SlideShare a Scribd company logo

Collections.compare(JDK, Eclipse, Guava, Apache...); GIDS 2017

Download as PPTX, PDF
N
Nikhil Nanivadekar

The document discusses the Java Collection Framework, highlighting its history, use cases, and comparing it with other collection frameworks like Apache Commons, Eclipse Collections, Google Guava, and Javaslang. It addresses specific problems such as creating a deck of cards using different collection types and performance benchmarks for operations like grouping and creating Cartesian products. Author Nikhil J. Nanivadekar, alongside others, focuses on various implementations and optimizations of collection structures within Java.

Technology
1 of 71
Downloaded 18 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Collections.compare(() -> { JDK; Apache; Eclipse; Guava...}); Nikhil J. Nanivadekar
#CollectionsCompare #GIDS17 @NikhilNanivade Agenda 1. Introduction 2. History of Java util collections framework 3. Gaps in use cases from the timeline 4. Timeline with other collection frameworks 5. Use case, domain diagrams 6. Deep dive in use case 7. Memory characteristics
#CollectionsCompare #GIDS17 @NikhilNanivade Authors • Donald Raab (@TheDonRaab) • Leonardo Lima (@leomrlima) • Nikhil J. Nanivadekar (@NikhilNanivade)
#CollectionsCompare #GIDS17 @NikhilNanivade Speaker Nikhil J. Nanivadekar – Mechanical Engineer – Software developer – Vice President, Technology at Goldman Sachs – From Pune, Maharashtra currently in Salt Lake City, Utah
java.util Collections
#CollectionsCompare #GIDS17 @NikhilNanivade java.util Collections • Java Collection Framework has been around since 1998/JDK 1.2. Very basic but critical collection support for Java. • Interfaces: Maps and Collections (List, Set, Queue, Deque) • Implementations: Basic and Concurrent versions of the interfaces, sorted and concurrent as well • Algorithms: Sorting, Shuffling, Routine Data Manipulation, Searching, Composition • Best way to get started is using the simple and concurrent tutorials
#CollectionsCompare #GIDS17 @NikhilNanivade Java Collections Frameworks Java 2 1998 • Interfaces • Collection • List • Set • Map • SortedSet • SortedMap Java 5 2005 • Generics • For-each loop • Interfaces • Iterable • Queue • ConcurrentMap • BlockingQueue Java 6 2006 •Interfaces •NavigableSet •NavigableMap •Deque •BlockingDeque •ConcurrentNavigableMap Java 7 2011 • Interfaces • TransferQueue Java 8 2014 • Lambdas • Method Refs • Default Methods • Interfaces • BaseStream • Stream • IntStream • LongStream • DoubleStream • Collector • Spliterator • PrimitiveIterator
!java.util Collections?
#CollectionsCompare #GIDS17 @NikhilNanivade !java.util Collections • We’re considering 4 collection frameworks, in alphabetical order: • Apache Commons Collection – version 4.1 • Eclipse Collections (Formerly GS Collections) – version 8.1.0 • Google Guava – version 21.0 • Javaslang (now called Vavr) – version 2.1.0-alpha
#CollectionsCompare #GIDS17 @NikhilNanivade Collection Framework Timelines Java 2 1998 •Jakarta Collections 1.0 •2001 •Apache Collections 3.0 •2004 Java 6 2006 •Google Collections 1.0 •Dec. 2009 Java 7 2011 •Google Guava 10.0 •Sept. 2011 •GS Collections 1.0 •Jan. 2012 •Apache Collections 4.0 •Nov. 2013 Java 8 2014 •Javaslang 1.0 •Mar. 2014 •Apache Collections 4.1 •Nov. 2015 •Eclipse Collections 7.0 •Jan. 2016 •Eclipse Collections 8.0 (Java 8) •Sept. 2016 •Google Guava 20.0 •Oct. 2016 •Javaslang 2.1a •Nov. 2016 March 2017 •Google Guava 21.0 (Java 8) •Jan. 2017 •Eclipse Collections 8.1 •Mar. 2017
Collections.compare();
#CollectionsCompare #GIDS17 @NikhilNanivade Use Case – Deck of Cards The Collections Compare Project: https://github.com/nikhilnanivadekar/CollectionsCompare
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Benchmarks • JMH – Java Microbenchmark Harness • http://openjdk.java.net/projects/code-tools/jmh/ • Measure Reported – Operations per second • Bigger numbers are better • 4 Core Intel I7, 50 Warm-up iterations, 30 measurement iterations, two forks
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deck of Cards 102,037 79,343 76,982 56,224 91,017 0 20,000 40,000 60,000 80,000 100,000 120,000 Scoreops/s Framework Deck of Cards Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Shared Code – Cartesian Product public static Stream<Card> streamCards() { return Card.cartesianProduct( EnumSet.allOf(Rank.class), EnumSet.allOf(Suit.class), Card::new); } private static <A, B, C> Stream<C> cartesianProduct( Set<A> set1, Set<B> set2, Function2<A, B, C> function) { return set1.stream().flatMap(first -> set2.stream().map(second -> function.apply(first, second))); }
