Pandas DataFrame to SQL

In this short guide, you’ll see how to get from Pandas DataFrame to SQL.

Steps

Step 1: Create a DataFrame

To start, create a DataFrame with the following data about products:

import pandas as pd

data = {
    "product_name": ["Computer", "Tablet", "Monitor", "Printer"],
    "price": [900, 300, 450, 150],
}

df = pd.DataFrame(data)

print(df)

Run the code, and you’ll get the following DataFrame:

  product_name  price
0     Computer    900
1       Tablet    300
2      Monitor    450
3      Printer    150

Step 2: Create a Database and a Table

For demonstration purposes, create a simple database and a table using sqlite3, where:

The database name will be: test_database
The table name will be: products

The “products” table will be used to store the information from the DataFrame.

Here is the full Python script to create the test_database and the products table:

import sqlite3

conn = sqlite3.connect("test_database")
c = conn.cursor()

c.execute("CREATE TABLE IF NOT EXISTS products (product_name text, price number)")
conn.commit()

Step 3: Get from Pandas DataFrame to SQL

You can use the following syntax to get from Pandas DataFrame to SQL:

df.to_sql("products", conn, if_exists="replace", index=False)

Here is the full Python code to get from Pandas DataFrame to SQL:

import pandas as pd
import sqlite3

conn = sqlite3.connect("test_database")
c = conn.cursor()

c.execute("CREATE TABLE IF NOT EXISTS products (product_name text, price number)")
conn.commit()

data = {
    "product_name": ["Computer", "Tablet", "Monitor", "Printer"],
    "price": [900, 300, 450, 150],
}

df = pd.DataFrame(data)

df.to_sql("products", conn, if_exists="replace", index=False)

c.execute(
    """  
    SELECT * FROM products
    """
)

for row in c.fetchall():
    print(row)

Run the code, and you’ll get the following results:

('Computer', 900)
('Tablet', 300)
('Monitor', 450)
('Printer', 150)

Going from the DataFrame to SQL and then back to the DataFrame

For this example, you’ll create a new database called test_database_2, and then find the maximum price among all the products:

import pandas as pd
import sqlite3

conn = sqlite3.connect('test_database_2')
c = conn.cursor()

c.execute('CREATE TABLE IF NOT EXISTS products (product_name text, price number)')
conn.commit()

data = {
    "product_name": ["Computer", "Tablet", "Monitor", "Printer"],
    "price": [900, 300, 450, 150],
}

df = pd.DataFrame(data)

df.to_sql("products", conn, if_exists="replace", index=False)

c.execute(
    """  
    SELECT * FROM products
    WHERE price = (SELECT max(price) FROM products)
    """
)

df = pd.DataFrame(c.fetchall(), columns=["product_name", "price"])

print(df)

Once you run the code in Python, you’ll get the product with the maximum price:

  product_name  price
0     Computer    900