We Keep Coding

Counting each distinct array occurrence in a list of arrays with duplicates

PROBLEM
I have a list of arrays and I want to count the occurrences of duplicates.
For example, if I have this :
{{1,2,3},
{1,0,3},
{1,2,3},
{5,2,6},
{5,2,6},
{5,2,6}}
I want a map (or any relevant collection) like this :
{ {1,2,3} -> 2,
{1,0,3} -> 1,
{5,2,6} -> 3 }
I can even lose the arrays values, I'm only interested in cardinals (e.g. 2, 1 and 3 here).
MY SOLUTION
I use the following algorithm :
First hash the arrays, and check if each hash is in an HashMap<Integer, ArrayList<int[]>>, let's name it distinctHash, where the key is the hash and the value is an ArrayList, let's name it rowList, containing the different arrays for this hash (to avoid collisions).
If the hash is not in distinctHash, put it with the value 1 in another HashMap<int[], Long> that counts each occurrence, let's call it distinctElements.
Then if the hash is in distinctHash, check if the corresponding array is contained in rowList. If it is, increment the value in distinctElements associated to the identical array found in rowList. (If you use the new array as a key you will create another key since their reference are different).
Here is the code, the boolean returned tells if a new distinct array was found, I apply this function sequentially on all of my arrays :
HashMap<int[], Long> distinctElements;
HashMap<Integer, ArrayList<int[]>> distinctHash;
private boolean addRow(int[] row) {
if (distinctHash.containsKey(hash)) {
int[] indexRow = distinctHash.get(hash).get(0);
for (int[] previousRow: distinctHash.get(hash)) {
if (Arrays.equals(previousRow, row)) {
distinctElements.put(
indexRow,
distinctElements.get(indexRow) + 1
);
return false;
}
}
distinctElements.put(row, 1L);
ArrayList<int[]> rowList = distinctHash.get(hash);
rowList.add(row);
distinctHash.put(hash, rowList);
return true;
} else {
distinctElements.put(row, 1L);
ArrayList<int[]> newValue = new ArrayList<>();
newValue.add(row);
distinctHash.put(hash, newValue);
return true;
}
}
QUESTION
The problem is that my algorithm is too slow for my needs (40s for 5,000,000 arrays, and 2h-3h for 20,000,000 arrays). Profiling with NetBeans told me that the hashing takes 70% of runtime (using Google Guava murmur3_128 hash function).
Is there another algorithm that could be faster? As I said I'm not interested in arrays values, only in the number of their occurrences. I am ready to sacrifice precision for speed so a probabilistic algorithm is fine.

Wrap the int[] in a class that implements equals and hashCode, then build Map of the wrapper class to instance count.
class IntArray {
private int[] array;
public IntArray(int[] array) {
this.array = array;
}
#Override
public int hashCode() {
return Arrays.hashCode(this.array);
}
#Override
public boolean equals(Object obj) {
return (obj instanceof IntArray && Arrays.equals(this.array, ((IntArray) obj).array));
}
#Override
public String toString() {
return Arrays.toString(this.array);
}
}
Test
int[][] input = {{1,2,3},
{1,0,3},
{1,2,3},
{5,2,6},
{5,2,6},
{5,2,6}};
Map<IntArray, Long> map = Arrays.stream(input).map(IntArray::new)
.collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));
map.entrySet().forEach(System.out::println);
Output
[1, 2, 3]=2
[1, 0, 3]=1
[5, 2, 6]=3
Note: The above solution is faster and uses less memory than solution by Ravindra Ranwala, but it does require the creation of an extra class, so it is debatable which is better.
For smaller arrays, use the simpler solution below by Ravindra Ranwala.
For larger arrays, the above solution is likely better.
Map<List<Integer>, Long> map = Stream.of(input)
.map(a -> Arrays.stream(a).boxed().collect(Collectors.toList()))
.collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));

You may do it like so,
Map<List<Integer>, Long> result = Stream.of(source)
.map(a -> Arrays.stream(a).boxed().collect(Collectors.toList()))
.collect(Collectors.groupingBy(Function.identity(), Collectors.counting()));
And here's the output,
{[1, 2, 3]=2, [1, 0, 3]=1, [5, 2, 6]=3}

If the sequence of elements for all duplication of that array is like each other and the length of each array is not much, you can map each array to an int number and using from last part of your method. Although this method decrease the time of hashing, there are some assumptions here which might not be true for your case.

