пет-проект полностью

This commit is contained in:
Bonifatsiy
2026-05-11 04:21:02 +04:00
parent c654840ffd
commit d280456d09
4 changed files with 377 additions and 0 deletions
@@ -0,0 +1,90 @@
DROP TABLE IF EXISTS dsolovianov.core_fact_lesson_views ON CLUSTER clickhouse_cluster SYNC;
DROP TABLE IF EXISTS dsolovianov.core_dim_user ON CLUSTER clickhouse_cluster SYNC;
DROP TABLE IF EXISTS dsolovianov.core_dim_course ON CLUSTER clickhouse_cluster SYNC;
DROP TABLE IF EXISTS dsolovianov.core_dim_lesson ON CLUSTER clickhouse_cluster SYNC;
DROP TABLE IF EXISTS dsolovianov.core_dim_date ON CLUSTER clickhouse_cluster SYNC;
DROP TABLE IF EXISTS dsolovianov.core_enrollments ON CLUSTER clickhouse_cluster SYNC;
CREATE TABLE dsolovianov.core_dim_user
ON CLUSTER clickhouse_cluster
(
user_id Int32,
name String,
age Int32,
email String,
registration_date Date
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_dim_user',
'{replica}'
)
ORDER BY user_id;
CREATE TABLE dsolovianov.core_dim_course
ON CLUSTER clickhouse_cluster
(
course_id Int32,
title String,
category String
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_dim_course',
'{replica}'
)
ORDER BY course_id;
CREATE TABLE dsolovianov.core_dim_lesson
ON CLUSTER clickhouse_cluster
(
lesson_id Int32,
title String,
duration_min Int32,
course_id Int32
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_dim_lesson',
'{replica}'
)
ORDER BY lesson_id;
CREATE TABLE dsolovianov.core_dim_date
ON CLUSTER clickhouse_cluster
(
date Date,
year Int32,
month Int32,
day Int32,
weekday String
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_dim_date',
'{replica}'
)
ORDER BY date;
CREATE TABLE dsolovianov.core_enrollments
ON CLUSTER clickhouse_cluster
(
user_id Int32,
course_id Int32,
enrolled_at DateTime64(3)
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_enrollments',
'{replica}'
)
ORDER BY user_id;
CREATE TABLE dsolovianov.core_fact_lesson_views
ON CLUSTER clickhouse_cluster
(
user_id Int32,
lesson_id Int32,
course_id Int32,
viewed_at DateTime64(3)
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/core_fact_lesson_views',
'{replica}'
)
ORDER BY user_id;
@@ -0,0 +1,186 @@
-- Этап 2. Создание витринных таблиц --
-- Витрина lesson_popularity_summary
DROP TABLE IF EXISTS dsolovianov.lesson_popularity_summary ON CLUSTER clickhouse_cluster SYNC;
CREATE TABLE dsolovianov.lesson_popularity_summary ON CLUSTER clickhouse_cluster (
lesson_id UInt32,
lesson_title String,
course_id UInt32,
course_title String,
total_views UInt64,
unique_users UInt64,
first_view DateTime,
last_view DateTime
) ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/lesson_popularity_summary',
'{replica}'
)
ORDER BY (course_id, lesson_id);
CREATE MATERIALIZED VIEW dsolovianov.lesson_popularity_summary_mv ON CLUSTER clickhouse_cluster
TO dsolovianov.lesson_popularity_summary
AS
SELECT
l.lesson_id AS lesson_id,
l.title AS lesson_title,
l.course_id AS course_id,
c.title AS course_title,
COUNT(v.viewed_at) AS total_views,
COUNT(DISTINCT v.user_id) AS unique_users,
MIN(v.viewed_at) AS first_view,
MAX(v.viewed_at) AS last_view
FROM dsolovianov.core_fact_lesson_views v
LEFT JOIN dsolovianov.core_dim_lesson l ON v.lesson_id = l.lesson_id
LEFT JOIN dsolovianov.core_dim_course c ON l.course_id = c.course_id
GROUP BY l.lesson_id, l.title, l.course_id, c.title;
INSERT INTO dsolovianov.lesson_popularity_summary
SELECT
l.lesson_id AS lesson_id,
l.title AS lesson_title,
l.course_id AS course_id,
c.title AS course_title,
COUNT(v.viewed_at) AS total_views,
