24. How to Load SQL Query Results into a pandas DataFrame

24.7 Practice, Tip and Close

By Ravindra Bagale · Last updated: 23 October 2026

Why. Connection + query + DataFrame must feel boringly repeatable before interview day (Chapter 18) or project week (Chapter 16 style refreshes).

What. One lab: build SQLite orders, pull with parameters, summarise in pandas, export a tiny CSV (Chapter 22 skill).

How.

import sqlite3
from pathlib import Path

import pandas as pd
from sqlalchemy import create_engine

Path("data").mkdir(exist_ok=True)
Path("outputs").mkdir(exist_ok=True)

# If you already ran 24.3, you can skip recreate — or rebuild for a clean lab.
con = sqlite3.connect("data/blinkit_lab.db")
df = pd.read_sql_query(
    """
    SELECT city, amount_inr
    FROM orders
    WHERE status = ? AND city IN (?, ?)
    """,
    con,
    params=("Delivered", "Pune", "Nashik"),
)
con.close()

summary = (
    df.groupby("city", as_index=False)["amount_inr"]
    .agg(orders="count", gmv="sum", median_basket="median")
)
summary.to_csv("outputs/sql_pull_city_gmv.csv", index=False)
print(summary)

engine = create_engine("sqlite:///data/blinkit_lab.db")
smoke = pd.read_sql_query("SELECT COUNT(*) AS n FROM orders", engine)
print(smoke)
engine.dispose()

Samajla ka? Aata SQL result seedha DataFrame madhe yenar — pudhe practice regular theva, lab DB var ek query/day.