Add quantileBFloat16Weighted aggregate function to ClickHouse

Why it matters

This feature provides users with a memory-efficient and fast method for calculating approximate weighted quantiles by utilizing the BFloat16 format. It solves the problem of high memory consumption in quantile computations while allowing weighted inputs, which are important for many analytical use cases.

How to use it

Use the new aggregate function quantileBFloat16Weighted in your queries similarly to other weighted quantile functions, for example:

SELECT quantileBFloat16Weighted(value, weight) FROM table;