GROUP_BITMAP

Introduction

The GROUP_BITMAP function is an aggregate function used to aggregate a set of unsigned integer values and compute the corresponding Bitmap data structure. This function is highly efficient when dealing with large-scale datasets, especially in scenarios where set operations (such as union, intersection, etc.) on a large number of integer values are required.

Syntax

group_bitmap(value)

Parameters

• value: An expression of type INTEGER or BIGINT, representing the integer values to be aggregated. Typically, these values are extracted from a column in a database table.

Return Value

The function returns a result of type BITMAP, which represents the aggregated bitmap. A bitmap is a compact data structure that stores the existence of integer values using bit representation. Each bit corresponds to an integer value. If a bit is set to 1, it indicates that the corresponding integer value exists in the set; if the bit is 0, it indicates that the value does not exist.

Usage Example

Example 1: Basic Usage

SELECT group_bitmap(v) AS bitmap_result FROM VALUES (1), (2), (3), (4), (5) AS t(v);

Result:

This returns a complete bitmap representing the set of all users.

Example 2: Counting distinct values per group

SELECT c, group_bitmap(v) AS bitmap_count FROM VALUES ('a', 1), ('a', 2), ('a', 3), ('b', 1), ('b', 2) AS t(c, v) GROUP BY c; +---+--------------+ | c | bitmap_count | +---+--------------+ | a | 3 | | b | 2 | +---+--------------+

Notes

1. Input Value Range: The group_bitmap function is designed for unsigned integer values. If negative values are provided as input, it may lead to unexpected results or errors.
2. Performance Optimization: The group_bitmap function is highly efficient for large-scale datasets. However, if the data volume is extremely large, performance impact should still be considered. Where possible, try to optimize the input data to improve the function's execution efficiency.
3. Intermediate State Usage: If you need to perform aggregate operations on multiple groups, it is recommended to use the group_bitmap_state and group_bitmap_merge functions to avoid redundant calculations.