Практика 'Синтаксис' - Clickhouse
This commit is contained in:
mnoskov
+120
@@ -0,0 +1,120 @@
|
||||
-- Задание 1
|
||||
|
||||
CREATE TABLE mnoskov.page_views
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
user_id UInt64,
|
||||
url String,
|
||||
view_time DateTime
|
||||
)
|
||||
ENGINE = MergeTree
|
||||
PARTITION BY toYYYYMM(view_time)
|
||||
ORDER BY (user_id, view_time);
|
||||
|
||||
|
||||
-- Задание 2
|
||||
CREATE TABLE mnoskov.users
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
id UInt64,
|
||||
name String,
|
||||
update_at DateTime
|
||||
)
|
||||
ENGINE = ReplacingMergeTree(update_at)
|
||||
ORDER BY id;
|
||||
|
||||
|
||||
-- Задание 3
|
||||
CREATE TABLE mnoskov.region_sales
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
region String,
|
||||
date Date,
|
||||
sales UInt64
|
||||
)
|
||||
ENGINE = SummingMergeTree
|
||||
PARTITION BY toYYYYMM(date)
|
||||
ORDER BY (region, date);
|
||||
|
||||
INSERT INTO mnoskov.region_sales (region, date, sales)
|
||||
VALUES
|
||||
('Moscow', '2026-05-01', 100),
|
||||
('Moscow', '2026-05-01', 150),
|
||||
('Moscow', '2026-05-01', 200),
|
||||
('Berlin', '2026-05-02', 300),
|
||||
('Berlin', '2026-05-02', 120),
|
||||
('Berlin', '2026-05-02', 80),
|
||||
('Paris', '2026-05-03', 500),
|
||||
('Paris', '2026-05-03', 250),
|
||||
('London', '2026-05-04', 400),
|
||||
('Tokyo', '2026-05-05', 700);
|
||||
|
||||
SELECT * FROM mnoskov.region_sales;
|
||||
-- Все четко работает.
|
||||
|
||||
|
||||
-- Задание 4
|
||||
CREATE TABLE mnoskov.user_metrics
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
region String,
|
||||
avg_age_state AggregateFunction(avg, UInt8),
|
||||
active_users_state AggregateFunction(uniq, UInt64)
|
||||
)
|
||||
ENGINE = AggregatingMergeTree
|
||||
ORDER BY region;
|
||||
|
||||
CREATE TABLE mnoskov.user_metrics_source
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
region String,
|
||||
age UInt8,
|
||||
user_id UInt64
|
||||
)
|
||||
ENGINE = MergeTree
|
||||
ORDER BY region;
|
||||
|
||||
INSERT INTO mnoskov.user_metrics_source
|
||||
(region, age, user_id)
|
||||
VALUES
|
||||
('Moscow', 25, 1001),
|
||||
('Moscow', 30, 1002),
|
||||
('Moscow', 22, 1003),
|
||||
('Moscow', 25, 1001),
|
||||
('Berlin', 40, 2001),
|
||||
('Berlin', 35, 2002),
|
||||
('Berlin', 29, 2003),
|
||||
('Berlin', 35, 2002),
|
||||
('Paris', 31, 3001),
|
||||
('Paris', 28, 3002),
|
||||
('Paris', 45, 3003),
|
||||
('Paris', 31, 3001),
|
||||
('London', 50, 4001),
|
||||
('London', 41, 4002),
|
||||
('London', 38, 4003),
|
||||
('London', 41, 4002),
|
||||
('Tokyo', 27, 5001),
|
||||
('Tokyo', 33, 5002),
|
||||
('Tokyo', 29, 5003),
|
||||
('Tokyo', 33, 5002);
|
||||
SELECT * FROM mnoskov.user_metrics_source;
|
||||
|
||||
INSERT INTO mnoskov.user_metrics
|
||||
SELECT
|
||||
region,
|
||||
avgState(age) AS avg_age_state,
|
||||
uniqState(user_id) AS active_users_state
|
||||
FROM
|
||||
mnoskov.user_metrics_source
|
||||
GROUP BY
|
||||
region;
|
||||
|
||||
SELECT
|
||||
region,
|
||||
avgMerge(avg_age_state),
|
||||
uniqMerge(active_users_state)
|
||||
FROM
|
||||
mnoskov.user_metrics
|
||||
GROUP BY
|
||||
region;
|
||||
-- Че то кликхаус странно себя ведет, то таблицы сами очищаются, то удаленные таблицы воскресают.
|
||||
+86
@@ -0,0 +1,86 @@
|
||||
-- Задание 1
|
||||
|
||||
CREATE TABLE mnoskov.logs
|
||||
ON CLUSTER clickhouse_cluster (
|
||||
dt DateTime,
|
||||
user_id UInt32,
|
||||
event_type String
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/logs',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY dt;
|
||||
|
||||
CREATE TABLE mnoskov.dist_logs
|
||||
ON CLUSTER clickhouse_cluster
|
||||
ENGINE = Distributed(
|
||||
'clickhouse_cluster',
|
||||
'mnoskov',
|
||||
'logs',
|
||||
user_id
|
||||
);
|
||||
-- параметр rand() - это ключ распределния между шардами. Если стоит он, то распределяются данные случайно, если укащать колонку, то по колонке.
