3 SQL Techniques
Every Data
Scientist Needs
to Know for
Faster Queries
Sluggish Queries?
Turn them Lightning-Fast with:
✔️ Indexing
✔️ Partitioning
✔️ Window Functions
@varshacbendre
The Struggle 😫
Slow queries that:
❌ Slow: Takes forever
❌ Inefficient: Large datasets
❌ Complex: Hard-to-read code
The Solution 👩💻
Advanced SQL = Better Queries
✅ Efficient: Faster execution
✅ Scalable: Handles large datasets
✅ Simple: Easy to understand
@varshacbendre
Indexing
A Shortcut for Your Queries
What it does
Speeds up row retrieval,
like an index in a book.
When to use
Searching/filtering on
frequently queried
columns.
Impact
Dramatically improves
SELECT query
performance.
@varshacbendre
Before vs After Indexing
QUERY: Find orders for customer_id = 123
BEFORE INDEXING
Full table scan (10M rows)
Time: 3.2 seconds
Scans ENTIRE table
AFTER INDEXING
Targeted row retrieval
Time: 0.03 seconds
Precise data path
CREATE INDEX Code
CREATE INDEX idx_customer_id
ON orders(customer_id);
Performance Boost
99% faster queries
Reduced overhead
Minimal storage impact
@varshacbendre
Partitioning
Slice Tables for Speed
What it does
Divides tables (e.g., by
date, region)
Boosts performance &
management
When to use
Specific data queries
Logical table divisions
Impact
Faster queries
Reduced complexity
Easier archiving
@varshacbendre
Before vs After Partitioning
QUERY: Analyze data for year = 2023
BEFORE PARTITIONING:
Full table scan (500M rows) ⏳ 12.5
seconds
AFTER PARTITIONING (BY YEAR):
Scans only 2023 data ✨
0.9 seconds
CREATE PARTITION Code:
CREATE PARTITION FUNCTION YearPF (datetime)
AS RANGE RIGHT FOR VALUES
('2022-01-01', '2023-01-01', '2024-01-01');
PERFORMANCE BOOST:
92% faster queries
Simplified data management
Better scalability
@varshacbendre
Window Functions
What it does
Perform calculations over
a set of rows
Examples: Ranking,
moving averages
When to use
Complex aggregations
(no GROUP BY)
Time-series analysis
Comparative calculations
Impact
Improved performance
Fewer subqueries
@varshacbendre
Before vs After Window
Functions
Example: 7-Day Sales Average 📊
Before:
SELECT date, sales,
(SELECT AVG(sales)
FROM sales s2
WHERE s2.date BETWEEN s1.date - 6 AND
s1.date) AS avg_7day
FROM sales s1;
⏳ Execution Time: ~4.2s
After:
SELECT date, sales,
AVG(sales) OVER (
ORDER BY dateROWS BETWEEN 6
PRECEDING AND CURRENT ROW
) AS avg_7day
FROM sales;
✨ Execution Time: ~0.4s
Before vs After Window
Functions
Performance Boost:
90% faster
Simplified structure
Pro Tip: Combine with PARTITION BY
for group-based calculations!
@varshacbendre
⚡ Recap: The Power of
Advanced SQL
1️⃣ Indexing: Locate rows faster with
smart shortcuts.
2️⃣ Partitioning: Optimize queries on
massive datasets.
3️⃣ Window Functions: Simplify
complex calculations.
🚀 These techniques take you from
struggling to scaling with SQL.
@varshacbendre
Ready to Elevate Your Data
Science and AI Journey?
Follow for Daily Insights
and Expert Tips!
@varshacbendre Save
