MongoDB flow for transforming JSON docs into pd.DataFrame

mongodb_docs_to_df.py

from pymongo import MongoClient

def get_db():
    # For local use
    from pymongo import MongoClient
    # if your db is run locally
    client = MongoClient(host='localhost', port=27017)
    print(client.list_database_names())
    return client.warehouse

# database
db = get_db()

# collections
collection1 = db.collection1
collection2 = db.collections2

# some aggregation query
# that gets a table of json documents that you need
query = [
    {}, ...
]

# documents
docs = collection1.aggregate(query)

# and transform it into df for easy data wrangling
df = pd.DataFrame()
for doc in docs:
    df = df.append(doc, ignore_index=True)