Convert PyMongo Cursor to Dataframe

Prerequisites: MongoDB Python Basics

This article is about converting the PyMongo Cursor to Pandas Dataframe. Functions like find() and find_one() returns the Cursor instance.

Let’s begin:

1. Importing Required Modules: Import the required module using the command:

   from pymongo import MongoClient
   from pandas import DataFrame

   If MongoDB is already not installed on your machine you can refer to the guide: Guide to Install MongoDB with Python

   If pandas not install you can install it using pip and if you are using Python3 then use pip3 instead of pip to install the required modules.

   pip install pandas
   

2. Creating a Connection: Now we had already imported the module, its time to establish a connection to the MongoDB server, presumably which is running on localhost (host name) at port 27017 (port number).

   client = MongoClient(‘localhost’, 27017)

3. Accessing the Database: Since the connection to the MongoDB server is established. We can now create or use the existing database.

   mydatabase = client.name_of_the_database

4. Accessing the Collection: We now select the collection from the database using the following syntax:

   collection_name = mydatabase.name_of_collection

5. Getting the documents: Getting all the documents from the collection using find() method. It returns the instance of the Cursor.

   cursor = collection_name.find()
   

6. Converting the Cursor to Dataframe: Converting the Cursor to the Pandas Dataframe.
   First, we convert the cursor to the list of dictionary.

   list_cur = list(cursor)
   

   Now, converting the list to the Dataframe

   df = DataFrame(list_cur)
   

Below is the implementation.

Sample Database:

# Python Program for demonstrating the
# PyMongo Cursor to Pandas DataFrame


# Importing required modules
from pymongo import MongoClient
from pandas import DataFrame


# Connecting to MongoDB server
# client = MongoClient('host_name',
# 'port_number')
client = MongoClient('localhost', 27017)

# Connecting to the database named
# GFG
mydatabase = client.GFG

# Accessing the collection named
# gfg_collection
mycollection = mydatabase.College

# Now creating a Cursor instance
# using find() function
cursor = mycollection.find()
print('Type of cursor:',type(cursor))

# Converting cursor to the list of
# dictionaries
list_cur = list(cursor)

# Converting to the DataFrame
df = DataFrame(list_cur)
print('Type of df:',type(df))

# Printing the df to console
print()
print(df.head())

Output:

Type of cursor: <class 'pymongo.cursor.Cursor'>
Type of df: <class 'pandas.core.frame.DataFrame'>

   _id      name Roll No Branch
0    1  Vishwash    1001    CSE
1    2   Vishesh    1002     IT
2    3    Shivam    1003     ME
3    4      Yash    1004    ECE
4    5      Raju    1005    CSE

Output Explanation:

As seen above when there is no argument is provided it only prints 5 records (numbered 0 to 4…brainfart).And if you put a positive int in the dataframe function, it will generate that many records.