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Cartesian Product 185,087 203,660 465,999 127,439 114,634 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 Scoreops/s Framework Cartesian Product Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Cartesian Product public class JDK8DeckOfCards { private SortedSet<Card> cards; public JDK8DeckOfCards() { this.cards = Card.streamCards() .collect(Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Cartesian Product public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; public ApacheCommonsDeckOfCards() { this.cards = Card.streamCards() .collect(Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Cartesian Product public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; public EclipseCollectionsDeckOfCards() { this.cards = SortedSets.immutable.with( Card.streamCards() .toArray(Card[]::new)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Cartesian Product public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; public GoogleGuavaDeckOfCards() { this.cards = Card.streamCards() .collect(ImmutableSortedSet .toImmutableSortedSet( Comparator.naturalOrder())); }
#CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Cartesian Product public class JavaSlangDeckOfCards { private SortedSet<Card> cards; public JavaSlangDeckOfCards() { this.cards = Card.streamCards() .collect(TreeSet.collector()); }
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - groupBy 540,396 183,848 231,321 176,412 353,651 - 100,000 200,000 300,000 400,000 500,000 600,000 Scoreops/s Framework groupBy Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Group By public class JDK8DeckOfCards { private Map<Suit, SortedSet<Card>> cardsBySuit; public JDK8DeckOfCards() { this.cardsBySuit = Collections.unmodifiableMap( this.cards.stream() .collect(Collectors.groupingBy( Card::getSuit, Collectors.mapping( Function.identity(), Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet))))); }
#CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Group By public class ApacheCommonsDeckOfCards { private MultiValuedMap<Suit, Card> cardsBySuit; public ApacheCommonsDeckOfCards() { SetValuedMap<Suit, Card> cbs = MultiMapUtils.newSetValuedHashMap(); this.cards.forEach(card -> cbs.put(card.getSuit(), card)); this.cardsBySuit = MultiMapUtils.unmodifiableMultiValuedMap(cbs); }
#CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Group By public class EclipseCollectionsDeckOfCards { private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public EclipseCollectionsDeckOfCards() { this.cardsBySuit = this.cards.groupBy(Card::getSuit); }
#CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Group By public class GoogleGuavaDeckOfCards { private ImmutableSetMultimap<Suit, Card> cardsBySuit; public GoogleGuavaDeckOfCards() { ImmutableSetMultimap.Builder<Suit, Card> builder = new ImmutableSetMultimap.Builder<Suit, Card>() .orderValuesBy(Comparator.naturalOrder()); this.cards.forEach(card -> builder.put(card.getSuit(), card)); this.cardsBySuit = builder.build(); }
#CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Group By public class JavaSlangDeckOfCards { private Map<Suit, ? extends SortedSet<Card>> cardsBySuit; public JavaSlangDeckOfCards() { this.cardsBySuit = this.cards.groupBy(Card::getSuit); }
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deck of Cards 102,037 79,343 76,982 56,224 91,017 0 20,000 40,000 60,000 80,000 100,000 120,000 Scoreops/s Framework Deck of Cards Apache EC Guava Javaslang JDK 540,396 183,848231,321176,412 353,651 - 500,000 1,000,000 Scoreops/s Framework groupBy Apache EC Guava Javaslang JDK 185,087 203,660 465,999 127,439 114,634 0 200,000 400,000 600,000 Scoreops/s Framework Cartesian Product Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - countsBySuit 706,272 1,096,527 533,728 440,786 724,526 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - countsByRank 539,252 827,986 444,832 390,134 591,869 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Count By public Map<Suit, Long> countsBySuit() { return this.cards.stream() .collect(Collectors.groupingBy( Card::getSuit, Collectors.counting())); } public Map<Rank, Long> countsByRank() { return this.cards.stream() .collect(Collectors.groupingBy( Card::getRank, Collectors.counting())); }
#CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Count By public Bag<Suit> countsBySuit() { return this.cards.stream() .map(Card::getSuit) .collect(Collectors.toCollection(HashBag::new)); } public MultiSet<Rank> countsByRank() { return this.cards.stream() .map(Card::getRank) .collect(Collectors.toCollection( HashMultiSet::new)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Count By public Bag<Suit> countsBySuit() { return this.cards.asLazy() .collect(Card::getSuit) .toBag(); } public Bag<Rank> countsByRank() { return this.cards.asLazy() .collect(Card::getRank) .toBag(); }
#CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Count By public Multiset<Suit> countsBySuit() { return this.cards.stream() .map(Card::getSuit) .collect(Collectors.toCollection( HashMultiset::create)); } public Multiset<Rank> countsByRank() { return this.cards.stream() .map(Card::getRank) .collect(Collectors.toCollection( HashMultiset::create)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Count By public java.util.Map<Suit, Long> countsBySuit() { return this.cards.collect( Collectors.groupingBy( Card::getSuit, Collectors.counting())); } public java.util.Map<Rank, Long> countsByRank() { return this.cards.collect( Collectors.groupingBy( Card::getRank, Collectors.counting())); }
#CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
#CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deal Hands 291,488 502,448 305,673 372,466 300,975 0 100,000 200,000 300,000 400,000 500,000 600,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
#CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Deal Hands public List<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect( Collectors.collectingAndThen( Collectors.toList(), Collections::unmodifiableList)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Deal Hands public List<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect( Collectors.collectingAndThen( Collectors.toList(), Collections::unmodifiableList)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Deal Hands public ImmutableList<Set<Card>> dealHands( MutableStack<Card> shuffled, int hands, int cardsPerHand) { return IntInterval.oneTo(hands) .collect(i -> this.deal(shuffled, cardsPerHand)); }
#CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Deal Hands public ImmutableList<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect(ImmutableList.toImmutableList()); }
#CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Deal Hands public List<Set<Card>> dealHands( Stack<Card> shuffled, int hands, int cardsPerHand) { List<Set<Card>> list = List.empty(); for (int i = 0; i < hands; i++) { Tuple2<Set<Card>, ? extends Stack<Card>> tuple2 = this.deal(shuffled, cardsPerHand); shuffled = tuple2._2(); list = list.append(tuple2._1()); } return list; }
Memory Benchmarks
#CollectionsCompare #GIDS17 @NikhilNanivade Memory Benchmarks • ObjectExplorer, by Dimitris Andrew • Open Source under Apache License 2.0 • Measure Reported – – Memory (Kb) vs Size – Memory (Kb) vs List, Set, Map with 1,000,000 elements • Smaller the better Sources – GitHub: https://github.com/DimitrisAndreou/memory-measurer – Found from StackOverflow: http://stackoverflow.com/a/9368930/5182052
#CollectionsCompare #GIDS17 @NikhilNanivade List<Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade List<Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade List<String>
#CollectionsCompare #GIDS17 @NikhilNanivade List<String>
#CollectionsCompare #GIDS17 @NikhilNanivade Set<Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade Set<Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade Set<String>
#CollectionsCompare #GIDS17 @NikhilNanivade Set<String>
#CollectionsCompare #GIDS17 @NikhilNanivade Map<Integer, Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade Map<Integer, Integer>
#CollectionsCompare #GIDS17 @NikhilNanivade Map<String, String>
#CollectionsCompare #GIDS17 @NikhilNanivade Map<String, String>
Summary
#CollectionsCompare #GIDS17 @NikhilNanivade Link • The Collections Compare Project https://github.com/nikhilnanivadekar/CollectionsCompare
#CollectionsCompare #GIDS17 @NikhilNanivade Collection Framework Comparison Feature JDK Apache Guava Eclipse Javaslang 1st, 2nd, 3rd List, Set, Map ✔  ✖ ✔  JDK, EC, JS Multiset / Bag ✖ ✔ ✔ ✔ ✖ GG, EC, AC Multimap ✖ ✔ ✔ ✔ ✔ GG, EC, AC BiMap ✖ ✔ ✔ ✔ ✖ GG, EC, AC Stack ✔  ✖ ✔ ✔ EC, JDK, JS Tree / Trie ✖ ✔ ✖ ✖ ✔ JS, AC Table ✖ ✖ ✔ ✖ ✖ GG Additional Types ✔ ✔ ✔   AC, JDK, GG Immutable? ✖ ✖ ✔ ✔ ✔ JS, EC, GG Primitives?  ✖  ✔ ✖ EC, JDK, GG Fluent API    ✔ ✔ EC, JS, JDK (E), (L), (S), (P)* , ✔, ✔, ✔ , , ✔, ✖ ✖, , ✔, ✖ ✔, ✔, ✔, ✔ ✔, ✔, ✔, ✖ EC, JDK, JS
Thanks! The Collections Compare Project https://github.com/nikhilnanivadekar/CollectionsCompare

More Related Content

PPTX
Collections.compare(JDK, Eclipse, Guava, Apache...);
Leonardo De Moura Rocha Lima
 
PPTX
Collections.compare(() -> JDK; Apache; Eclipse, Guava...});
Java Usergroup Berlin-Brandenburg
 
PPTX
Collections.compare(JDK, Eclipse, Guava, Apache...); DevoxxUS 2017
Nikhil Nanivadekar
 
PDF
JavaOne 2017 - Collections.compare:JDK, Eclipse, Guava, Apache... [CON1754]
Leonardo De Moura Rocha Lima
 
PDF
So S.O.L.I.D Fu - Designing Better Code
Neil Crookes
 
PDF
Find your data - using GraphDB capabilities in XPages applications - ICS.UG 2016
ICS User Group
 
PDF
Apache Spark for Library Developers with Erik Erlandson and William Benton
Databricks
 
PDF
Visage Android Hands-on Lab (OSCON)
Stephen Chin
 
Collections.compare(JDK, Eclipse, Guava, Apache...);
Leonardo De Moura Rocha Lima
 
Collections.compare(() -> JDK; Apache; Eclipse, Guava...});
Java Usergroup Berlin-Brandenburg
 
Collections.compare(JDK, Eclipse, Guava, Apache...); DevoxxUS 2017
Nikhil Nanivadekar
 
JavaOne 2017 - Collections.compare:JDK, Eclipse, Guava, Apache... [CON1754]
Leonardo De Moura Rocha Lima
 
So S.O.L.I.D Fu - Designing Better Code
Neil Crookes
 
Find your data - using GraphDB capabilities in XPages applications - ICS.UG 2016
ICS User Group
 
Apache Spark for Library Developers with Erik Erlandson and William Benton
Databricks
 
Visage Android Hands-on Lab (OSCON)
Stephen Chin
 

Similar to Collections.compare(JDK, Eclipse, Guava, Apache...); GIDS 2017 (20)

PPTX
10 Strategies for Developing Reliable Jakarta EE & MicroProfile Applications ...
Payara
 
PPTX
jCardSim - development platform for Java Card Applications
Mikhail Dudarev
 
PDF
The cost of learning - advantage of mixer2
Y Watanabe
 
PDF
JavaCro 2014 Scala and Java EE 7 Development Experiences
Peter Pilgrim
 
PDF
JavaCro'14 - Scala and Java EE 7 Development Experiences – Peter Pilgrim
HUJAK - Hrvatska udruga Java korisnika / Croatian Java User Association
 
PPTX
Preparing for java 9 modules upload
Ryan Cuprak
 
PDF
DIとトレイとによるAndroid開発の効率化
Tomoharu ASAMI
 
PPTX
Apache spark core
Thành Nguyễn
 
PDF
Scala Frustrations
takezoe
 
PPTX
Modules in Java? Finally! (OpenJDK 9 Jigsaw, JSR376)
Mihail Stoynov
 
PDF
Module, AMD, RequireJS
偉格 高
 
PDF
[Pgday.Seoul 2018] 이기종 DB에서 PostgreSQL로의 Migration을 위한 DB2PG
PgDay.Seoul
 
PPT
Backbone js-slides
DrupalCamp Kyiv Рысь
 
PDF
Using Backbone.js with Drupal 7 and 8
Ovadiah Myrgorod
 
PDF
[DSC 2016] 系列活動：李泳泉 / 星火燎原 - Spark 機器學習初探
台灣資料科學年會
 
PPTX
Packing for the Web with Webpack
Thiago Temple
 
PDF
"Xapi-lang For declarative code generation" By James Nelson
GWTcon
 
PDF
Webpack Encore - Asset Management for the rest of us
Stefan Adolf
 
PPTX
Php on the Web and Desktop
Elizabeth Smith
 
PDF
S313431 JPA 2.0 Overview
Ludovic Champenois
 
10 Strategies for Developing Reliable Jakarta EE & MicroProfile Applications ...
Payara
 
jCardSim - development platform for Java Card Applications
Mikhail Dudarev
 
The cost of learning - advantage of mixer2
Y Watanabe
 
JavaCro 2014 Scala and Java EE 7 Development Experiences
Peter Pilgrim
 
JavaCro'14 - Scala and Java EE 7 Development Experiences – Peter Pilgrim
HUJAK - Hrvatska udruga Java korisnika / Croatian Java User Association
 
