Joining in Snowflake

INTRODUCTION TO SNOWFLAKE

Palak Raina
Senior Data Engineer
JOINS
INNER JOIN
OUTER JOINS
LEFT OUTER JOIN or LEFT JOIN

RIGHT OUTER JOIN or RIGHT JOIN

FULL OUTER JOIN or FULL JOIN

CROSS JOINS

SELF JOINS

NATURAL JOIN

LATERAL JOIN

INTRODUCTION TO SNOWFLAKE
Pizza dataset

INTRODUCTION TO SNOWFLAKE
Similarities with PostgreSQL - INNER JOIN
SELECT
pt.name AS pizza_type_name,
pt.category,
pt.ingredients,
p.size,
p.price
FROM
pizza_type AS pt
INNER JOIN -- Using INNER JOIN
pizzas AS p
ON
pt.pizza_type_id = p.pizza_type_id

INTRODUCTION TO SNOWFLAKE
NATURAL JOIN
Syntax:

SELECT ...
FROM <table_one> [
{
| NATURAL [ { LEFT | RIGHT | FULL } [ OUTER ] ]
}
]
JOIN <table_two>
[ ... ]

INTRODUCTION TO SNOWFLAKE
NATURAL JOIN
Without NATURAL JOIN With NATURAL JOIN

SELECT * SELECT *
FROM pizzas AS p FROM pizzas AS p
JOIN pizza_type AS t NATURAL JOIN pizza_type AS t
ON t.pizza_type_id = p.pizza_type_id

INTRODUCTION TO SNOWFLAKE
NATURAL JOIN

NOT ALLOWED

select *
FROM pizzas AS p
NATURAL JOIN pizza_type AS t
ON t.pizza_type_id = p.pizza_type_id

INTRODUCTION TO SNOWFLAKE
NATURAL JOIN
ALLOWED

WHERE clause

SELECT *
FROM pizzas AS p
NATURAL JOIN pizza_type AS t
WHERE pizza_type_id = 'bbq_ckn'

INTRODUCTION TO SNOWFLAKE
LATERAL JOIN
Syntax:

SELECT ...
FROM <left_hand_expression> , --
LATERAL
(<right_hand_expression>)

left_hand_expression - Table, view, or subquery

right_hand_expression - Inline view or subquery

INTRODUCTION TO SNOWFLAKE
LATERAL JOIN with a subquery
SELECT
p.pizza_id,
lat.name,
lat.category
FROM pizzas AS p,
LATERAL -- Keyword LATERAL
( SELECT *
FROM pizza_type AS t
-- Referencing outer query column: p.pizza_type_id
WHERE p.pizza_type_id = t.pizza_type_id
) AS lat

INTRODUCTION TO SNOWFLAKE
Why LATERAL JOIN?
SELECT
*
FROM orders AS o,
LATERAL (
-- Subquery calculating total_spent
SELECT
SUM(p.price * od.quantity) AS total_spent
FROM order_details AS od
JOIN pizzas AS p
ON od.pizza_id = p.pizza_id
WHERE o.order_id = od.order_id
) AS t
ORDER BY o.order_id

INTRODUCTION TO SNOWFLAKE
 Let's practice!
INTRODUCTION TO SNOWFLAKE
 Subquerying and
Common Table
Expressions
INTRODUCTION TO SNOWFLAKE

Palak Raina
Senior Data Engineer
Subquerying
Nested queries
Used in FROM , WHERE , HAVING or SELECT clauses

Example:

SELECT column1
FROM table1
WHERE column1 = (SELECT column2 FROM table2 WHERE condition)

Types: Correlated and uncorrelated subqueries

INTRODUCTION TO SNOWFLAKE
Correlated subquery
References columns from the outer query

SELECT pt.name,
pz.price,
pt.category
FROM pizzas AS pz
JOIN pizza_type AS pt
ON pz.pizza_type_id = pt.pizza_type_id
WHERE pz.price < (
-- Identifies highest price for each piza category
SELECT MAX(p2.price) -- Max price
FROM pizzas AS p2
WHERE -- Correlated: uses outer query column
p2.pizza_type_id = pz.pizza_type_id
)

