WITH users (user_id, action_date, action) 
AS (VALUES 
(1, CAST('01-01-20' AS date), 'Join'), 
(1, CAST('01-02-20' AS date), 'Access'), 
(2, CAST('01-02-20' AS date), 'Join'), 
(3, CAST('01-02-20' AS date), 'Join'), 
(1, CAST('01-03-20' AS date), 'Access'), 
(3, CAST('01-03-20' AS date), 'Access'),
(1, CAST('01-04-20' AS date), 'Access')),
-- get join dates for each user
join_dates AS (
SELECT user_id, action_date AS join_date
FROM users
WHERE action = 'Join' ),
-- create vector containing all dates in date range
date_vector AS (
SELECT cast(generate_series(min(action_date), max(action_date), 
            '1 day'::interval) AS date) AS dates
FROM users ),
-- cross join to get all possible user-date combinations
all_users_dates AS (
SELECT DISTINCT user_id, d.dates
FROM users
CROSS JOIN date_vector d ),
-- left join users table onto all user-date combinations on matching user ID and date (null on days where user didn't engage), join onto this each user's signup date, exclude user-date combinations falling before user signup
t1 AS (
SELECT a.dates - c.join_date AS day_no, b.user_id
FROM all_users_dates a
LEFT JOIN users b
ON a.user_id = b.user_id
AND a.dates = b.action_date
JOIN join_dates c
ON a.user_id = c.user_id 
WHERE a.dates - c.join_date >= 0 )
-- grouping by days since signup, count (non-null) user IDs as active users, total users, and the quotient as retention rate
SELECT day_no, count(*) AS n_total, 
       count(DISTINCT user_id) AS n_active, 
       round(1.0*count(DISTINCT user_id)/count(*), 2) AS retention
FROM t1
GROUP BY 1

PostgreSQL online editor

Write, Run & Share PostgreSQL queries online using OneCompiler's PostgreSQL online editor and compiler for free. It's one of the robust, feature-rich online editor and compiler for PostgreSQL. Getting started with the OneCompiler's PostgreSQL editor is really simple and pretty fast. The editor shows sample boilerplate code when you choose database as 'PostgreSQL' and start writing queries to learn and test online without worrying about tedious process of installation.

About PostgreSQL

PostgreSQL is a open source relational database system and is also knows as Postgres.

Key Features:

  • Postgres is not only free and open-source but also it is highly extensible.
  • Custom Data types and funtions from various programming languaues can be introduced and the good part is compiling entire database is not required.
  • ACID(Atomicity, Consistency, Isolation, Durability) compliant.
  • First DBMS which implemented Multi-version concurrency control (MVCC) feature.
  • It's the default database server for MacOS.
  • It supports all major operating systems like Linux, Windows, OpenBSD,FreeBSD etc.

Syntax help

1. CREATE

CREATE command is used to create a table, schema or an index.

Syntax:

         CREATE TABLE table_name (
                column1 datatype,
                column2 datatype,
                ....);

2. ALTER

ALTER command is used to add, modify or delete columns or constraints from the database table.

Syntax

ALTER TABLE Table_name ADD column_name datatype;

3. TRUNCATE:

TRUNCATE command is used to delete the data present in the table but this will not delete the table.

Syntax

TRUNCATE table table_name;

4. DROP

DROP command is used to delete the table along with its data.

Syntax

DROP TABLE table_name;

5. RENAME

RENAME command is used to rename the table name.

Syntax

ALTER TABLE table_name1 RENAME to new_table_name1;

6. INSERT

INSERT Statement is used to insert new records into the database table.

Syntax

INSERT INTO table_name (column1, column2, column3, ...) VALUES (value1, value2, value3, ...);

7. SELECT

Select statement is used to select data from database tables.

Syntax:

SELECT column1, column2, ...
FROM table_name;

8. UPDATE

UPDATE statement is used to modify the existing values of records present in the database table.

Syntax

UPDATE table_name
SET column1 = value1, column2 = value2, ...
WHERE condition;

9. DELETE

DELETE statement is used to delete the existing records present in the database table.

Syntax

DELETE FROM table_name where condition;