From e1bdb102846f56e4296f8a3ed6793485f25c54da Mon Sep 17 00:00:00 2001
From: Andrew Atkinson <andyatkinson@gmail.com>
Date: Fri, 6 Jun 2025 15:09:00 -0500
Subject: [PATCH] Report on cardinality for columns for all tables

Show n_distinct, show estimated rows

Exclude non-app db schema
---
 high_cardinality_columns.sql | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)
 create mode 100644 high_cardinality_columns.sql

diff --git a/high_cardinality_columns.sql b/high_cardinality_columns.sql
new file mode 100644
index 0000000..35147af
--- /dev/null
+++ b/high_cardinality_columns.sql
@@ -0,0 +1,23 @@
+-- run "analyze" to analyze all tables
+SELECT
+    schemaname,
+    tablename,
+    attname AS column,
+    n_distinct,
+    CASE WHEN n_distinct >= 0 THEN
+        n_distinct::text
+    ELSE
+        ('~' || (- n_distinct * rel.reltuples)) -- negative means "estimated fraction of table"
+    END AS estimated_distinct,
+    rel.reltuples::numeric AS estimated_rows
+FROM
+    pg_stats s
+    JOIN pg_class rel ON rel.relname = s.tablename
+    JOIN pg_namespace nsp ON nsp.oid = rel.relnamespace
+        AND nsp.nspname = s.schemaname
+WHERE
+    n_distinct IS NOT NULL
+AND schemaname NOT IN ('pg_catalog','information_schema')
+ORDER BY
+    ABS(n_distinct) DESC
+LIMIT 10;