Java predicate - match against first predicate [duplicate]

I've just started playing with Java 8 lambdas and I'm trying to implement some of the things that I'm used to in functional languages.
For example, most functional languages have some kind of find function that operates on sequences, or lists that returns the first element, for which the predicate is true. The only way I can see to achieve this in Java 8 is:
lst.stream()
.filter(x -> x > 5)
.findFirst()
However this seems inefficient to me, as the filter will scan the whole list, at least to my understanding (which could be wrong). Is there a better way?

No, filter does not scan the whole stream. It's an intermediate operation, which returns a lazy stream (actually all intermediate operations return a lazy stream). To convince you, you can simply do the following test:
List<Integer> list = Arrays.asList(1, 10, 3, 7, 5);
int a = list.stream()
.peek(num -> System.out.println("will filter " + num))
.filter(x -> x > 5)
.findFirst()
.get();
System.out.println(a);
Which outputs:
will filter 1
will filter 10
10
You see that only the two first elements of the stream are actually processed.
So you can go with your approach which is perfectly fine.

However this seems inefficient to me, as the filter will scan the whole list
No it won't - it will "break" as soon as the first element satisfying the predicate is found. You can read more about laziness in the stream package javadoc, in particular (emphasis mine):
Many stream operations, such as filtering, mapping, or duplicate removal, can be implemented lazily, exposing opportunities for optimization. For example, "find the first String with three consecutive vowels" need not examine all the input strings. Stream operations are divided into intermediate (Stream-producing) operations and terminal (value- or side-effect-producing) operations. Intermediate operations are always lazy.

return dataSource.getParkingLots()
.stream()
.filter(parkingLot -> Objects.equals(parkingLot.getId(), id))
.findFirst()
.orElse(null);
I had to filter out only one object from a list of objects. So i used this, hope it helps.

In addition to Alexis C's answer, If you are working with an array list, in which you are not sure whether the element you are searching for exists, use this.
Integer a = list.stream()
.peek(num -> System.out.println("will filter " + num))
.filter(x -> x > 5)
.findFirst()
.orElse(null);
Then you could simply check whether a is null.

Already answered by #AjaxLeung, but in comments and hard to find.
For check only
lst.stream()
.filter(x -> x > 5)
.findFirst()
.isPresent()
is simplified to
lst.stream()
.anyMatch(x -> x > 5)

import org.junit.Test;
import java.util.Arrays;
import java.util.List;
import java.util.Optional;
// Stream is ~30 times slower for same operation...
public class StreamPerfTest {
int iterations = 100;
List<Integer> list = Arrays.asList(1, 10, 3, 7, 5);
// 55 ms
#Test
public void stream() {
for (int i = 0; i < iterations; i++) {
Optional<Integer> result = list.stream()
.filter(x -> x > 5)
.findFirst();
System.out.println(result.orElse(null));
}
}
// 2 ms
#Test
public void loop() {
for (int i = 0; i < iterations; i++) {
Integer result = null;
for (Integer walk : list) {
if (walk > 5) {
result = walk;
break;
}
}
System.out.println(result);
}
}
}

A generic utility function with looping seems a lot cleaner to me:
static public <T> T find(List<T> elements, Predicate<T> p) {
for (T item : elements) if (p.test(item)) return item;
return null;
}
static public <T> T find(T[] elements, Predicate<T> p) {
for (T item : elements) if (p.test(item)) return item;
return null;
}
In use:
List<Integer> intList = Arrays.asList(1, 2, 3, 4, 5);
Integer[] intArr = new Integer[]{1, 2, 3, 4, 5};
System.out.println(find(intList, i -> i % 2 == 0)); // 2
System.out.println(find(intArr, i -> i % 2 != 0)); // 1
System.out.println(find(intList, i -> i > 5)); // null

Improved One-Liner answer: If you are looking for a boolean return value, we can do it better by adding isPresent:
return dataSource.getParkingLots().stream().filter(parkingLot -> Objects.equals(parkingLot.getId(), id)).findFirst().isPresent();

Java 8 lambda get and remove element from list

Given a list of elements, I want to get the element with a given property and remove it from the list. The best solution I found is:
ProducerDTO p = producersProcedureActive
.stream()
.filter(producer -> producer.getPod().equals(pod))
.findFirst()
.get();
producersProcedureActive.remove(p);
Is it possible to combine get and remove in a lambda expression?

To Remove element from the list
objectA.removeIf(x -> conditions);
eg:
objectA.removeIf(x -> blockedWorkerIds.contains(x));
List<String> str1 = new ArrayList<String>();
str1.add("A");
str1.add("B");
str1.add("C");
str1.add("D");
List<String> str2 = new ArrayList<String>();
str2.add("D");
str2.add("E");
str1.removeIf(x -> str2.contains(x));
str1.forEach(System.out::println);
OUTPUT:
A
B
C

Although the thread is quite old, still thought to provide solution - using Java8.
Make the use of removeIf function. Time complexity is O(n)
producersProcedureActive.removeIf(producer -> producer.getPod().equals(pod));
API reference: removeIf docs
Assumption: producersProcedureActive is a List
NOTE: With this approach you won't be able to get the hold of the deleted item.