Preparing for java 9 modules upload
Ryan Cuprak
 
DIとトレイとによるAndroid開発の効率化
Tomoharu ASAMI
 
Apache spark core
Thành Nguyễn
 
Scala Frustrations
takezoe
 
Modules in Java? Finally! (OpenJDK 9 Jigsaw, JSR376)
Mihail Stoynov
 
Module, AMD, RequireJS
偉格 高
 
[Pgday.Seoul 2018] 이기종 DB에서 PostgreSQL로의 Migration을 위한 DB2PG
PgDay.Seoul
 
Backbone js-slides
DrupalCamp Kyiv Рысь
 
Using Backbone.js with Drupal 7 and 8
Ovadiah Myrgorod
 
[DSC 2016] 系列活動：李泳泉 / 星火燎原 - Spark 機器學習初探
台灣資料科學年會
 
Packing for the Web with Webpack
Thiago Temple
 
"Xapi-lang For declarative code generation" By James Nelson
GWTcon
 
Webpack Encore - Asset Management for the rest of us
Stefan Adolf
 
Php on the Web and Desktop
Elizabeth Smith
 
S313431 JPA 2.0 Overview
Ludovic Champenois
 
Ad

Recently uploaded (20)

PPT
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
PDF
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
Libreria ERP
 
PPTX
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
PDF
Encapsulation_ Review paper, used for researhc scholars
gurumoop
 
PDF
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
PDF
NewMind AI Weekly Chronicles - August'25-Week II
NewMind AI
 
PDF
Advanced methodologies resolving dimensionality complications for autism neur...
IAESIJAI
 
PDF
MIND Revenue Release Quarter 2 2025 Press Release
MIND CTI
 
PDF
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
PDF
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
PPTX
Spectroscopy.pptx food analysis technology
nawahassan45
 
PDF
The Rise and Fall of 3GPP – Time for a Sabbatical?
3G4G
 
PPTX
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
PDF
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
PDF
A comparative analysis of optical character recognition models for extracting...
IAESIJAI
 
PDF
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 
PDF
cuic standard and advanced reporting.pdf
SimoneTitaniumTomase
 
PDF
Per capita expenditure prediction using model stacking based on satellite ima...
IAESIJAI
 
“AI and Expert System Decision Support & Business Intelligence Systems”
ZubinRadhakrishnan
 
Profit Center Accounting in SAP S/4HANA, S4F28 Col11
Libreria ERP
 
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
Encapsulation_ Review paper, used for researhc scholars
gurumoop
 
7 ChatGPT Prompts to Help You Define Your Ideal Customer Profile.pdf
SOFTTECHHUB
 
NewMind AI Weekly Chronicles - August'25-Week II
NewMind AI
 
Advanced methodologies resolving dimensionality complications for autism neur...
IAESIJAI
 
MIND Revenue Release Quarter 2 2025 Press Release
MIND CTI
 
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
Safe Software
 
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
Spectroscopy.pptx food analysis technology
nawahassan45
 
The Rise and Fall of 3GPP – Time for a Sabbatical?
3G4G
 
20250228 LYD VKU AI Blended-Learning.pptx
DatVoNgoc
 
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
Reach Out and Touch Someone: Haptics and Empathic Computing
Mark Billinghurst
 
A comparative analysis of optical character recognition models for extracting...
IAESIJAI
 
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 
cuic standard and advanced reporting.pdf
SimoneTitaniumTomase
 
Per capita expenditure prediction using model stacking based on satellite ima...
IAESIJAI
 
