Skip to content

The easy-to-use open source Business Intelligence and Embedded Analytics tool that lets everyone work with data :bar_chart:

Document bigquery median behaviour (#82201)

Riley Thompsona155c1f4eea1
diff --git a/docs/questions/query-builder/expressions-list.md b/docs/questions/query-builder/expressions-list.md
index 3221338fcb1..ef918042370 100644
--- a/docs/questions/query-builder/expressions-list.md
+++ b/docs/questions/query-builder/expressions-list.md
@@ -186,7 +186,9 @@ Syntax: `Median(column)`
Example: `Median([Age])` would find the midpoint age where half of the ages are older, and half of the ages are younger.
-Databases that don't support `median`: Druid, MariaDB, MongoDB, MySQL, SQLite, Vertica, and SQL Server. Presto only provides approximate results.
+Databases that don't support `median`: Druid, MariaDB, MongoDB, MySQL, SQLite, Vertica, and SQL Server.
+
+Presto only provides approximate results. BigQuery uses the lower of the middle 2 values, rather than their mean, when there is an even number of values.
Related: [Min](#min), [Max](#max), [Average](#average).
@@ -1050,6 +1052,8 @@ Limitations are noted for each aggregation and function above, and here they are
**Athena**: `convertTimezone`, `datetime`, `float`, `splitPart`.
+**BigQuery**: `Median`.
+
**Databricks**: `convertTimezone`, `datetime`, `float`, `splitPart`.
**Druid**: `Median`, `Percentile`, `StandardDeviation`, `power`, `log`, `exp`, `sqrt`, `Offset`, `datetime`, `float`, `splitPart`. Function `regexExtract` and some [type casting functions](#type-casting-functions) are only available for the Druid-JDBC driver.