example-python-read-and-write-from-mongo

import pandas as pd
from pymongo import MongoClient
import os

# ====== Connection ====== #
# Connection to Mongo
client_mongo = MongoClient(os.environ['IP_MONGO'],27017)
# Connection to the database
db = client_mongo.sandbox
# Authenticating to the database
db.authenticate(os.environ['MONGO_USER'],os.environ['MONGO_PASSWORD'])
# Connection to the collection
collection = db.helloworld

# ====== Inserting Documents ====== #
# Creating a simple Pandas DataFrame
liste_hello = ['hello1','hello2']
liste_world = ['world1','world2']
df = pd.DataFrame(data = {'hello' : liste_hello, 'world': liste_world})
# Bulk inserting documents. Each row in the DataFrame will be a document in Mongo
db.insert_many(df.to_dict('records'))

# ====== Finding Documents ====== #
documents = collection.find({'message': 'helloworld1'})
df = pd.DataFrame(list(documents))

Tried this out & it's awesome!!!! One correction though

collection.insert_many(df.to_dict('records'))

or

db.helloworld.insert_many(df.to_dict('records'))

or

db.['helloworld'].insert_many(df.to_dict('records'))

Also, if you don't want to insert duplicates. You need to create a unique index on a particular "KEY" How to create unique Index in Mongo DB and insert as follows

Create index in mongo CLI making the hello "Key" to be unique

db.helloworld.ensureIndex( { hello: 1 }, { unique: true, sparse: true } )

Insert ignoring duplicates

# Other imports
from pymongo.errors import BulkWriteError

# code omitted for brevity 
try:
    db.helloworld.insert_many(df.to_dict('records'), ordered=False)
except BulkWriteError as e:
    pass

Duplicates will throw an exception but the new records will be inserted. You can safely ignore the error thrown by duplicate records