Ad

Collections.compare(JDK, Eclipse, Guava, Apache...); GIDS 2017

  • 1. Collections.compare(() -> { JDK; Apache; Eclipse; Guava...}); Nikhil J. Nanivadekar
  • 2. #CollectionsCompare #GIDS17 @NikhilNanivade Agenda 1. Introduction 2. History of Java util collections framework 3. Gaps in use cases from the timeline 4. Timeline with other collection frameworks 5. Use case, domain diagrams 6. Deep dive in use case 7. Memory characteristics
  • 3. #CollectionsCompare #GIDS17 @NikhilNanivade Authors • Donald Raab (@TheDonRaab) • Leonardo Lima (@leomrlima) • Nikhil J. Nanivadekar (@NikhilNanivade)
  • 4. #CollectionsCompare #GIDS17 @NikhilNanivade Speaker Nikhil J. Nanivadekar – Mechanical Engineer – Software developer – Vice President, Technology at Goldman Sachs – From Pune, Maharashtra currently in Salt Lake City, Utah
  • 6. #CollectionsCompare #GIDS17 @NikhilNanivade java.util Collections • Java Collection Framework has been around since 1998/JDK 1.2. Very basic but critical collection support for Java. • Interfaces: Maps and Collections (List, Set, Queue, Deque) • Implementations: Basic and Concurrent versions of the interfaces, sorted and concurrent as well • Algorithms: Sorting, Shuffling, Routine Data Manipulation, Searching, Composition • Best way to get started is using the simple and concurrent tutorials
  • 7. #CollectionsCompare #GIDS17 @NikhilNanivade Java Collections Frameworks Java 2 1998 • Interfaces • Collection • List • Set • Map • SortedSet • SortedMap Java 5 2005 • Generics • For-each loop • Interfaces • Iterable • Queue • ConcurrentMap • BlockingQueue Java 6 2006 •Interfaces •NavigableSet •NavigableMap •Deque •BlockingDeque •ConcurrentNavigableMap Java 7 2011 • Interfaces • TransferQueue Java 8 2014 • Lambdas • Method Refs • Default Methods • Interfaces • BaseStream • Stream • IntStream • LongStream • DoubleStream • Collector • Spliterator • PrimitiveIterator
  • 9. #CollectionsCompare #GIDS17 @NikhilNanivade !java.util Collections • We’re considering 4 collection frameworks, in alphabetical order: • Apache Commons Collection – version 4.1 • Eclipse Collections (Formerly GS Collections) – version 8.1.0 • Google Guava – version 21.0 • Javaslang (now called Vavr) – version 2.1.0-alpha
  • 10. #CollectionsCompare #GIDS17 @NikhilNanivade Collection Framework Timelines Java 2 1998 •Jakarta Collections 1.0 •2001 •Apache Collections 3.0 •2004 Java 6 2006 •Google Collections 1.0 •Dec. 2009 Java 7 2011 •Google Guava 10.0 •Sept. 2011 •GS Collections 1.0 •Jan. 2012 •Apache Collections 4.0 •Nov. 2013 Java 8 2014 •Javaslang 1.0 •Mar. 2014 •Apache Collections 4.1 •Nov. 2015 •Eclipse Collections 7.0 •Jan. 2016 •Eclipse Collections 8.0 (Java 8) •Sept. 2016 •Google Guava 20.0 •Oct. 2016 •Javaslang 2.1a •Nov. 2016 March 2017 •Google Guava 21.0 (Java 8) •Jan. 2017 •Eclipse Collections 8.1 •Mar. 2017
  • 12. #CollectionsCompare #GIDS17 @NikhilNanivade Use Case – Deck of Cards The Collections Compare Project: https://github.com/nikhilnanivadekar/CollectionsCompare
  • 13. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 14. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Benchmarks • JMH – Java Microbenchmark Harness • http://openjdk.java.net/projects/code-tools/jmh/ • Measure Reported – Operations per second • Bigger numbers are better • 4 Core Intel I7, 50 Warm-up iterations, 30 measurement iterations, two forks
  • 15. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 16. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deck of Cards 102,037 79,343 76,982 56,224 91,017 0 20,000 40,000 60,000 80,000 100,000 120,000 Scoreops/s Framework Deck of Cards Apache EC Guava Javaslang JDK
  • 17. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 18. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 19. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 20. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 21. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 22. #CollectionsCompare #GIDS17 @NikhilNanivade A Deck of Cards – Only the Types public class JDK8DeckOfCards { private SortedSet<Card> cards; private Map<Suit, SortedSet<Card>> cardsBySuit; public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; private MultiValuedMap<Suit, Card> cardsBySuit; public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; private ImmutableSetMultimap<Suit, Card> cardsBySuit; public class JavaSlangDeckOfCards { private SortedSet<Card> cards; private Map<Suit, ? extends SortedSet<Card>> cardsBySuit;
  • 23. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 24. #CollectionsCompare #GIDS17 @NikhilNanivade Shared Code – Cartesian Product public static Stream<Card> streamCards() { return Card.cartesianProduct( EnumSet.allOf(Rank.class), EnumSet.allOf(Suit.class), Card::new); } private static <A, B, C> Stream<C> cartesianProduct( Set<A> set1, Set<B> set2, Function2<A, B, C> function) { return set1.stream().flatMap(first -> set2.stream().map(second -> function.apply(first, second))); }
