Skip to content

Conversation

@DamianMaslanka5
Copy link

Some time ago I looked for similar information and I was not able find any data which clearly shows how popular uv really is.
Stars on github don't always translate into real world usage.

I think this chart might be useful for other people deciding if they should replace pip with uv

Current sql query is slow (10s), so new table only for this query is needed.

SELECT toStartOfWeek(date) AS x, installer AS name, sum(count) AS y
FROM pypi.pypi_downloads_per_day_by_version_by_installer_by_type
WHERE date > toStartOfMonth(now() - toIntervalMonth(18)) and x != toStartOfWeek(now()) AND installer IN ('pip', 'uv', 'poetry')
GROUP BY x, name
ORDER BY x ASC

New table

CREATE OR REPLACE TABLE pypi.pypi_downloads_per_day_by_installer
(
    `date` Date,
    `installer` String,
    `count` Int64
)
ENGINE = SummingMergeTree
ORDER BY (date, installer)

image

@vercel
Copy link

vercel bot commented Nov 26, 2025

@DamianMaslanka5 is attempting to deploy a commit to the ClickHouse Team on Vercel.

A member of the Team first needs to authorize it.

@gingerwizard
Copy link
Collaborator

@DamianMaslanka5 i think we don't need the new table - the view forces filtering by project.

SELECT
    toStartOfWeek(date) AS x,
    installer AS name,
    sum(count) AS y
FROM pypi.pypi_downloads_per_day_by_version_by_installer_by_type
WHERE (project = 'boto3') AND (date > toStartOfMonth(now() - toIntervalMonth(18))) AND (x != toStartOfWeek(now())) AND (installer IN ('pip', 'uv', 'poetry'))
GROUP BY
    x,
    name
ORDER BY x ASC
FORMAT `Null`

Query id: 223afc04-b0a1-437a-8281-f14bb28fb42f

Ok.

0 rows in set. Elapsed: 0.078 sec. Processed 15.85 million rows, 516.45 MB (202.70 million rows/s., 6.60 GB/s.)
Peak memory usage: 503.13 MiB.

@gingerwizard
Copy link
Collaborator

Ah i see you want this on the summary page, we can add this to both maybe this page and the package page @lio-p

@gingerwizard gingerwizard self-requested a review December 15, 2025 11:20
@gingerwizard
Copy link
Collaborator

@lio-p visually id prefer we think about how to visualize this - i dont like the colors on the main page. Maybe we can do as a area chart and split between this and top contributors (something we've wanted to do for a while)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants