From 844a7fb5b0045c8d2fdc24a3221140f8546d9cd1 Mon Sep 17 00:00:00 2001
From: Michelle Lohr <98989192+chellebodnar-google@users.noreply.github.com>
Date: Thu, 17 Mar 2022 15:31:26 -0700
Subject: [PATCH] Feat: Add 311 example (#310)

* Create category_by_complaint_source.sql

Add a query snippet

* Create artifact.yaml

* Improve formatting

* Add better comments

Add comments and change variable "num_instances" to "num_compaints" for clarity
---
 .../docs/queries/artifact.yaml                |  5 ++++
 .../category_by_complaint_source.sql          | 30 +++++++++++++++++++
 2 files changed, 35 insertions(+)
 create mode 100644 datasets/san_francisco_311/docs/queries/artifact.yaml
 create mode 100644 datasets/san_francisco_311/docs/queries/category_by_complaint_source/category_by_complaint_source.sql

diff --git a/datasets/san_francisco_311/docs/queries/artifact.yaml b/datasets/san_francisco_311/docs/queries/artifact.yaml
new file mode 100644
index 000000000..db9c4cfee
--- /dev/null
+++ b/datasets/san_francisco_311/docs/queries/artifact.yaml
@@ -0,0 +1,5 @@
+artifact:
+  title: Analyze most prevelant category by complaint source of issue
+  description: In this tutorial we analyze the most likely category corresponding to each complaint source from all 311 reports in San Francisco.
+  vertical: government
+  tier: free
diff --git a/datasets/san_francisco_311/docs/queries/category_by_complaint_source/category_by_complaint_source.sql b/datasets/san_francisco_311/docs/queries/category_by_complaint_source/category_by_complaint_source.sql
new file mode 100644
index 000000000..71f245707
--- /dev/null
+++ b/datasets/san_francisco_311/docs/queries/category_by_complaint_source/category_by_complaint_source.sql
@@ -0,0 +1,30 @@
+# What is the most common category for each complaint source?
+
+WITH source_category_counts AS (
+  SELECT
+    source,
+    category,
+    COUNT(1) AS num_complaints
+  FROM
+    `bigquery-public-data`.san_francisco_311.311_service_requests 
+  GROUP BY
+    source, category
+)
+SELECT 
+  source,
+  category, 
+  num_complaints, 
+  num_complaints/total AS fraction_of_source
+FROM
+  (SELECT 
+     source,
+     category, 
+     num_complaints,
+     # Within each source, rank the categories by number of complaints in descending order. 
+     ROW_NUMBER() OVER (PARTITION BY source ORDER BY num_complaints DESC) AS rank,
+     # Compute the total number of complaints reported per source
+     SUM(num_complaints) OVER (PARTITION BY source) total
+   FROM source_category_counts)
+ WHERE
+   # Extract the most common category of complaint
+   rank = 1;