INTRODUCTION TO SNOWFLAKE
Uncorrelated subquery
No reference to outer or main query

SELECT order_id
FROM order_details AS od
WHERE pizza_id = (
-- Uncorrelated: standalone subquery
-- Not referencing to outer query columns
SELECT pizza_id
FROM pizzas
ORDER BY price DESC
LIMIT 1
)

INTRODUCTION TO SNOWFLAKE
Correlated subquery limitations
Can't use LIMIT with corelated subquery

SELECT pt.name,
pz.price,
pt.category
FROM pizzas AS pz
JOIN pizza_type AS pt
ON pz.pizza_type_id = pt.pizza_type_id
WHERE pz.price < (
SELECT p2.price -- Get price
FROM pizzas AS p2
WHERE p2.pizza_type_id = pz.pizza_type_id -- Correlated to outer query
ORDER BY p2.price DESC -- Order with max price first
LIMIT 1 -- LIMIT 1 fetches only the top record, i.e., the max price
)

INTRODUCTION TO SNOWFLAKE
Correlated subquery limitations
Result:

INTRODUCTION TO SNOWFLAKE
Using LIMIT in uncorrelated subquery
SELECT pt.name,
Result:
pz.price,
pt.category
FROM pizzas AS pz
JOIN pizza_type AS pt
ON pz.pizza_type_id = pt.pizza_type_id
WHERE pz.price < (
SELECT p2.price
FROM pizzas AS p2 -- No where clause
ORDER BY p2.price DESC
-- Can use limit here
LIMIT 1
)

INTRODUCTION TO SNOWFLAKE
Common Table Expressions
Basic Syntax:

-- WITH keyword
WITH cte1 AS ( -- CTE name
SELECT col_1, col_2
FROM table1
)
...
SELECT ...
FROM cte1 -- Query CTE
;

INTRODUCTION TO SNOWFLAKE
Common Table Expressions
WITH max_price AS ( -- CTE:max_price
SELECT pizza_type_id,
MAX(price) AS max_price
FROM pizzas
GROUP BY pizza_type_id
)
-- Main query
SELECT pt.name,
pz.price,
pt.category
FROM pizzas AS pz
JOIN pizza_type AS pt ON pz.pizza_type_id = pt.pizza_type_id
JOIN max_price AS mp -- Joining with CTE:max_price
ON pt.pizza_type_id = mp.pizza_type_id
WHERE pz.price < mp.max_price -- Compare the price with max_price CTE column

INTRODUCTION TO SNOWFLAKE
Multiple CTEs
-- Define multiple CTEs separated by commas
WITH cte1 AS (
SELECT ...
FROM ...
),
cte2 AS (
SELECT ...
FROM ...
)
-- Main query combining both CTEs
SELECT ...
FROM cte1
JOIN cte2 ON ...
WHERE ...

INTRODUCTION TO SNOWFLAKE
Why Use CTEs?
Managing complex operations
Readable

Modular

Reusable

INTRODUCTION TO SNOWFLAKE
 Let's practice!
INTRODUCTION TO SNOWFLAKE
 Snowflake Query
Optimization
INTRODUCTION TO SNOWFLAKE

Palak Raina
Senior Data Engineer
What's Snowflake Query Optimization?
Transforming into more efficient queries
Snowflake's Cloud Services Layer

INTRODUCTION TO SNOWFLAKE
Why Optimize Queries in Snowflake?
Achieve faster results
Cost efficiency
Shorter query times consumes fewer resources like CPU and memory.

INTRODUCTION TO SNOWFLAKE
Common query problems
Exploding Joins: Be cautious!
Incorrect:

SELECT *
FROM order_details AS od
JOIN pizzas AS p -- Missing ON condition leading to exploding joins

INTRODUCTION TO SNOWFLAKE
Common query problems
Exploding Joins: Be cautious!
Correct:

SELECT *
FROM order_details AS od
JOIN pizzas AS p
ON od.pizza_id = p.pizza_id

INTRODUCTION TO SNOWFLAKE
Common query problems
UNION or UNION ALL : Know the difference.
UNION removes duplicates, slows down the query.

