Book Image

Apache Spark 2.x Machine Learning Cookbook

By : Mohammed Guller, Siamak Amirghodsi, Shuen Mei, Meenakshi Rajendran, Broderick Hall
Book Image

Apache Spark 2.x Machine Learning Cookbook

By: Mohammed Guller, Siamak Amirghodsi, Shuen Mei, Meenakshi Rajendran, Broderick Hall

Overview of this book

Machine learning aims to extract knowledge from data, relying on fundamental concepts in computer science, statistics, probability, and optimization. Learning about algorithms enables a wide range of applications, from everyday tasks such as product recommendations and spam filtering to cutting edge applications such as self-driving cars and personalized medicine. You will gain hands-on experience of applying these principles using Apache Spark, a resilient cluster computing system well suited for large-scale machine learning tasks. This book begins with a quick overview of setting up the necessary IDEs to facilitate the execution of code examples that will be covered in various chapters. It also highlights some key issues developers face while working with machine learning algorithms on the Spark platform. We progress by uncovering the various Spark APIs and the implementation of ML algorithms with developing classification systems, recommendation engines, text analytics, clustering, and learning systems. Toward the final chapters, we’ll focus on building high-end applications and explain various unsupervised methodologies and challenges to tackle when implementing with big data ML systems.

Related Content you might be interested in

Current Title:

Small book image
Apache Spark 2.x Machine Learning Cookbook
Mohammed Guller | Siamak Amirghodsi | Shuen Mei | Meenakshi Rajendran | Broderick Hall
Sep, 2017 | 666 pages
Small book image
Machine Learning with Spark
Manpreet Singh Ghotra | Rajdeep Dua
Apr, 2017 | 532 pages
Small book image
Scala and Spark for Big Data Analytics
Md Rezaul Karim | Sridhar Alla
Jul, 2017 | 796 pages
Small book image
Hands-On Data Analysis with Scala
Rajesh Gupta
May, 2019 | 298 pages
Table of Contents (20 chapters)
Title Page
Title Page
Credits
Credits
About the Authors
About the Authors
About the Reviewer
About the Reviewer
www.PacktPub.com
www.PacktPub.com
Customer Feedback
Customer Feedback
Preface
Preface
Free Chapter
1
Practical Machine Learning with Spark Using Scala
Practical Machine Learning with Spark Using Scala
Introduction
Downloading and installing the JDK
Downloading and installing IntelliJ
Downloading and installing Spark
Configuring IntelliJ to work with Spark and run Spark ML sample codes
Running a sample ML code from Spark
Identifying data sources for practical machine learning
Running your first program using Apache Spark 2.0 with the IntelliJ IDE
How to add graphics to your Spark program
2
Just Enough Linear Algebra for Machine Learning with Spark
Just Enough Linear Algebra for Machine Learning with Spark
Introduction
Package imports and initial setup for vectors and matrices
Creating DenseVector and setup with Spark 2.0
Creating SparseVector and setup with Spark
Creating dense matrix and setup with Spark 2.0
Using sparse local matrices with Spark 2.0
Performing vector arithmetic using Spark 2.0
Performing matrix arithmetic using Spark 2.0
Exploring RowMatrix in Spark 2.0
Exploring Distributed IndexedRowMatrix in Spark 2.0
Exploring distributed CoordinateMatrix in Spark 2.0
Exploring distributed BlockMatrix in Spark 2.0
3
Spark's Three Data Musketeers for Machine Learning - Perfect Together
Spark's Three Data Musketeers for Machine Learning - Perfect Together
Introduction
Creating RDDs with Spark 2.0 using internal data sources
Creating RDDs with Spark 2.0 using external data sources
Transforming RDDs with Spark 2.0 using the filter() API
Transforming RDDs with the super useful flatMap() API
Transforming RDDs with set operation APIs
RDD transformation/aggregation with groupBy() and reduceByKey()
Transforming RDDs with the zip() API
Join transformation with paired key-value RDDs
Reduce and grouping transformation with paired key-value RDDs
Creating DataFrames from Scala data structures
Operating on DataFrames programmatically without SQL
Loading DataFrames and setup from an external source
Using DataFrames with standard SQL language - SparkSQL
Working with the Dataset API using a Scala Sequence
Creating and using Datasets from RDDs and back again
Working with JSON using the Dataset API and SQL together
Functional programming with the Dataset API using domain objects
4
Common Recipes for Implementing a Robust Machine Learning System
Common Recipes for Implementing a Robust Machine Learning System
Introduction
Spark's basic statistical API to help you build your own algorithms
ML pipelines for real-life machine learning applications
Normalizing data with Spark
Splitting data for training and testing
Common operations with the new Dataset API
Creating and using RDD versus DataFrame versus Dataset from a text file in Spark 2.0
LabeledPoint data structure for Spark ML
Getting access to Spark cluster in Spark 2.0
Getting access to Spark cluster pre-Spark 2.0
Getting access to SparkContext vis-a-vis SparkSession object in Spark 2.0
New model export and PMML markup in Spark 2.0
Regression model evaluation using Spark 2.0
Binary classification model evaluation using Spark 2.0
Multiclass classification model evaluation using Spark 2.0
Multilabel classification model evaluation using Spark 2.0
Using the Scala Breeze library to do graphics in Spark 2.0
5
Practical Machine Learning with Regression and Classification in Spark 2.0 - Part I
Practical Machine Learning with Regression and Classification in Spark 2.0 - Part I
Introduction
Fitting a linear regression line to data the old fashioned way
Generalized linear regression in Spark 2.0
Linear regression API with Lasso and L-BFGS in Spark 2.0
Linear regression API with Lasso and 'auto' optimization selection in Spark 2.0
Linear regression API with ridge regression and 'auto' optimization selection in Spark 2.0
Isotonic regression in Apache Spark 2.0
Multilayer perceptron classifier in Apache Spark 2.0
One-vs-Rest classifier (One-vs-All) in Apache Spark 2.0
Survival regression – parametric AFT model in Apache Spark 2.0
6
Practical Machine Learning with Regression and Classification in Spark 2.0 - Part II
Practical Machine Learning with Regression and Classification in Spark 2.0 - Part II
Introduction
Linear regression with SGD optimization in Spark 2.0
Logistic regression with SGD optimization in Spark 2.0
Ridge regression with SGD optimization in Spark 2.0
Lasso regression with SGD optimization in Spark 2.0
Logistic regression with L-BFGS optimization in Spark 2.0
Support Vector Machine (SVM) with Spark 2.0
Naive Bayes machine learning with Spark 2.0 MLlib
Exploring ML pipelines and DataFrames using logistic regression in Spark 2.0
7
Recommendation Engine that Scales with Spark
Recommendation Engine that Scales with Spark
Introduction
Setting up the required data for a scalable recommendation engine in Spark 2.0
Exploring the movies data details for the recommendation system in Spark 2.0
Exploring the ratings data details for the recommendation system in Spark 2.0
Building a scalable recommendation engine using collaborative filtering in Spark 2.0
8
Unsupervised Clustering with Apache Spark 2.0
Unsupervised Clustering with Apache Spark 2.0
Introduction
Building a KMeans classifying system in Spark 2.0
Bisecting KMeans, the new kid on the block in Spark 2.0
Using Gaussian Mixture and Expectation Maximization (EM) in Spark to classify data
Classifying the vertices of a graph using Power Iteration Clustering (PIC) in Spark 2.0
Latent Dirichlet Allocation (LDA) to classify documents and text into topics
Streaming KMeans to classify data in near real-time
9
Optimization - Going Down the Hill with Gradient Descent
Optimization - Going Down the Hill with Gradient Descent
Introduction
Optimizing a quadratic cost function and finding the minima using just math to gain insight
Coding a quadratic cost function optimization using Gradient Descent (GD) from scratch
Coding Gradient Descent optimization to solve Linear Regression from scratch
Normal equations as an alternative for solving Linear Regression in Spark 2.0
10
Building Machine Learning Systems with Decision Tree and Ensemble Models
Building Machine Learning Systems with Decision Tree and Ensemble Models
Introduction
Getting and preparing real-world medical data for exploring Decision Trees and Ensemble models in Spark 2.0
Building a classification system with Decision Trees in Spark 2.0
Solving Regression problems with Decision Trees in Spark 2.0
Building a classification system with Random Forest Trees in Spark 2.0
Solving regression problems with Random Forest Trees in Spark 2.0
Building a classification system with Gradient Boosted Trees (GBT) in Spark 2.0
Solving regression problems with Gradient Boosted Trees (GBT) in Spark 2.0
11
Curse of High-Dimensionality in Big Data
Curse of High-Dimensionality in Big Data
Introduction
Two methods of ingesting and preparing a CSV file for processing in Spark
Singular Value Decomposition (SVD) to reduce high-dimensionality in Spark
Principal Component Analysis (PCA) to pick the most effective latent factor for machine learning in Spark
12
Implementing Text Analytics with Spark 2.0 ML Library
Implementing Text Analytics with Spark 2.0 ML Library
Introduction
Doing term frequency with Spark - everything that counts
Displaying similar words with Spark using Word2Vec
Downloading a complete dump of Wikipedia for a real-life Spark ML project
Using Latent Semantic Analysis for text analytics with Spark 2.0
Topic modeling with Latent Dirichlet allocation in Spark 2.0
13
Spark Streaming and Machine Learning Library
Spark Streaming and Machine Learning Library
Introduction
Structured streaming for near real-time machine learning
Streaming DataFrames for real-time machine learning
Streaming Datasets for real-time machine learning
Streaming data and debugging with queueStream
Downloading and understanding the famous Iris data for unsupervised classification
Streaming KMeans for a real-time on-line classifier
Downloading wine quality data for streaming regression
Streaming linear regression for a real-time regression
Downloading Pima Diabetes data for supervised classification
Streaming logistic regression for an on-line classifier

Running your first program using Apache Spark 2.0 with the IntelliJ IDE

The purpose of this is to get you comfortable with compiling and running a recipe using the Spark 2.0 development environment you just set up. We will explore the components and steps in later chapters.

