# Mariadb Aggregate Functions > Reference for MariaDB aggregate functions, covering counting, totals, extrema, string and bitwise aggregation, and statistical functions, with guidance on NULL handling and common pitfalls. - Skill: `mariadb-corporation/mariadb-aggregate-functions` (Agent Skill) - Install (CLI): `npx skillmds add mariadb-corporation/mariadb-aggregate-functions` - Raw SKILL.md: https://api.skillmd.com/api/skills/mariadb-corporation/mariadb-aggregate-functions/raw - Safety review: PASS (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Data & Analytics, SQL & Databases - Tags: Aggregate Functions, Count, Group By, Group Concat, Mariadb, Sql, Sum, Window Functions - Author: mariadb-corporation (https://skillmd.com/u/mariadb-corporation) - Updated: 2026-08-22 - Page: https://skillmd.com/skills/mariadb-corporation/mariadb-aggregate-functions --- # MariaDB Aggregate Functions *Last updated: 2026-06-24* Catalog of every built-in aggregate function in MariaDB, with signature and a one-line semantic summary per entry. For a function not listed here, fall back to the canonical reference at . For `GROUP BY`, `HAVING`, `WITH ROLLUP`, and window-function `OVER()` mechanics, see `mariadb-select`. > **Default context:** Assume MariaDB **11.8 LTS** (GA May 2025) unless the user states another version. Functions with a `*(since X.Y)*` annotation are only available from that version onward; everything else is in every current LTS branch (10.6, 10.11, 11.4, 11.8). ## What LLMs Often Miss | If the agent writes / assumes… | …prefer the MariaDB form | |---|---| | `COUNT(col)` counts every row, the same as `COUNT(*)` | `COUNT(*)` counts **all rows** (including `NULL`s); `COUNT(expr)` counts only **non-NULL** `expr`; `COUNT(DISTINCT expr)` counts distinct non-NULL values. All three can differ on the same column | | `SUM`/`AVG`/`MIN`/`MAX` factor `NULL` rows into the result | Every aggregate except `COUNT(*)` **skips `NULL`** rows. `AVG(col)` is sum-of-non-null ÷ count-of-non-null — `NULL`s change the denominator, not just the numerator | | `SUM`/`AVG` over an empty set returns `0` | On no matching rows, `SUM`, `AVG`, `MIN`, `MAX`, `GROUP_CONCAT`, and the `STDDEV`/`VAR` family return **`NULL`**; only `COUNT` returns `0` | | `GROUP_CONCAT(...)` returns the full concatenation | Output is silently capped at **`group_concat_max_len` (default 1 MB)** and truncated. A truncation **warning** is raised (not an error) — check `SHOW WARNINGS`, or raise the variable | | `GROUP_CONCAT` is plain concatenation with no ordering or dedup | It takes `[DISTINCT]`, `ORDER BY`, `SEPARATOR str` (default comma; `SEPARATOR ''` for none), and `LIMIT`, all inside the call: `GROUP_CONCAT(DISTINCT x ORDER BY y SEPARATOR '; ' LIMIT 10)` | | Selecting non-grouped columns will be rejected (ONLY_FULL_GROUP_BY) | MariaDB's **default `sql_mode` does not include `ONLY_FULL_GROUP_BY`**, so the query runs and returns an **arbitrary** row's value per group — a silent-bug source. Add the column to `GROUP BY`, or enable `ONLY_FULL_GROUP_BY` to make it an error | | `STDDEV(x)` / `STD(x)` is the sample standard deviation | `STD`, `STDDEV`, and `STDDEV_POP` all return the **population** standard deviation (÷N). For the sample value (÷N−1) use `STDDEV_SAMP`. Likewise `VARIANCE`/`VAR_POP` are population; `VAR_SAMP` is sample | | Any aggregate works as a window function with `OVER()` | Most do (`COUNT`, `SUM`, `AVG`, `MIN`, `MAX`, `BIT_*`, the stats family). But `SUM`/`AVG` **with `DISTINCT`** cannot be window functions, and `JSON_ARRAYAGG`/`JSON_OBJECTAGG` cannot either | | `MIN(enum_col)` / `MAX(enum_col)` orders by the declared `ENUM` order | `MIN`/`MAX` compare `ENUM`/`SET` by **string value**, not declared position — the result can differ from `ORDER BY enum_col LIMIT 1` | | `BIT_AND`/`BIT_OR`/`BIT_XOR` work at the column's native width | They always compute at **64-bit** precision and ignore `NULL`s. On an all-`NULL`/empty set they return the identity (`BIT_AND` → all bits set, `BIT_OR`/`BIT_XOR` → 0) | | Aggregating JSON needs a custom function | Use `JSON_ARRAYAGG` / `JSON_OBJECTAGG` — but they're catalogued under **`mariadb-json-functions`** (they live in the JSON tree), and their output is also bounded by `group_concat_max_len` | For `COUNT(DISTINCT expr [, expr...])` specifically, note it counts distinct **non-NULL** combinations across the listed expressions. ## Functions Auto-generated from the canonical `server/reference/sql-functions/aggregate-functions/` pages by `agent-skills/extractor/extract_function_category.py`; regenerate when the doc tree changes. Do not hand-edit between the markers. (The JSON aggregates `JSON_ARRAYAGG`/`JSON_OBJECTAGG` live in the JSON tree — see `mariadb-json-functions`.) ### AVG `AVG([DISTINCT] expr)` Returns the average value of expr. ### BIT_AND `BIT_AND(expr) [over_clause]` Returns the bitwise AND of all bits in _expr_. ### BIT_OR `BIT_OR(expr) [over_clause]` Returns the bitwise OR of all bits in `expr`. ### BIT_XOR `BIT_XOR(expr) [over_clause]` Returns the bitwise XOR of all bits in `expr`. ### COUNT `COUNT(expr)` Returns a count of the number of non-NULL values of expr in the rows retrieved by a SELECT statement. ### GROUP_CONCAT `GROUP_CONCAT(expr)` This function returns a string result with the concatenated non-NULL values from a group. ### MAX `MAX([DISTINCT] expr)` Returns the largest, or maximum, value of _`expr`_. ### MIN `MIN([DISTINCT] expr)` Returns the minimum value of _`expr`_. ### STD `STD(expr)` Returns the population standard deviation of _`expr`_. ### STDDEV `STDDEV(expr)` Returns the population standard deviation of _`expr`_. ### STDDEV_POP `STDDEV_POP(expr)` Returns the population standard deviation of _`expr`_ (the square root of VAR_POP()). ### STDDEV_SAMP `STDDEV_SAMP(expr)` Returns the sample standard deviation of `expr` (the square root of VAR_SAMP()). ### SUM `SUM([DISTINCT] expr)` Returns the sum of _`expr`_. ### VARIANCE `VARIANCE(expr)` Returns the population standard variance of `expr`. ### VAR_POP `VAR_POP(expr)` Returns the population standard variance of `expr`. ### VAR_SAMP `VAR_SAMP(expr)` Returns the sample variance of _`expr`_. ## See Also - **`mariadb-select`** — `GROUP BY`, `HAVING`, `WITH ROLLUP`, and window-function `OVER()` mechanics; the `ONLY_FULL_GROUP_BY` interaction - **`mariadb-json-functions`** — `JSON_ARRAYAGG` / `JSON_OBJECTAGG` - **`mariadb-string-functions`** — non-aggregate string building (`CONCAT`, `CONCAT_WS`) - Canonical reference on `mariadb.com/docs`: _This page is: Copyright © 2026 MariaDB. All rights reserved._