Notebooks — schema_evolution-gh_archive-spark-iceberg¶

Auto-extracted from jupyter/notebook.ipynb and zeppelin/notebook.zpln. Both notebooks implement identical logic in PySpark and Scala.

1. Section map¶

Subsection	Scala (Zeppelin)	PySpark (Jupyter)
2.1 Setup	✓	✓
2.2 Read	✓	✓
2.3 Transform	✓	✓
2.4 Write	✓	✓
2.5 Verify	✓	✓

2. Walkthrough¶

2.1 Setup¶

Scala (Zeppelin):

import spark.implicits._
import org.apache.spark.sql.functions._
// spark is pre-bound by the Atlas Zeppelin interpreter

PySpark (Jupyter):

from pyspark.sql import SparkSession

spark = SparkSession.builder.remote("sc://spark-connect:15002").getOrCreate()

2.2 Read¶

Scala (Zeppelin):

spark.sql("CREATE TABLE IF NOT EXISTS lakehouse.silver.gh_events_se (id string, type string, actor_login string) USING iceberg")
spark.sql("INSERT INTO lakehouse.silver.gh_events_se VALUES ('1','PushEvent','octocat')")

PySpark (Jupyter):

spark.sql("CREATE TABLE IF NOT EXISTS lakehouse.silver.gh_events_se (id string, type string, actor_login string) USING iceberg")
spark.sql("INSERT INTO lakehouse.silver.gh_events_se VALUES ('1','PushEvent','octocat')")

2.3 Transform¶

Scala (Zeppelin):

spark.sql("ALTER TABLE lakehouse.silver.gh_events_se ADD COLUMN repo_name string")
spark.sql("ALTER TABLE lakehouse.silver.gh_events_se RENAME COLUMN type TO event_type")

PySpark (Jupyter):

spark.sql("ALTER TABLE lakehouse.silver.gh_events_se ADD COLUMN repo_name string")
spark.sql("ALTER TABLE lakehouse.silver.gh_events_se RENAME COLUMN type TO event_type")

2.4 Write¶

Scala (Zeppelin):

spark.sql("INSERT INTO lakehouse.silver.gh_events_se VALUES ('2','WatchEvent','torvalds','linux')")

PySpark (Jupyter):

spark.sql("INSERT INTO lakehouse.silver.gh_events_se VALUES ('2','WatchEvent','torvalds','linux')")

2.5 Verify¶

Scala (Zeppelin):

spark.sql("SELECT id, event_type, actor_login, repo_name FROM lakehouse.silver.gh_events_se ORDER BY id").show()

PySpark (Jupyter):

spark.sql("SELECT id, event_type, actor_login, repo_name FROM lakehouse.silver.gh_events_se ORDER BY id").show()

3. Scala / PySpark parity¶

Both notebooks share the same numbered sections and produce identical Iceberg tables; only the language and interpreter differ.

4. How to run¶

Open the scenario's zeppelin/notebook.zpln on the Atlas Zeppelin UI or jupyter/notebook.ipynb on JupyterHub, then run all paragraphs/cells top to bottom.