add example

patterns/unnest_data.malloynb

@@ -0,0 +1,25 @@
+>>>markdown
+# Unnesting Data
+Malloy can naturally read unnested data.  Currently there is no mechanism to join something that unnest so for now you have to do it in SQL
+>>>malloy
+source: airports is duckdb.sql("""
+  SELECT 
+    *,
+    SPLIT(city,' ') as city_words
+  FROM '../data/airports.parquet' 
+""") extend {
+  measure:
+    airport_count is count()
+}
+>>>markdown
+We can now 'city_words', the words that appear in the city column
+>>>malloy
+run: airports -> {
+  group_by: city_word is city_words.value
+  aggregate: airport_count
+  # list 
+  nest: by_city is {
+    group_by: city
+    aggregate: airport_count
+  }
+}