SQL

INTERVIEW
CHEAT SHEET
 SQL Interview Cheat Sheet

There are three main question types in a SQL interview:

1. Basic Functions

2. Complex queries for calculating metrics

3. Conceptual questions

This summary walks you through all three types of questions when preparing for a SQL Interview.

This quick reference lists common pitfalls and important details to keep in mind when working with
SQL, making it easier for you to avoid mistakes and write bug-free queries.

Practice makes perfect! Speed matters!

Quick Checks
Keywords
Order of Statements
Table and Column Names
Parentheses
Aliases
Single Quotes
Carefully use Indentation & White spaces
Go for the ANSI-92 JOIN Syntax (explicit join)
Basic Functions
SELECT
CASE WHEN
ROUND
JOIN
Multiple Filters
LIKE
<>
GROUP BY

SQL Interview Cheat Sheet 1

 ORDER BY
LIMIT N
UNION and UNION ALL
Complex Queries
DATE Format
Window function
Common Table Expression (CTE)
Conceptual Questions
How to write efficient SQL query
Primary key vs foreign key
[Advanced] SQL Rewrite in Pandas Equivalently

Quick Checks
Keywords
Check if the query contains all the necessary keywords

1. SELECT

2. FROM

3. WHERE

4. GROUP BY

5. HAVING

6. ORDER BY

Order of Statements
The order of the statement is the following. Mixing up the statement order will result in a syntax error.

1. SELECT

2. FROM

3. WHERE

4. GROUP BY

5. HAVING

6. ORDER BY

The most common mistake is to have WHERE after GROUP BY .

// Correct Syntax // Wrong Syntax

SELECT SELECT
COUNT(*) COUNT(*)
FROM table1 FROM table1
WHERE date = '2022-01-01' GROUP BY col2

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 2

 GROUP BY col2 WHERE date = '2022-01-01'
HAVING COUNT(*) > 100; HAVING COUNT(*) > 100;

Table and Column Names

Make sure the table and columns are EXACTLY the same as in the problem statement, including
CAPITALIZATION.

Pro tip: copy and paste the table and column names to avoid typos

Table: dropoff_records

id dropoff_date
// Wrong table and column names
1 2022-01-01
SELECT dropoff_data
2 2022-01-02 FROM dropoff_record;

… …

Parentheses
Check if ALL parentheses are paired.

Can you spot the mistake in the query?

SELECT p.name, d.department_count, s.total_sales

FROM persons AS p
JOIN (
SELECT
department_id,
COUNT(people) AS department_count
FROM department
GROUP BY department_id
) AS d
ON d.department_id = p.department_id
JOIN (
SELECT
person_id,
SUM(sales) AS total_sales
FROM orders
GROUP BY person_id
AS s
ON s.person_id = p.person_id;

Aliases
Don’t forget to create aliases when tables have the same names, i.e. when self-joining tables.

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 3

 // Correct Syntax // Wrong Syntax

SELECT SELECT
* *
FROM FROM
table1 AS today, table1,
table1 AS yesterday; table1;

Single Quotes
Use single quotes in queries for strings.

[S]ingle quotes are for [S]tring Literals (date literals are also strings).

[D]ouble quotes are for [D]atabase Identifiers.

// Correct Syntax // Wrong Syntax

SELECT * SELECT *
FROM table1 FROM table1
WHERE date = '2022-01-01' WHERE date = "2022-01-01"
AND status = 'ready'; AND status = "ready";

Carefully use Indentation & White spaces

Ident after a keyword, and when you use a subquery or a derived table.