Consider using vanilla java iterators to perform the task:
public static <T> T findAndRemoveFirst(Iterable<? extends T> collection, Predicate<? super T> test) {
T value = null;
for (Iterator<? extends T> it = collection.iterator(); it.hasNext();)
if (test.test(value = it.next())) {
it.remove();
return value;
}
return null;
}
Advantages:
It is plain and obvious.
It traverses only once and only up to the matching element.
You can do it on any Iterable even without stream() support (at least those implementing remove() on their iterator).
Disadvantages:
You cannot do it in place as a single expression (auxiliary method or variable required)
As for the
Is it possible to combine get and remove in a lambda expression?
other answers clearly show that it is possible, but you should be aware of
Search and removal may traverse the list twice
ConcurrentModificationException may be thrown when removing element from the list being iterated

The direct solution would be to invoke ifPresent(consumer) on the Optional returned by findFirst(). This consumer will be invoked when the optional is not empty. The benefit also is that it won't throw an exception if the find operation returned an empty optional, like your current code would do; instead, nothing will happen.
If you want to return the removed value, you can map the Optional to the result of calling remove:
producersProcedureActive.stream()
.filter(producer -> producer.getPod().equals(pod))
.findFirst()
.map(p -> {
producersProcedureActive.remove(p);
return p;
});
But note that the remove(Object) operation will again traverse the list to find the element to remove. If you have a list with random access, like an ArrayList, it would be better to make a Stream over the indexes of the list and find the first index matching the predicate:
IntStream.range(0, producersProcedureActive.size())
.filter(i -> producersProcedureActive.get(i).getPod().equals(pod))
.boxed()
.findFirst()
.map(i -> producersProcedureActive.remove((int) i));
With this solution, the remove(int) operation operates directly on the index.

Use can use filter of Java 8, and create another list if you don't want to change the old list:
List<ProducerDTO> result = producersProcedureActive
.stream()
.filter(producer -> producer.getPod().equals(pod))
.collect(Collectors.toList());

I'm sure this will be an unpopular answer, but it works...
ProducerDTO[] p = new ProducerDTO[1];
producersProcedureActive
.stream()
.filter(producer -> producer.getPod().equals(pod))
.findFirst()
.ifPresent(producer -> {producersProcedureActive.remove(producer); p[0] = producer;}
p[0] will either hold the found element or be null.
The "trick" here is circumventing the "effectively final" problem by using an array reference that is effectively final, but setting its first element.

With Eclipse Collections you can use detectIndex along with remove(int) on any java.util.List.
List<Integer> integers = Lists.mutable.with(1, 2, 3, 4, 5);
int index = Iterate.detectIndex(integers, i -> i > 2);
if (index > -1) {
integers.remove(index);
}
Assert.assertEquals(Lists.mutable.with(1, 2, 4, 5), integers);
If you use the MutableList type from Eclipse Collections, you can call the detectIndex method directly on the list.
MutableList<Integer> integers = Lists.mutable.with(1, 2, 3, 4, 5);
int index = integers.detectIndex(i -> i > 2);
if (index > -1) {
integers.remove(index);
}
Assert.assertEquals(Lists.mutable.with(1, 2, 4, 5), integers);
Note: I am a committer for Eclipse Collections

The below logic is the solution without modifying the original list
List<String> str1 = new ArrayList<String>();
str1.add("A");
str1.add("B");
str1.add("C");
str1.add("D");
List<String> str2 = new ArrayList<String>();
str2.add("D");
str2.add("E");
List<String> str3 = str1.stream()
.filter(item -> !str2.contains(item))
.collect(Collectors.toList());
str1 // ["A", "B", "C", "D"]
str2 // ["D", "E"]
str3 // ["A", "B", "C"]

When we want to get multiple elements from a List into a new list (filter using a predicate) and remove them from the existing list, I could not find a proper answer anywhere.
Here is how we can do it using Java Streaming API partitioning.
Map<Boolean, List<ProducerDTO>> classifiedElements = producersProcedureActive
.stream()
.collect(Collectors.partitioningBy(producer -> producer.getPod().equals(pod)));
// get two new lists
List<ProducerDTO> matching = classifiedElements.get(true);
List<ProducerDTO> nonMatching = classifiedElements.get(false);
// OR get non-matching elements to the existing list
producersProcedureActive = classifiedElements.get(false);
This way you effectively remove the filtered elements from the original list and add them to a new list.
Refer the 5.2. Collectors.partitioningBy section of this article.