UNION ALL is faster if no duplicates.

Handling big data

Use filters to narrow down data.

Apply limits for quicker results.

INTRODUCTION TO SNOWFLAKE
How to optimize queries?
SELECT * SELECT TOP 10*
SELECT SELECT
* TOP 10*
FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS

INTRODUCTION TO SNOWFLAKE
How to optimize queries?
SELECT * LIMIT
SELECT * SELECT *
FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS
LIMIT 10

INTRODUCTION TO SNOWFLAKE
How to optimize queries?
SELECT * Avoid SELECT *
FROM SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS
SELECT
o_orderdate,
o_orderstatus
FROM
SNOWFLAKE_SAMPLE_DATA.TPCH_SF100.ORDERS

INTRODUCTION TO SNOWFLAKE
How to optimize queries?
Filter Early
Use WHERE Clause Early On.

Apply filters before JOIN s

INTRODUCTION TO SNOWFLAKE
Without early filtering
SELECT orders.order_id,
orders.order_date,
pizza_type.name,
pizzas.pizza_size
FROM orders
JOIN order_details
ON orders.order_id = order_details.order_id
JOIN pizzas
ON order_details.pizza_id = pizzas.pizza_id
JOIN pizza_type
ON pizzas.pizza_type_id = pizza_type.pizza_type_id
WHERE orders.order_date = '2015-01-01'; -- Filtering after JOIN

INTRODUCTION TO SNOWFLAKE
With early filtering
WITH filtered_orders AS (
SELECT *
FROM orders
WHERE order_date = '2015-01-01' -- Filtering in CTE before JOIN
)
SELECT filtered_orders.order_id,
filtered_orders.order_date,
pizza_type.name,
pizzas.pizza_size
FROM filtered_orders -- Joining with CTE
JOIN order_details
ON filtered_orders.order_id = order_details.order_id
JOIN pizzas

INTRODUCTION TO SNOWFLAKE
Query history
Query History
snowflake.account_usage.query_history

SELECT
query_text,
start_time,
end_time,
execution_time
FROM
snowflake.account_usage.query_history
WHERE query_text ilike '%order_details%'

INTRODUCTION TO SNOWFLAKE
Query history
Spot slow or frequently running queries

SELECT
query_text,
start_time,
end_time,
execution_time
FROM
snowflake.account_usage.query_history
WHERE
execution_time > 1000

INTRODUCTION TO SNOWFLAKE
 Let's practice!
INTRODUCTION TO SNOWFLAKE
 Handling semi-
structured data
INTRODUCTION TO SNOWFLAKE

Palak Raina
Senior Data Engineer
Structured versus semi-structured
Example of structured data Example of semi-structured data

| cust_id | cust_name | cust_age | cust_email |

|---------|-----------|----------|-----------------------|
| 1 | cust1 | 40 | cust1***@gmail.com |
| 2 | cust2 | 35 | cust2***@gmail.com |
| 3 | cust3 | 42 | cust3***@gmail.com |

INTRODUCTION TO SNOWFLAKE
Introducing JSON
JavaScript Object Notation
Common use cases: Web APIs, Mobile Apps, Config files.

JSON data structure:

Key-Value Pairs, e.g., cust_id: 1

INTRODUCTION TO SNOWFLAKE
JSON in Snowflake
Native JSON support
Flexible for evolving schemas

Comparisons:

Postgres: Uses JSONB

Snowflake: Uses VARIANT

INTRODUCTION TO SNOWFLAKE
How Snowflake stores JSON data
VARIANT supports OBJECT and ARRAY data types
OBJECT: { "key": "value"}

ARRAY: ["list", "of", "values"]

Creating a Snowflake Table to handle JSON data

CREATE TABLE cust_info_json_data(

customer_id INT,
customer_info VARIANT -- VARIANT data type
);

INTRODUCTION TO SNOWFLAKE
Semi-structured data functions
PARSE_JSON
expr : JSON data in string format.

Returns: VARIANT type, valid JSON

object.