// Correct Syntax // Correct Syntax but low readability

SELECT c.id, SELECT c.id, c.name, p.date

c.name, FROM customers c
p.date LEFT JOIN ( SELECT customer_id, MIN(date) as date
FROM customers c FROM purchases GROUP BY customer_id ) AS p
LEFT JOIN ( ON p.customer_id = c.id
SELECT customer_id, WHERE c.age<=30;
MIN(date) as date
FROM purchases
GROUP BY customer_id
) AS p
ON p.customer_id = c.id
WHERE c.age <= 30;

Go for the ANSI-92 JOIN Syntax (explicit join)

// Correct Syntax // Wrong Syntax

SELECT c.id, SELECT c.id,

c.name, c.name,
COUNT(t.id) as count_transactions COUNT(t.id) as count_transactions
FROM customers c FROM customers c, transactions t
JOIN transactions t ON c.id = t.customer_id WHERE c.id = t.customer_id

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 4

 WHERE c.age <= 30 AND c.age <= 30
GROUP BY 1,2; GROUP BY 1,2;

Basic Functions
SELECT
Add commas between columns. No comma after the last column.

// Correct Syntax // Wrong Syntax

SELECT SELECT
col1, col1
col2, col2,
col3 col3,
FROM table1; FROM table1;

CASE WHEN
Don’t forget the END keyword at the end.

// Correct Syntax // Wrong Syntax

SELECT SELECT
CASE CASE
WHEN score > 95 THEN 'Excellent' WHEN score > 95 THEN 'Excellent'
WHEN score > 80 THEN 'Good' WHEN score > 80 THEN 'Good'
WHEN score > 60 THEN 'Fair' WHEN score > 60 THEN 'Fair'
ELSE 'Poor' ELSE 'Poor'
END FROM table1;
FROM table1;

Syntax Details

If there is no ELSE part and no conditions are true, it returns NULL.

The CASE statement evaluates its conditions sequentially and stops with the first condition
whose condition is satisfied.

ROUND
Don’t forget to use round for a ratio result to increase the readability.

// Correct Syntax // Correct Syntax but not clean result

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 5

 SELECT SELECT
seller, seller,
ROUND(COUNT(shipped_flag)::FLOAT/COUNT(order_i COUNT(shipped)::FLOAT/COUNT(shipped_flag) as shi
d),2) as ship_rate p_rate
FROM sales FROM sales
GROUP BY seller; GROUP BY seller;

JOIN
(INNER) JOIN : Returns records that have matching values in both tables

LEFT (OUTER) JOIN : Returns all records from the left table, and the matched records from the right table

RIGHT (OUTER) JOIN : Returns all records from the right table, and the matched records from the left table

FULL (OUTER) JOIN : Returns all records when there is a match in either left or right table

Don’t forget the ON statement unless you are using implicit join.

// Correct Syntax // Wrong Syntax

SELECT SELECT
* *
FROM table1 AS t2 FROM table1 AS t2
JOIN table2 AS t1 JOIN table2 AS t1;
ON t2.col1 = t1.col1;

Multiple Filters
Be careful when combining OR with AND statements. Use parentheses () to combine filters when
necessary.

Example:
Select records with the status ‘ready’ or ‘shipped’ on 2022-01-01.

// Correct Logic // Wrong Logic

SELECT SELECT
* *
FROM table1 FROM table1
WHERE date = '2022-01-01' WHERE date = '2022-01-01'
AND (status = 'ready' AND status = 'ready'
OR status = 'shipped'); OR status = 'shipped';

LIKE
The percent sign (%) represents zero, one, or multiple characters

The underscore sign (_) represents one, single character

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 6

 Example:
Find any values that have "94401" in any position

// Correct Syntax // Correct Syntax but wrong result

SELECT SELECT
* *
FROM table1 AS t1 FROM table1 AS t1
WHERE address LIKE '%94401%'; WHERE address LIKE '_94401_';

<>
Find the best combo sell items, use <> to filter out the same item in an order.

// Correct Syntax // Correct Syntax but wrong result

SELECT t1.item_id, SELECT t1.item_id,

t2.item_id, t2.item_id,
COUNT(*) COUNT(*)
FROM table1 AS t1 FROM table1 AS t1
JOIN table1 AS t2 JOIN table1 AS t2
ON t1.order_id = t2.order_id ON t1.order_id = t2.order_id
WHERE t1.item_id <> t2.item_id GROUP BY 1,2
GROUP BY 1,2 ORDER BY 3 DESC
ORDER BY 3 DESC LIMIT 1;
LIMIT 1;

You can use either != or <> both in your queries as both are technically the same but <> is
preferred as that is SQL-92 standard.

