

# Retrieve all documents
mov.find()

<pymongo.synchronous.cursor.Cursor at 0x7f...760>

list(mov.find())

[{'_id': '68...ff', 'title': 'superbad', ...

Cursor = pointer to query results
Allows fetching results one by one
Control is useful

Let's have a look

# Set up client
from pymongo import MongoClient 
client = MongoClient()
mov = client.film.movies

# find(), cursor to list
list(mov.find())

[
    {
        "_id": "6824bb...e53adbf274ff",
        "title": "superbad",
        "genre": ["comedy", "teen"],
        "release_year": 2007,
        "rating": 7.6
    },
    {
        "_id": "6824bb...e53adbf27500",
        "title": "interstellar",
        "genre": ["adventure", "drama", "sci-fi"],
        "release_year": 2014,
        "rating": 8.6,
        "won_oscar": True
    },
    ...
]

Query filters

from pymongo import MongoClient 
client = MongoClient()
mov = client.film.movies


curs = mov.find({ "won_oscar": True })
list(curs)

[
    {
        '_id': '6824bbd05644e53adbf27500',
        'genre': ['adventure', 'drama', 'sci-fi'],
        'rating': 8.6,
        'release_year': 2014,
        'title': 'interstellar',
        'won_oscar': True
    },
     {
        '_id': '6824bbd05644e53adbf27501',
        'genre': ['action', 'sci-fi', 'thriller'],
        'rating': 8.8,
        'release_year': 2010,
        'title': 'inception',
        'won_oscar': True
    },
    ...
]

Target specific records with .find_one()

from pymongo import MongoClient 

client = MongoClient()
mov = client.film.movies


mov.find_one({ "title": "parasite" })

{
    '_id': '6824bbd05644e53adbf27518',
    'genre': ['drama', 'thriller'],
    'rating': 8.5,
    'release_year': 2019,
    'title': 'parasite',
    'won_oscar': True
}

Return first document that matches query
Returns document, not Cursor object

Let's practice!

Introduction to MongoDB in Python