week_5_batch_processing

Week 5: Batch Processing

5.1 Introduction

• 🎥 5.1.1 Introduction to Batch Processing
• 🎥 5.1.2 Introduction to Spark

5.2 Installation

Follow these intructions to install Spark:

• Windows
• Linux
• MacOS

And follow this to run PySpark in Jupyter

• 🎥 5.2.1 (Optional) Installing Spark (Linux)

5.3 Spark SQL and DataFrames

• 🎥 5.3.1 First Look at Spark/PySpark
• 🎥 5.3.2 Spark Dataframes
• 🎥 5.3.3 (Optional) Preparing Yellow and Green Taxi Data

Script to prepare the Dataset download_data.sh

Note: The other way to infer the schema (apart from pandas) for the csv files, is to set the inferSchema option to true while reading the files in Spark.

• 🎥 5.3.4 SQL with Spark

5.4 Spark Internals

• 🎥 5.4.1 Anatomy of a Spark Cluster
• 🎥 5.4.2 GroupBy in Spark
• 🎥 5.4.3 Joins in Spark

5.5 (Optional) Resilient Distributed Datasets

• 🎥 5.5.1 Operations on Spark RDDs
• 🎥 5.5.2 Spark RDD mapPartition

5.6 Running Spark in the Cloud

• 🎥 5.6.1 Connecting to Google Cloud Storage
• 🎥 5.6.2 Creating a Local Spark Cluster
• 🎥 5.6.3 Setting up a Dataproc Cluster
• 🎥 5.6.4 Connecting Spark to Big Query

Homework

• Homework

Community notes

Did you take notes? You can share them here.

• Notes by Alvaro Navas
• Sandy's DE Learning Blog
• Notes by Alain Boisvert
• Alternative : Using docker-compose to launch spark by rafik
• Marcos Torregrosa's blog (spanish)
• Notes by Victor Padilha
• Add your notes here (above this line)