  • 25. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Cartesian Product 185,087 203,660 465,999 127,439 114,634 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 500,000 Scoreops/s Framework Cartesian Product Apache EC Guava Javaslang JDK
  • 26. #CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Cartesian Product public class JDK8DeckOfCards { private SortedSet<Card> cards; public JDK8DeckOfCards() { this.cards = Card.streamCards() .collect(Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet)); }
  • 27. #CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Cartesian Product public class ApacheCommonsDeckOfCards { private SortedSet<Card> cards; public ApacheCommonsDeckOfCards() { this.cards = Card.streamCards() .collect(Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet)); }
  • 28. #CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Cartesian Product public class EclipseCollectionsDeckOfCards { private ImmutableSortedSet<Card> cards; public EclipseCollectionsDeckOfCards() { this.cards = SortedSets.immutable.with( Card.streamCards() .toArray(Card[]::new)); }
  • 29. #CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Cartesian Product public class GoogleGuavaDeckOfCards { private ImmutableSortedSet<Card> cards; public GoogleGuavaDeckOfCards() { this.cards = Card.streamCards() .collect(ImmutableSortedSet .toImmutableSortedSet( Comparator.naturalOrder())); }
  • 30. #CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Cartesian Product public class JavaSlangDeckOfCards { private SortedSet<Card> cards; public JavaSlangDeckOfCards() { this.cards = Card.streamCards() .collect(TreeSet.collector()); }
  • 31. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 32. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - groupBy 540,396 183,848 231,321 176,412 353,651 - 100,000 200,000 300,000 400,000 500,000 600,000 Scoreops/s Framework groupBy Apache EC Guava Javaslang JDK
  • 33. #CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Group By public class JDK8DeckOfCards { private Map<Suit, SortedSet<Card>> cardsBySuit; public JDK8DeckOfCards() { this.cardsBySuit = Collections.unmodifiableMap( this.cards.stream() .collect(Collectors.groupingBy( Card::getSuit, Collectors.mapping( Function.identity(), Collectors.collectingAndThen( Collectors.toCollection(TreeSet::new), Collections::unmodifiableSortedSet))))); }
  • 34. #CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Group By public class ApacheCommonsDeckOfCards { private MultiValuedMap<Suit, Card> cardsBySuit; public ApacheCommonsDeckOfCards() { SetValuedMap<Suit, Card> cbs = MultiMapUtils.newSetValuedHashMap(); this.cards.forEach(card -> cbs.put(card.getSuit(), card)); this.cardsBySuit = MultiMapUtils.unmodifiableMultiValuedMap(cbs); }
  • 35. #CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Group By public class EclipseCollectionsDeckOfCards { private ImmutableSortedSetMultimap<Suit, Card> cardsBySuit; public EclipseCollectionsDeckOfCards() { this.cardsBySuit = this.cards.groupBy(Card::getSuit); }
  • 36. #CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Group By public class GoogleGuavaDeckOfCards { private ImmutableSetMultimap<Suit, Card> cardsBySuit; public GoogleGuavaDeckOfCards() { ImmutableSetMultimap.Builder<Suit, Card> builder = new ImmutableSetMultimap.Builder<Suit, Card>() .orderValuesBy(Comparator.naturalOrder()); this.cards.forEach(card -> builder.put(card.getSuit(), card)); this.cardsBySuit = builder.build(); }
  • 37. #CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Group By public class JavaSlangDeckOfCards { private Map<Suit, ? extends SortedSet<Card>> cardsBySuit; public JavaSlangDeckOfCards() { this.cardsBySuit = this.cards.groupBy(Card::getSuit); }
  • 38. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deck of Cards 102,037 79,343 76,982 56,224 91,017 0 20,000 40,000 60,000 80,000 100,000 120,000 Scoreops/s Framework Deck of Cards Apache EC Guava Javaslang JDK 540,396 183,848231,321176,412 353,651 - 500,000 1,000,000 Scoreops/s Framework groupBy Apache EC Guava Javaslang JDK 185,087 203,660 465,999 127,439 114,634 0 200,000 400,000 600,000 Scoreops/s Framework Cartesian Product Apache EC Guava Javaslang JDK
  • 39. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 40. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - countsBySuit 706,272 1,096,527 533,728 440,786 724,526 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
  • 41. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test - countsByRank 539,252 827,986 444,832 390,134 591,869 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 800,000 900,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