INTRODUCTION TO SNOWFLAKE
PARSE_JSON
Example:

SELECT PARSE_JSON(
' -- enclosed in strings
{
"cust_id": 1,
"cust_name": "cust1",
"cust_age": 40,
"cust_email":"cust1***@gmail.com"
}
'-- enclosed in strings
) AS customer_info_json

INTRODUCTION TO SNOWFLAKE
OBJECT_CONSTRUCT
OBJECT_CONSTRUCT
Syntax: OBJECT_CONSTRUCT( [<key1>, <value1> [, <keyN>, <valueN> ...]] )

Returns: JSON object

SELECT OBJECT_CONSTRUCT(
'cust_id', 1,
'cust_name', 'cust1',
'cust_age', 40,
'cust_email', 'cust1***@gmail.com'
)

Customer info data stored in table

INTRODUCTION TO SNOWFLAKE
Querying JSON data in Snowflake
Simple JSON Data

SELECT
customer_info:cust_age, -- using colon to access data from column
customer_info:cust_name,
customer_info:cust_email,
FROM
cust_info_json_data;

INTRODUCTION TO SNOWFLAKE
Querying nested JSON Data in Snowflake
Example of nested object

Colon: :

Dot: .

INTRODUCTION TO SNOWFLAKE
Querying nested JSON using colon/dot notations
Accessing values using colon notation Accessing values using dot notation

<column>:<level1_element>: <column>:<level1_element>.
<level2_element>:<level3_element> <level2_element>.<level3_element>

SELECT SELECT
customer_info:address:street AS street_name customer_info:address.street AS street_name
FROM FROM
cust_info_json_data cust_info_json_data

INTRODUCTION TO SNOWFLAKE
 Let's practice!
INTRODUCTION TO SNOWFLAKE
 Wrap-up
INTRODUCTION TO SNOWFLAKE

Palak Raina
Senior Data Engineer
Journey

INTRODUCTION TO SNOWFLAKE
Chapter 1: Architecture, Competitors, and
SnowflakeSQL

INTRODUCTION TO SNOWFLAKE
Chapter 1: Architecture, Competitors, and Snowflake
SQL

INTRODUCTION TO SNOWFLAKE
Chapter2: Snowflake SQL and key concepts
Connecting to Snowflake

WEB UI

Drivers & Connectors

SnowSQL

INTRODUCTION TO SNOWFLAKE
Chapter2: Snowflake SQL and key concepts
Data Definition Language (DDL) Database structures and DML commands

CREATE SHOW

ALTER DESCRIBE

DROP INSERT

RENAME UPDATE

COMMENT MERGE

COPY

INTRODUCTION TO SNOWFLAKE
Chapter2: Snowflake SQL and key concepts
VARCHAR STRING functions: CONCAT , UPPER , LOWER
NUMERIC
DATE & TIME functions: CURRENT_DATE ,
INT CURRENT_TIME

DATE
EXTRACT functions: GROUP BY ALL
TIME

TIMESTAMP

VARIANT -> Semi-structured data

Data Type conversion - What, Why, How?

Conversion Fucntions: TO_VARCHAR , TO_DATE

INTRODUCTION TO SNOWFLAKE
Chapter 3: Advance Snowflake SQL Concepts
JOINS

NATURAL JOIN

LATERAL JOIN

Subquerying

CTEs

INTRODUCTION TO SNOWFLAKE
Chapter 3: Advance Snowflake SQL Concepts
Snowflake Query Optimization

Common query problems: Exploding Joins, UNION vs UNION ALL

Rewriting queries: TOP , LIMIT , Early filtering, Avoid Select *`

Semi structured data

PARSE_JSON , OBJECT_CONSTRUCT

Querying JSON data in Snowflake

INTRODUCTION TO SNOWFLAKE
Is this all?
Much more to unfold.
Not addressed

Setting context
Roles, Users

Setting up Virtual Warehouses

Window functions

Query profiling

Materialized Views

Clustering
...

INTRODUCTION TO SNOWFLAKE
Useful resources
Snowflake documentation: https://docs.snowflake.com/
Snowflake forums: https://community.snowflake.com/s/forum

INTRODUCTION TO SNOWFLAKE
This is just the
beginning!
INTRODUCTION TO SNOWFLAKE