Lisa Cao created SPARK-60399:
--------------------------------
Summary: Add a Metric Views programming guide and a CREATE METRIC
VIEW reference page
Key: SPARK-60399
URL: https://issues.apache.org/jira/browse/SPARK-60399
Project: Spark
Issue Type: Documentation
Components: SQL, Documentation
Affects Versions: 4.2.0, 4.3.0
Reporter: Lisa Cao
Spark 4.2.0 added metric views (SPARK-53524 and follow-ups): {{CREATE VIEW ...
WITH METRICS LANGUAGE YAML AS $$ ... $$}} and the {{MEASURE()}} aggregate
function. The docs have no page for either. The SQL reference has no entry for
the statement or the YAML definition format, and the only mention of
{{MEASURE}} is the generated built-in function entry.
This adds two pages:
* {{docs/metric-views-programming-guide.md}}, listed in the Programming Guides
menu next to Declarative Pipelines: what a metric view is and why a measure is
evaluated at the grain of the query that asks for it, a quick start, the key
concepts (source, dimensions, measures, filter, grain), defining metric views
in SQL, querying them from SQL and from the DataFrame API, modeling patterns
(ratios as single measures, date dimensions, derived dimensions in a SQL
source, a metric view over a metric view), lifecycle, and considerations.
* {{docs/sql-ref-syntax-ddl-create-metric-view.md}}, in the shape of the other
DDL reference pages (Description, Syntax, Parameters, Examples, Related
Statements): the statement and its clauses, the v0.1 YAML definition, how
column types are derived, what {{DESCRIBE EXTENDED}} shows, and the querying
rules.
It also links the reference page from the DDL list in {{sql-ref-syntax.md}},
{{CREATE VIEW}}, {{ALTER VIEW}} and {{SHOW CREATE TABLE}}, and adds a
{{MEASURE}} section to the aggregate function page.
Every statement on both pages was run against Spark 4.3.0 and the output shown
is what it produced.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