  • 42. #CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Count By public Map<Suit, Long> countsBySuit() { return this.cards.stream() .collect(Collectors.groupingBy( Card::getSuit, Collectors.counting())); } public Map<Rank, Long> countsByRank() { return this.cards.stream() .collect(Collectors.groupingBy( Card::getRank, Collectors.counting())); }
  • 43. #CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Count By public Bag<Suit> countsBySuit() { return this.cards.stream() .map(Card::getSuit) .collect(Collectors.toCollection(HashBag::new)); } public MultiSet<Rank> countsByRank() { return this.cards.stream() .map(Card::getRank) .collect(Collectors.toCollection( HashMultiSet::new)); }
  • 44. #CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Count By public Bag<Suit> countsBySuit() { return this.cards.asLazy() .collect(Card::getSuit) .toBag(); } public Bag<Rank> countsByRank() { return this.cards.asLazy() .collect(Card::getRank) .toBag(); }
  • 45. #CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Count By public Multiset<Suit> countsBySuit() { return this.cards.stream() .map(Card::getSuit) .collect(Collectors.toCollection( HashMultiset::create)); } public Multiset<Rank> countsByRank() { return this.cards.stream() .map(Card::getRank) .collect(Collectors.toCollection( HashMultiset::create)); }
  • 46. #CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Count By public java.util.Map<Suit, Long> countsBySuit() { return this.cards.collect( Collectors.groupingBy( Card::getSuit, Collectors.counting())); } public java.util.Map<Rank, Long> countsByRank() { return this.cards.collect( Collectors.groupingBy( Card::getRank, Collectors.counting())); }
  • 47. #CollectionsCompare #GIDS17 @NikhilNanivade Problem Statement – Deck of Cards 1. Create Deck of Cards – Store Cards in an “Immutable” SortedSet (Cartesian product of Suit x Rank) – Group the cards by Suit in an “Immutable” SortedSet “Multimap” (Group By) 2. Get the number of cards – Count By Suit returning “Multiset” or “Bag” – Count By Rank returning “Multiset” or “Bag” 3. Deal five hands of five cards each – Return the cards as an “Immutable” List of five Sets of five cards
  • 48. #CollectionsCompare #GIDS17 @NikhilNanivade Performance Test – Deal Hands 291,488 502,448 305,673 372,466 300,975 0 100,000 200,000 300,000 400,000 500,000 600,000 Scoreops/s Framework Apache EC Guava Javaslang JDK
  • 49. #CollectionsCompare #GIDS17 @NikhilNanivade JDK Collections – Deal Hands public List<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect( Collectors.collectingAndThen( Collectors.toList(), Collections::unmodifiableList)); }
  • 50. #CollectionsCompare #GIDS17 @NikhilNanivade Apache Collections – Deal Hands public List<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect( Collectors.collectingAndThen( Collectors.toList(), Collections::unmodifiableList)); }
  • 51. #CollectionsCompare #GIDS17 @NikhilNanivade Eclipse Collections – Deal Hands public ImmutableList<Set<Card>> dealHands( MutableStack<Card> shuffled, int hands, int cardsPerHand) { return IntInterval.oneTo(hands) .collect(i -> this.deal(shuffled, cardsPerHand)); }
  • 52. #CollectionsCompare #GIDS17 @NikhilNanivade Google Guava – Deal Hands public ImmutableList<Set<Card>> dealHands( Deque<Card> shuffled, int hands, int cardsPerHand) { return IntStream.range(0, hands) .mapToObj(i -> this.deal(shuffled, cardsPerHand)) .collect(ImmutableList.toImmutableList()); }
  • 53. #CollectionsCompare #GIDS17 @NikhilNanivade Javaslang – Deal Hands public List<Set<Card>> dealHands( Stack<Card> shuffled, int hands, int cardsPerHand) { List<Set<Card>> list = List.empty(); for (int i = 0; i < hands; i++) { Tuple2<Set<Card>, ? extends Stack<Card>> tuple2 = this.deal(shuffled, cardsPerHand); shuffled = tuple2._2(); list = list.append(tuple2._1()); } return list; }
  • 55. #CollectionsCompare #GIDS17 @NikhilNanivade Memory Benchmarks • ObjectExplorer, by Dimitris Andrew • Open Source under Apache License 2.0 • Measure Reported – – Memory (Kb) vs Size – Memory (Kb) vs List, Set, Map with 1,000,000 elements • Smaller the better Sources – GitHub: https://github.com/DimitrisAndreou/memory-measurer – Found from StackOverflow: http://stackoverflow.com/a/9368930/5182052
  • 69. #CollectionsCompare #GIDS17 @NikhilNanivade Link • The Collections Compare Project https://github.com/nikhilnanivadekar/CollectionsCompare
  • 70. #CollectionsCompare #GIDS17 @NikhilNanivade Collection Framework Comparison Feature JDK Apache Guava Eclipse Javaslang 1st, 2nd, 3rd List, Set, Map ✔  ✖ ✔  JDK, EC, JS Multiset / Bag ✖ ✔ ✔ ✔ ✖ GG, EC, AC Multimap ✖ ✔ ✔ ✔ ✔ GG, EC, AC BiMap ✖ ✔ ✔ ✔ ✖ GG, EC, AC Stack ✔  ✖ ✔ ✔ EC, JDK, JS Tree / Trie ✖ ✔ ✖ ✖ ✔ JS, AC Table ✖ ✖ ✔ ✖ ✖ GG Additional Types ✔ ✔ ✔   AC, JDK, GG Immutable? ✖ ✖ ✔ ✔ ✔ JS, EC, GG Primitives?  ✖  ✔ ✖ EC, JDK, GG Fluent API    ✔ ✔ EC, JS, JDK (E), (L), (S), (P)* , ✔, ✔, ✔ , , ✔, ✖ ✖, , ✔, ✖ ✔, ✔, ✔, ✔ ✔, ✔, ✔, ✖ EC, JDK, JS
  • 71. Thanks! The Collections Compare Project https://github.com/nikhilnanivadekar/CollectionsCompare