GROUP BY
When mixing aggregate functions ( SUM , AVG , COUNT , etc) with an unaggregated column, ensure it’s in the
GROUP BY statement.

// Correct Syntax // Correct Syntax but wrong

// result
SELECT
COUNT(*), SELECT
city COUNT(*),
FROM table1 city
GROUP BY city; FROM table1;

When there are multiple unaggregated columns, ensure ALL are in the GROUP BY statement.

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 7

 // Correct Syntax // Wrong Syntax

SELECT SELECT
COUNT(*), COUNT(*),
city, city,
country country
FROM table1 FROM table1
GROUP BY city, country; GROUP BY city;

ORDER BY
If you use ORDER BY without any keywords, it is ascending order by default. To sort the records in
descending order, use the DESC keyword.
Example:
Return score from highest to lowest:

// Correct Syntax // Correct Syntax but wrong result, missing key

word DESC
SELECT
name, SELECT
score name,
FROM table1 score
ORDER BY score DESC; FROM table1 t1
ORDER BY score;

LIMIT N
Example:
Return top 5 scores

// Correct Syntax // Correct Syntax in MS SQL Server

SELECT SELECT TOP 5

name, name,
score score
FROM table1 FROM table1 t1
ORDER BY score DESC ORDER BY score DESC;
LIMIT 5;

UNION and UNION ALL

UNION returns unique rows and UNION ALL returns all rows.
Example:
Combine ALL data from table a and table b

Table a Table b

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 8

 id id

1 2

2 3

// Correct Logic // Wrong Logic

SELECT * SELECT *
FROM a FROM a
UNION ALL UNION
SELECT * SELECT *
FROM b; FROM b;

Complex Queries
DATE Format
DATE()

Check the date column format before using it. If it is TIMESTAMP or VARCHAR , you need to
convert it to DATE first. Otherwise, the result will not be aggregated to the date level.

Example:

Aggregate the sales by date:

// Correct Syntax // Correct Syntax but wrong result

SELECT SELECT
DATE(timestamp), timestamp,
SUM(sales) SUM(sales)
FROM table1 FROM table1
GROUP BY 1; GROUP BY 1;

CURDATE()
Example:
Find out last 7 days page visitor count

// Correct Syntax // Correct Syntax but hard coding is not pref

erred
SELECT
pagename, SELECT
COUNT(DISTINCT user_id) pagename,
FROM table1 COUNT(DISTINCT user_id)
FROM table1

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 9

 WHERE visit_timestamp > CURDATE -7 WHERE visit_timestamp BETWEEN 'YYYY-MM-DD' AND
GROUP BY 1; 'YYYY-MM-DD'
GROUP BY 1;

DATEDIFF()

If we want positive results, the first value must be less than the 2nd value.

// Correct Syntax // Correct Syntax but wrong result, confused

with start date and end date
SELECT
COUMT(seller_id) SELECT
FROM table1 COUMT(seller_id)
WHERE DATEDIFF(signup_date, order_date) >=7; FROM table1
WHERE DATEDIFF(order_date, signup_date) >=7;

Go to this link for more detailed info:

https://dev.mysql.com/doc/refman/8.0/en/date-and-time-functions.html#function_datediff

Window function
ROW_NUMBER
ROW_NUMBER always returns unique rankings.

Window function column cannot be called by where clause when it is generated

// Correct Syntax // Wrong Syntax

SELECT dep_name, SELECT dep_name,

emp_no, emp_no,
salary, salary,
enroll_date enroll_date,
FROM rank() OVER (PARTITION BY depname ORDER
(SELECT *, BY salary DESC, empno) AS pos
rank() OVER (PARTITION BY depname ORDER BY sa FROM empsalary
lary DESC, empno) AS pos WHERE pos < 3;
FROM empsalary
) AS salary_pos
WHERE pos < 3;

RANK/DENSE RANK

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 10

 RANK and DENSE_RANK return the rank of values in a column with and without gaps, respectively.

