Use clickpy data

[astrojuanlu]

I tried to use Clickpy data https://clickhouse.com/blog/clickpy-one-trillion-rows like this:

import clickhouse_connect

client = clickhouse_connect.get_client(host="clickpy-clickhouse.clickhouse.com", username="play", database="pypi", secure=True)

kedro_all_total_downloads = pl.from_arrow(client.query_arrow("SELECT project, count FROM pypi_downloads WHERE project LIKE 'kedro%'"))

However:

• The aggregations are quite coarse, which improves performance but requires thinking what data do we want
• There are some minor annoyances in the data https://github.com/ClickHouse/clickpy/issues/89
• Some things didn't quite work as expected out of the box https://github.com/ClickHouse/clickhouse-connect/issues/394

[astrojuanlu]

All of these issues are minor. Let's do this.