COUNT(DISTINCT v.user_id) AS unique_users,
MIN(v.viewed_at) AS first_view,
MAX(v.viewed_at) AS last_view
FROM dsolovianov.core_dim_lesson l
JOIN dsolovianov.core_dim_course c ON l.course_id = c.course_id
LEFT JOIN dsolovianov.core_fact_lesson_views v ON l.lesson_id = v.lesson_id
GROUP BY l.lesson_id, l.title, l.course_id, c.title;
-- Витрину inactive_users_summary
DROP TABLE IF EXISTS dsolovianov.inactive_users_summary ON CLUSTER clickhouse_cluster SYNC;
CREATE TABLE dsolovianov.inactive_users_summary ON CLUSTER clickhouse_cluster (
user_id UInt32,
name String,
email String,
age UInt8,
registered_courses_count UInt32,
registration_date Date
) ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/inactive_users_summary',
'{replica}'
)
ORDER BY user_id;
CREATE MATERIALIZED VIEW dsolovianov.inactive_users_summary_mv ON CLUSTER clickhouse_cluster
TO dsolovianov.inactive_users_summary
AS
SELECT
u.user_id AS user_id,
u.name AS name,
u.email AS email,
u.age AS age,
COUNT(DISTINCT e.course_id) AS registered_courses_count,
u.registration_date AS registration_date
FROM dsolovianov.core_dim_user u
LEFT JOIN dsolovianov.core_enrollments e ON u.user_id = e.user_id
WHERE u.user_id NOT IN (SELECT DISTINCT user_id FROM dsolovianov.core_fact_lesson_views)
GROUP BY u.user_id, u.name, u.email, u.age, u.registration_date;
INSERT INTO dsolovianov.inactive_users_summary
SELECT
u.user_id AS user_id,
u.name AS name,
u.email AS email,
u.age AS age,
COUNT(DISTINCT e.course_id) AS registered_courses_count,
u.registration_date AS registration_date
FROM dsolovianov.core_dim_user u
LEFT JOIN dsolovianov.core_enrollments e ON u.user_id = e.user_id
WHERE u.user_id NOT IN (SELECT DISTINCT user_id FROM dsolovianov.core_fact_lesson_views)
GROUP BY u.user_id, u.name, u.email, u.age, u.registration_date;
-- Витрина course_completion_rate
DROP TABLE IF EXISTS dsolovianov.course_completion_rate ON CLUSTER clickhouse_cluster SYNC;
CREATE TABLE dsolovianov.course_completion_rate ON CLUSTER clickhouse_cluster (
user_id UInt32,
user_name String,
course_id UInt32,
course_title String,
lessons_in_course UInt32,
lessons_viewed UInt32,
completion_rate Float32
) ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/dsolovianov/course_completion_rate',
'{replica}'
)
ORDER BY (user_id, course_id);
DROP TABLE IF EXISTS dsolovianov.course_completion_rate_mv ON CLUSTER clickhouse_cluster SYNC;
CREATE MATERIALIZED VIEW dsolovianov.course_completion_rate_mv ON CLUSTER clickhouse_cluster
TO dsolovianov.course_completion_rate
AS
SELECT
e.user_id AS user_id,
u.name AS user_name,
e.course_id AS course_id,
c.title AS course_title,
(
SELECT COUNT(*)
FROM dsolovianov.core_dim_lesson l
WHERE l.course_id = e.course_id
) AS lessons_in_course,
(
SELECT COUNT(DISTINCT v.lesson_id)
FROM dsolovianov.core_fact_lesson_views v
JOIN dsolovianov.core_dim_lesson l ON v.lesson_id = l.lesson_id
WHERE v.user_id = e.user_id AND l.course_id = e.course_id
) AS lessons_viewed,
(
SELECT COUNT(DISTINCT v.lesson_id)
FROM dsolovianov.core_fact_lesson_views v
JOIN dsolovianov.core_dim_lesson l ON v.lesson_id = l.lesson_id
WHERE v.user_id = e.user_id AND l.course_id = e.course_id
) / (
SELECT COUNT(*)
FROM dsolovianov.core_dim_lesson l
WHERE l.course_id = e.course_id
) AS completion_rate
FROM dsolovianov.core_enrollments e
JOIN dsolovianov.core_dim_user u ON e.user_id = u.user_id
JOIN dsolovianov.core_dim_course c ON e.course_id = c.course_id
GROUP BY e.user_id, u.name, e.course_id, c.title;
INSERT INTO dsolovianov.course_completion_rate
SELECT
