Merge pull request 'completed stage 1 creation tables and relations' (#40) from OS-310 into main

Reviewed-on: https://gitea.de-infra.servehttp.com/de-infrastructure/DWH/pulls/40
Reviewed-by: aleksey bogatyrev <abogatyrev@example.com>
This commit is contained in:
2026-01-07 21:15:41 +04:00
4 changed files with 178 additions and 0 deletions
+70
View File
@@ -0,0 +1,70 @@
--drop table is exists users;
create table users(
id int primary key,
name varchar(50) not null,
age int not null,
email varchar(50) not null,
registration_date date not null default current_date
)
insert into users (id, name, age, email, registration_date) values
(1, 'Alice', 25, 'alice@mail.com', '2023-01-10'),
(2, 'Bob', 30, 'bob@gmail.com', '2023-02-05'),
(3, 'Charlie',22, 'charlie@mail.com', '2023-02-20'),
(4, 'Diana', 28, 'diana@mail.com', '2023-03-01'),
(5, 'Ethan', 35, 'ethan@gmail.com', '2023-03-10')
create table courses(
id int primary key,
title text not null,
category varchar(30) not null,
created_at date not null default current_date
)
insert into courses (id, title, category, created_at) values
(1, 'SQL для начинающих', 'data', '2023-01-01'),
(2, 'Python для анализа данных', 'programming', '2023-01-15'),
(3, 'BI с нуля', 'business', '2023-03-01');
create table lessons(
id int primary key,
course_id int references courses(id),
title text not null,
duration_min int
)
insert into lessons (id, course_id, title, duration_min) values
(1, 1, 'SELECT и FROM', 10),
(2, 1, 'JOIN', 15),
(3, 2, 'Pandas', 20),
(4, 3, 'Основы BI', 12),
(5, 3, 'Метрики и дашборды', 18)
create table enrollments(
id int primary key,
user_id int references users(id),
course_id int references courses(id),
enrolled_at date not null default current_date
)
insert into enrollments (id, user_id, course_id, enrolled_at) values
(1, 1, 1, '2023-01-15'),
(2, 1, 2, '2023-02-01'),
(3, 2, 1, '2023-01-20'),
(4, 3, 2, '2023-03-05'),
(5, 4, 3, '2023-04-01')
create table lessons_views(
id int primary key,
user_id int references users(id),
lesson_id int references lessons(id),
viewed_at timestamp not null default now()
)
insert into lessons_views(id, user_id, lesson_id, viewed_at) values
(1, 1, 1, '2023-01-16 10:00:00'),
(2, 1, 2, '2023-01-16 10:15:00'),
(3, 2, 1, '2023-01-21 09:00:00'),
(4, 3, 3, '2023-03-06 12:00:00'),
(5, 4, 4, '2023-04-02 10:00:00'),
(6, 4, 5, '2023-04-03 10:00:00')
+16
View File
@@ -0,0 +1,16 @@
create materialized view dm_course_summary as
select
e.course_id,
c.title,
count(distinct e.user_id) as unique_users,
count(distinct l.id) as lessons_count,
count(distinct lv.id) as total_views,
round(count(distinct lv.id)*1.0/nullif(count(distinct e.user_id), 0), 2) as avg_views_per_user,
min(lv.viewed_at) as first_viewed_at,
max(lv.viewed_at) as last_viewed_at
from enrollments e
join courses c on e.course_id = c.id
join lessons l on c.id = l.course_id
join lessons_views lv on lv.lesson_id = l.id
group by e.course_id, c.title
order by e.course_id
@@ -0,0 +1,8 @@
create materialized view dm_platform_summary as
select
(select count(id) from users) as total_users,
(select count(id) from courses) as total_courses,
(select count(id) from lessons) as total_lessons,
(select count(distinct user_id) from lessons_views) as users_with_views,
round((select count(id) from lessons)*1.0/nullif((select count(id) from courses), 0), 2) as avg_lessons_per_course,
round((select count(id) from lessons_views)*1.0/nullif((select count(id) from lessons), 0), 2) as avg_views_per_lesson
@@ -0,0 +1,84 @@
CREATE MATERIALIZED VIEW dm_user_activity_summary as
with enrollment_counts as(
select user_id,
count(course_id) as enrollment_count
from enrollments e
group by user_id
),views as(
select user_id,
count(distinct lesson_id) as viewed_lessons_count,
min(viewed_at) as first_view,
max(viewed_at) as last_view
from lessons_views
group by user_id
), total_lessons as(
select user_id,
count(*) as total_lessons
from enrollments e
join lessons using(course_id)
group by user_id
order by user_id
), engagment_rate as (
select tl.user_id, round(v.viewed_lessons_count*100.0/nullif(total_lessons), 2) as engagment_rate
from total_lessons tl
join views v using(user_id)
)
select u.id, u.name, enrollment_count, viewed_lessons_count, last_view, first_view, engagment_rate
from users u
join enrollment_counts ec on u.id = ec.user_id
join views v on u.id = v.user_id
join engagment_rate er on u.id = er.user_id
order by u.id
/*
--первая попытка сделать витрину, делал избыточно с проверкой каждого подзапроса, поэтому забыл убрать order by, а также больше практиковался в оконных функциях
CREATE MATERIALIZED VIEW user_activity_summary as
with enrollment_counts as(
select distinct user_id,
count(course_id) over (partition by user_id) as enrollment_count
from enrollments e
order by user_id
), viewed_lessons_count as (
select
user_id, count(distinct lesson_id) as viewed_lessons_count
from lessons_views lv
group by user_id
order by user_id
), first_views as(
select
user_id,
viewed_at,
row_number() over (partition by user_id order by viewed_at asc) as rnf
from lessons_views
), last_views as(
select
user_id,
viewed_at,
row_number() over (partition by user_id order by viewed_at desc) as rnl
from lessons_views
), result_views as(
select user_id,
fw.viewed_at as first_view,
lw.viewed_at as last_view
from first_views fw
join last_views lw using(user_id)
where rnf = 1 and rnl = 1
), total_lessons as(
select user_id,
count(*) as total_lessons
from enrollments e
join lessons using(course_id)
group by user_id
order by user_id
), engagment_rate as (
select tl.user_id, round(vlc.viewed_lessons_count*100.0/total_lessons, 2) as engagment_rate
from total_lessons tl
join viewed_lessons_count vlc using(user_id)
)
select u.id, u.name, enrollment_count, viewed_lessons_count, last_view, first_view, engagment_rate
from users u
join enrollment_counts ec on u.id = ec.user_id
join viewed_lessons_count vlc on u.id = vlc.user_id
join result_views rv on u.id = rv.user_id
join engagment_rate er on u.id = er.user_id
order by u.id
*/