|
||||
|
||||
|
||||
-- Задание 2
|
||||
|
||||
CREATE TABLE mnoskov.events
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
user_id UInt32,
|
||||
event_type String,
|
||||
ts DateTime
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/events',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY (ts, user_id);
|
||||
|
||||
CREATE VIEW mnoskov.top_active_users AS
|
||||
SELECT
|
||||
user_id,
|
||||
count(*) AS cnt_user
|
||||
FROM
|
||||
mnoskov.events
|
||||
GROUP BY
|
||||
user_id
|
||||
ORDER BY
|
||||
cnt_user DESC, user_id
|
||||
LIMIT 100;
|
||||
-- VIEW не подойдет, тк вьюшка будет каждый раз считаться заново, а это энергозатратно, поэтому надо использовать MATERIALIZED VIEW.
|
||||
|
||||
|
||||
-- Задание 3
|
||||
|
||||
CREATE TABLE mnoskov.raw_events
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
event_time DateTime,
|
||||
user_id UInt32,
|
||||
event_type String
|
||||
)
|
||||
ENGINE=MergeTree
|
||||
ORDER BY(event_time,user_id);
|
||||
|
||||
CREATE TABLE mnoskov.agr_clicks_per_day
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
event_date Date,
|
||||
count UInt64
|
||||
)
|
||||
ENGINE=SummingMergeTree
|
||||
ORDER BY event_date;
|
||||
|
||||
CREATE MATERIALIZED VIEW mnoskov.mv_clicks_per_day
|
||||
ON CLUSTER clickhouse_cluster
|
||||
TO mnoskov.agr_clicks_per_day AS
|
||||
SELECT
|
||||
toDate(event_time) AS event_date,
|
||||
count(*) AS count
|
||||
FROM mnoskov.raw_events
|
||||
GROUP BY event_date;
|
||||
|
||||
SELECT * FROM mnoskov.raw_events;
|
||||
SELECT * FROM mnoskov.agr_clicks_per_day;
|
||||
+115
@@ -0,0 +1,115 @@
|
||||
-- Задание 1
|
||||
|
||||
CREATE TABLE mnoskov.raw_data1
|
||||
ON CLUSTER clickhouse_cluster (
|
||||
user_id String,
|
||||
balance String,
|
||||
created_at String
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/raw_data1',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY tuple();
|
||||
|
||||
SELECT
|
||||
user_id,
|
||||
balance,
|
||||
created_at
|
||||
FROM (
|
||||
SELECT
|
||||
user_id,
|
||||
toDecimal64OrNull(balance,2) AS balance,
|
||||
toDateTime64OrNull(created_at, 3) AS created_at
|
||||
FROM
|
||||
mnoskov.raw_data1
|
||||
)
|
||||
WHERE
|
||||
balance IS NOT NULL AND created_at IS NOT NULL;
|
||||
|
||||
|
||||
-- Задание 2
|
||||
|
||||
CREATE TABLE mnoskov.events
|
||||
ON CLUSTER clickhouse_cluster (
|
||||
event_id String,
|
||||
event_date String, -- форматы: "2023-01-01", "01.02.2023", "invalid"
|
||||
event_time String -- форматы: "12:30:45", "25:99:99"
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/events',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY
|
||||
event_id;
|
||||
|
||||
SELECT * FROM mnoskov.events;
|
||||
|
||||
|
||||
SELECT
|
||||
event_id,
|
||||
event_datetime
|
||||
FROM (
|
||||
SELECT
|
||||
event_id,
|
||||
parseDateTimeBestEffortOrNull(concat(event_date, ' ', event_time)) AS event_datetime
|
||||
FROM
|
||||
mnoskov.events
|
||||
)
|
||||
ORDER BY
|
||||
event_datetime;
|
||||
|
||||
|
||||
-- Задание 3
|
||||
|
||||
CREATE TABLE mnoskov.user_activity
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
user_id UInt32,
|
||||
actions String -- JSON: ["login", "purchase", "logout"]
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/user_activity',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY
|
||||
user_id;
|
||||
|
||||
SELECT
|
||||
user_id,
|
||||
count(*) as user_actions_cnt
|
||||
FROM
|
||||
mnoskov.user_activity
|
||||
WHERE
|
||||
has(JSONExtract(actions, 'Array(String)'), 'purchase')
|
||||
GROUP BY
|
||||
user_id
|
||||
ORDER BY
|
||||
user_id;
|
||||
|
||||
|
||||
-- Задание 4
|
||||
|
||||
CREATE TABLE mnoskov.sensor_data
|
||||
ON CLUSTER clickhouse_cluster
|
||||
(
|
||||
sensor_id UInt32,
|
||||
value Float64,
|
||||
status Enum('ok' = 1, 'error' = 2)
|
||||
)
|
||||
ENGINE = ReplicatedMergeTree(
|
||||
'/clickhouse/tables/{shard}/mnoskov/sensor_data',
|
||||
'{replica}'
|
||||
)
|
||||
ORDER BY sensor_id;
|
||||
|
||||
--1-2.
|
||||
SELECT
|
||||
toString(status) AS str_status,
|
||||
avg(value) AS avg_value
|
||||
FROM
|
||||
mnoskov.sensor_data
|
||||
GROUP BY
|
||||
str_status;
|
||||
--3
|
||||
-- не понял задание. все строки в value можно преобразовать в Decimal(10,2), потому что value - Float64, а он все null переделывает в 0.
|
||||
Reference in New Issue
Block a user