// Correct Syntax // Correct Syntax but wrong result, missing k

eyword DESC
SELECT name, score
FROM SELECT name, score
(SELECT name, score, FROM
RANK() OVER(ORDER BY grades DESC) AS ranki (SELECT name, score,
ng RANK() OVER(ORDER BY grades) AS ranking
FROM t1) scores FROM t1) scores
WHERE ranking <=10; WHERE ranking <=10;

LEAD/LAG
LAG pulls from previous rows and LEAD pulls from following rows.
Example:
Compute MoM growth Rate from a monthly sales table.

// Correct Syntax // Correct Syntax but wrong result, wrong key

word is used
SELECT seller,
years, SELECT seller,
months, years,
sales, months,
sales/LAG(sales) OVER(PARTITION BY selle sales,
r ORDER BY years, months)-1 AS growth_rate sales/LEAD(sales) OVER(PARTITION BY sell
FROM table1; er ORDER BY years, months)-1 AS growth_rate
FROM table1;

COUNT(), SUM(), MIN(), MAX(), MAX()

COUNT , SUM , MIN , MAX , and AVG can all be used in functions to compute aggregated results of each

partition and return the result to every row.

Example:
Return the cumulative balance by account. Use ORDER BY for cumulative, otherwise, it returns a total for
each row.

// Correct Syntax // Correct Syntax but wrong result, missing O

RDER BY clause
SELECT
dates, SELECT
SUM(Amount) OVER (PARTITION BY Account ORDER dates,
BY Months) SUM(Amount) OVER (PARTITION BY Account)
FROM FROM
AccountBalances; AccountBalances;

Common Table Expression (CTE)

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 11

 Using CTE improves the readability of your query

A CTE is defined once then can be referred multiple times

You declare a CTE with the instruction WITH … AS:

WITH my_cte AS
(
SELECT col1, col2 FROM table
)
SELECT * FROM my_cte;

Use “,” to separate multiple CTEs

// Correct Syntax // Correct Syntax but low readability

WITH first_purchase AS SELECT c.names,

( p.avg_salary as avg_salary,
SELECT customer_id, fp.date
MIN(date) as date FROM customers c
FROM purchases JOIN (
GROUP BY customer_id SELECT customer_id,
), MIN(date) as date
persona AS FROM purchases
( GROUP BY customer_id
SELECT age, ) AS fp
gender, ON fp.customer_id = c.id
AVG(salary) as avg_salary JOIN (
FROM customers SELECT age,
GROUP BY age, gender gender,
) AVG(salary) as avg_salary
SELECT c.name, FROM customers
p.avg_salary as avg_salary, GROUP BY age, gender
fp.date ) AS p
FROM customers c ON p.age = c.age
JOIN first_purchase fp ON fp.customer_id = c.id AND p.gender = c.gender
JOIN persona p ON p.age = p.age WHERE c.age <= 30;
AND p.gender = p.gender
WHERE c.age <= 30;

Conceptual Questions
How to write efficient SQL query
Use indexes

Indexes are used to quickly locate rows in a table based on the values in a specific column. Make
sure that you have proper indexes on the columns used in your WHERE clause and JOIN
conditions, as this can greatly improve the performance of your queries.

Avoid using SELECT *

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 12

 Instead of " SELECT * ", list only the columns you need in your result set. This will reduce the amount
of data that needs to be transferred from the database to your application, which can improve
performance.

Avoid using subqueries

Subqueries can be slow, especially when they are used in the WHERE clause. Instead, try to use
JOIN s or derived tables to achieve the same result.

Avoid using functions in the WHERE clause

Functions like UPPER() , LOWER() , and CONVERT() can prevent the use of indexes, as they change the
values of the columns. Instead, try to use these functions in the SELECT statement, after the data
has been filtered.

Use proper JOIN conditions

When using JOINs, make sure to specify the correct JOIN conditions. Using a wrong join type or
missing join conditions can result in a large number of unnecessary rows being returned, which
can negatively impact performance.