We are going to write our own version of the Spark 2.0.0 program and examine the output so we can understand how it works. To emphasize, this short recipe is only a simple RDD program with Scala sugar syntax to make sure you have set up your environment correctly before starting to work more complicated recipes.

How to do it...

  1. Start a new project in IntelliJ or in an IDE of your choice. Make sure that the necessary JAR files are included.
  2. Download the sample code for the book, find the myFirstSpark20.scala file, and place the code in the following directory.

We installed Spark 2.0 in the C:\spark-2.0.0-bin-hadoop2.7\ directory on a Windows machine.

  1. Place the myFirstSpark20.scala file in the C:\spark-2.0.0-bin-hadoop2.7\examples\src\main\scala\spark\ml\cookbook\chapter1 directory:

Mac users note that we installed Spark 2.0 in the /Users/USERNAME/spark/spark-2.0.0-bin-hadoop2.7/ directory on a Mac machine.

Place the myFirstSpark20.scala file in the /Users/USERNAME/spark/spark-2.0.0-bin-hadoop2.7/examples/src/main/scala/spark/ml/cookbook/chapter1 directory.

  1. Set up the package location where the program will reside:
package spark.ml.cookbook.chapter1
  1. Import the necessary packages for the Spark session to gain access to the cluster and log4j.Logger to reduce the amount of output produced by Spark:
import org.apache.spark.sql.SparkSession 
import org.apache.log4j.Logger 
import org.apache.log4j.Level
  1. Set output level to ERROR to reduce Spark's logging output:
Logger.getLogger("org").setLevel(Level.ERROR)
  1. Initialize a Spark session by specifying configurations with the builder pattern, thus making an entry point available for the Spark cluster:
val spark = SparkSession 
.builder 
.master("local[*]")
 .appName("myFirstSpark20") 
.config("spark.sql.warehouse.dir", ".") 
.getOrCreate()

The myFirstSpark20 object will run in local mode. The previous code block is a typical way to start creating a SparkSession object.

  1. We then create two array variables:
val x = Array(1.0,5.0,8.0,10.0,15.0,21.0,27.0,30.0,38.0,45.0,50.0,64.0) 
val y = Array(5.0,1.0,4.0,11.0,25.0,18.0,33.0,20.0,30.0,43.0,55.0,57.0)
  1. We then let Spark create two RDDs based on the array created before:
val xRDD = spark.sparkContext.parallelize(x) 
val yRDD = spark.sparkContext.parallelize(y)
  1. Next, we let Spark operate on the RDD; the zip() function will create a new RDD from the two RDDs mentioned before:
val zipedRDD = xRDD.zip(yRDD) 
zipedRDD.collect().foreach(println)

In the console output at runtime (more details on how to run the program in the IntelliJ IDE in the following steps), you will see this:

  1. Now, we sum up the value for xRDD and yRDD and calculate the new zipedRDD sum value. We also calculate the item count for zipedRDD:
val xSum = zipedRDD.map(_._1).sum() 
val ySum = zipedRDD.map(_._2).sum() 
val xySum= zipedRDD.map(c => c._1 * c._2).sum() 
val n= zipedRDD.count()
  1. We print out the value calculated previously in the console:
println("RDD X Sum: " +xSum) 
println("RDD Y Sum: " +ySum) 
println("RDD X*Y Sum: "+xySum) 
println("Total count: "+n)

Here's the console output:

  1. We close the program by stopping the Spark session:
spark.stop()
  1. Once the program is complete, the layout of myFirstSpark20.scala in the IntelliJ project explorer will look like the following:
  1. Make sure there is no compiling error. You can test this by rebuilding the project:

Once the rebuild is complete, there should be a build completed message on the console:

Information: November 18, 2016, 11:46 AM - Compilation completed successfully with 1 warning in 55s 648ms
  1. You can run the previous program by right-clicking on the myFirstSpark20 object in the project explorer and selecting the context menu option (shown in the next screenshot) called Run myFirstSpark20.

Note

You can also use the Run menu from the menu bar to perform the same action.

  1. Once the program is successfully executed, you will see the following message:
Process finished with exit code 0

This is also shown in the following screenshot:

  1. Mac users with IntelliJ will be able to perform this action using the same context menu.

Note

Place the code in the correct path.

How it works...

In this example, we wrote our first Scala program, myFirstSpark20.scala, and displayed the steps to execute the program in IntelliJ. We placed the code in the path described in the steps for both Windows and Mac.

In the myFirstSpark20 code, we saw a typical way to create a SparkSession object and how to configure it to run in local mode using the master() function. We created two RDDs out of the array objects and used a simple zip() function to create a new RDD.

We also did a simple sum calculation on the RDDs that were created and then displayed the result in the console. Finally, we exited and released the resource by calling spark.stop().

There's more...

can be downloaded from http://spark.apache.org/downloads.html.

Documentation for Spark 2.0 related to RDD can be found at http://spark.apache.org/docs/latest/programming-guide.html#rdd-operations.

See also

Previous Section
End of Section 9
Next Section