e.user_id AS user_id,
u.name AS user_name,
e.course_id AS course_id,
c.title AS course_title,
(
SELECT COUNT(*)
FROM dsolovianov.core_dim_lesson l
WHERE l.course_id = e.course_id
) AS lessons_in_course,
COALESCE(
(
SELECT COUNT(DISTINCT v.lesson_id)
FROM dsolovianov.core_fact_lesson_views v
JOIN dsolovianov.core_dim_lesson l ON v.lesson_id = l.lesson_id
WHERE v.user_id = e.user_id AND l.course_id = e.course_id
), 0
) AS lessons_viewed,
COALESCE(
(
SELECT COUNT(DISTINCT v.lesson_id)
FROM dsolovianov.core_fact_lesson_views v
JOIN dsolovianov.core_dim_lesson l ON v.lesson_id = l.lesson_id
WHERE v.user_id = e.user_id AND l.course_id = e.course_id
), 0
) / (
SELECT COUNT(*)
FROM dsolovianov.core_dim_lesson l
WHERE l.course_id = e.course_id
) AS completion_rate
FROM dsolovianov.core_enrollments e
JOIN dsolovianov.core_dim_user u ON e.user_id = u.user_id
JOIN dsolovianov.core_dim_course c ON e.course_id = c.course_id
GROUP BY e.user_id, u.name, e.course_id, c.title;
@@ -0,0 +1,88 @@
-- Этап 3. Анализ и запросы - Напиши SQL-запросы к витринам -- Напиши свои вариации для практики--
-- Топ-5 самых просматриваемых уроков
SELECT
lesson_title,
course_title,
total_views
FROM lesson_popularity_summary
ORDER BY total_views DESC
LIMIT 5;
-- Неактивные пользователи, записавшиеся на курсы
SELECT
name,
email,
registered_courses_count
FROM inactive_users_summary
WHERE registered_courses_count > 0
ORDER BY registration_date DESC;
-- Курсы с самым высоким процентом завершения
SELECT
course_title,
AVG(completion_rate) AS avg_completion
FROM course_completion_rate
GROUP BY course_title
ORDER BY avg_completion DESC
LIMIT 5;
-- Мои вариации --
-- Среднее количество учеников на урок по каждому курсу
SELECT
course_title,
ROUND(AVG(unique_users), 2) AS avg_users_lesson,
SUM(total_views) AS total_course_views
FROM dsolovianov.lesson_popularity_summary
GROUP BY course_title
ORDER BY avg_users_lesson DESC;
-- Возрастное распределение неактивных пользователей
SELECT
CASE
WHEN age < 25 THEN 'до 25'
WHEN age BETWEEN 25 AND 30 THEN '25-30'
WHEN age BETWEEN 31 AND 35 THEN '31-35'
ELSE '35+'
END AS age_group,
COUNT(*) AS inactive_users_count,
AVG(registered_courses_count) AS avg_courses_enrolled
FROM dsolovianov.inactive_users_summary
GROUP BY age_group
ORDER BY age_group;
-- Пользователи, которые начали, но не закончили курс (completion_rate между 0 и 1)
SELECT
user_name,
course_title,
lessons_viewed,
lessons_in_course,
ROUND(completion_rate * 100, 0) AS no_completion,
(lessons_in_course - lessons_viewed) AS remaining_lessons
FROM dsolovianov.course_completion_rate
WHERE completion_rate BETWEEN 0.01 AND 0.99
ORDER BY completion_rate DESC, user_name;
-- Прогресс каждого пользователя по курсам (с указанием статуса)
SELECT
user_name,
course_title,
lessons_viewed,
lessons_in_course,
ROUND(completion_rate * 100, 0) AS progress_percent,
CASE
WHEN completion_rate = 1.0 THEN 'Пройден'
WHEN completion_rate = 0 THEN 'Не начат'
WHEN completion_rate >= 0.5 THEN 'В процессе (>50%)'
ELSE 'В процессе (<50%)'
END AS status
FROM dsolovianov.course_completion_rate
ORDER BY user_name, completion_rate DESC;
@@ -0,0 +1,13 @@
-- Этап 4. Тестирование и валидация --
/*
### **Задание 6: Сравни данные с Greenplum**
1. Выбери 2–3 значения из таблиц `_summary_source` в Greenplum
2. Сравни с ClickHouse-результатами:
- Совпадает ли `SUM`, `AVG`, `MIN`, `MAX`
- Проверяй `NULL`, `0`, округления
3. Напиши мини-отчёт: **"где совпадает, где нет, почему"**
*/
-- Ничего не понятно по этому этапу. У меня в GP нет таблиц "_summary_source"