Use LIMIT and OFFSET clauses

If you only need a limited number of rows from your query, use the LIMIT and OFFSET clauses to
specify the number of rows to return. This can help reduce the amount of data returned, improving
performance.

Use the appropriate data type

Choose the appropriate data type for each column in your table. For example, using an INT data
type for a column that only contains numbers with two decimal places may not be the best choice,
as it will result in more storage space being used than necessary.

Primary key vs foreign key

PRIMARY KEY FOREIGN KEY

To ensure data in the specific column is A column or group of columns in a relational database table provides a
unique. link between data in two tables.
Uniquely identifies a record in the
It refers to the field in a table which is the primary key of another table.
relational database table.
Only one primary key is allowed in a table. More than one foreign key is allowed in a table.

[Advanced] SQL Rewrite in Pandas Equivalently

SQL Pandas

select * from df df

select * from df limit 3 df.head(3)

select id from df where

df[df.name == 'LAX'].id
name = 'LAX'

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 13

 SQL Pandas

select distinct type from df df.type.unique()

select * from df where

region = 'US-CA' and type df[(df.region == 'US-CA') & (df.type == 'sea')]
= 'sea'
select id, name, length
from df where region = 'US- df[(df.region == 'US-CA') & (df.type == 'large')][['id', 'name', 'length']]
CA' and type = 'large'
select * from df where
df[df.name == 'LAX'].sort_values('type')
name = 'LAX' order by type
select * from df where
name = 'LAX' order by type df[df.name == 'LAX'].sort_values('type', ascending=False)
desc
select * from df where type
df[df.type.isin(['heliport', 'balloonport'])]
in ('heliport', 'balloonport')

select * from df where type

not in ('heliport', df[~df.type.isin(['heliport', 'balloonport'])]
'balloonport')

select country, type,

count(*) from df group by
df.groupby(['country', 'type']).size()
country, type order by
country, type
select country, type,
count(*) from df group by df.groupby(['country','type']).size().to_frame('size').reset_index().sort_values(['country',
country, type order by 'size'], ascending=[True, False])
country, count(*) desc
select type, count(*) from df
where country = 'US' group df[df.country == 'US'].groupby('type').filter(lambda g: len(g) >
by type having count(*) > 1000).groupby('type').size().sort_values(ascending=False)
1000 order by count(*) desc
select country from df order
df.nlargest(10, 'country')
by size desc limit 10
select country from df order
by size desc limit 10 df.nlargest(20, 'country').tail(10)
offset 10
select max(length_ft),
min(length_ft),
df.agg({'length_ft': ['min', 'max', 'mean', 'median']})
avg(length_ft),
median(length_ft) from df
select name, type,
description, frequency from
df_freq.merge(df[df.name == 'LAX'][['id']], left_on='df_ref', right_on='id', how='inner')
df_freq join df on
[['name', 'type', 'description', 'frequency']]
df_freq.ref_id = df.id where
df.name = 'LAX'

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 14

 SQL Pandas

row_number()
over(partition by seller_id df.groupby([‘seller_id’])[‘order_date’].rank(method=’first’, ascending = False)
order by order_date)
rank() over(partition by
seller_id order by df.groupby(['seller_id'])['order_date'].rank(method='min')
order_date)
dense_rank() over(partition
by seller_id order by df.groupby(['seller_id'])['order_date'].rank(method='dense')
order_date)
sum(amount) over(partition
by seller_id, order_month
df.groupby(['seller_id', 'order_month'])['amount'].cumsum()
order by order_date rows
unbounded preceding)

avg(amount) over(partition
df.groupby(['seller_id', 'order_month'])['amount'].transform('mean').round(1)
by seller_id, order_month)
lag(sales, 1) over(partition
df.groupby('seller')['sales'].shift(-1)
by seller order by date)
avg(sales) over(order by
date rows between 6
df['sales'].rolling(7).mean().round(1)
preceding and current
row)

SQL Interview Cheat Sheet - emmading.com. All rights reserved. 15