As others have suggested, this might be a use case for loops and iterables. In my opinion, this is the simplest approach. If you want to modify the list in-place, it cannot be considered "real" functional programming anyway. But you could use Collectors.partitioningBy() in order to get a new list with elements which satisfy your condition, and a new list of those which don't. Of course with this approach, if you have multiple elements satisfying the condition, all of those will be in that list and not only the first.

the task is: get ✶and✶ remove element from list
p.stream().collect( Collectors.collectingAndThen( Collector.of(
ArrayDeque::new,
(a, producer) -> {
if( producer.getPod().equals( pod ) )
a.addLast( producer );
},
(a1, a2) -> {
return( a1 );
},
rslt -> rslt.pollFirst()
),
(e) -> {
if( e != null )
p.remove( e ); // remove
return( e ); // get
} ) );

resumoRemessaPorInstrucoes.removeIf(item ->
item.getTipoOcorrenciaRegistro() == TipoOcorrenciaRegistroRemessa.PEDIDO_PROTESTO.getNome() ||
item.getTipoOcorrenciaRegistro() == TipoOcorrenciaRegistroRemessa.SUSTAR_PROTESTO_BAIXAR_TITULO.getNome());

Combining my initial idea and your answers I reached what seems to be the solution
to my own question:
public ProducerDTO findAndRemove(String pod) {
ProducerDTO p = null;
try {
p = IntStream.range(0, producersProcedureActive.size())
.filter(i -> producersProcedureActive.get(i).getPod().equals(pod))
.boxed()
.findFirst()
.map(i -> producersProcedureActive.remove((int)i))
.get();
logger.debug(p);
} catch (NoSuchElementException e) {
logger.error("No producer found with POD [" + pod + "]");
}
return p;
}
It lets remove the object using remove(int) that do not traverse again the
list (as suggested by #Tunaki) and it lets return the removed object to
the function caller.
I read your answers that suggest me to choose safe methods like ifPresent instead of get but I do not find a way to use them in this scenario.
Are there any important drawback in this kind of solution?
Edit following #Holger advice
This should be the function I needed
public ProducerDTO findAndRemove(String pod) {
return IntStream.range(0, producersProcedureActive.size())
.filter(i -> producersProcedureActive.get(i).getPod().equals(pod))
.boxed()
.findFirst()
.map(i -> producersProcedureActive.remove((int)i))
.orElseGet(() -> {
logger.error("No producer found with POD [" + pod + "]");
return null;
});
}

A variation of the above:
import static java.util.function.Predicate.not;
final Optional<MyItem> myItem = originalCollection.stream().filter(myPredicate(someInfo)).findFirst();
final List<MyItem> myOtherItems = originalCollection.stream().filter(not(myPredicate(someInfo))).toList();
private Predicate<MyItem> myPredicate(Object someInfo) {
return myItem -> myItem.someField() == someInfo;
}

How to use Java 8 streams to find all values preceding a larger value?

Use Case
Through some coding Katas posted at work, I stumbled on this problem that I'm not sure how to solve.
Using Java 8 Streams, given a list of positive integers, produce a
list of integers where the integer preceded a larger value.
[10, 1, 15, 30, 2, 6]
The above input would yield:
[1, 15, 2]
since 1 precedes 15, 15 precedes 30, and 2 precedes 6.
Non-Stream Solution
public List<Integer> findSmallPrecedingValues(final List<Integer> values) {
List<Integer> result = new ArrayList<Integer>();
for (int i = 0; i < values.size(); i++) {
Integer next = (i + 1 < values.size() ? values.get(i + 1) : -1);
Integer current = values.get(i);
if (current < next) {
result.push(current);
}
}
return result;
}
What I've Tried
The problem I have is I can't figure out how to access next in the lambda.
return values.stream().filter(v -> v < next).collect(Collectors.toList());
Question
Is it possible to retrieve the next value in a stream?
Should I be using map and mapping to a Pair in order to access next?

Using IntStream.range:
static List<Integer> findSmallPrecedingValues(List<Integer> values) {
return IntStream.range(0, values.size() - 1)
.filter(i -> values.get(i) < values.get(i + 1))
.mapToObj(values::get)
.collect(Collectors.toList());
}
It's certainly nicer than an imperative solution with a large loop, but still a bit meh as far as the goal of "using a stream" in an idiomatic way.
Is it possible to retrieve the next value in a stream?
Nope, not really. The best cite I know of for that is in the java.util.stream package description:
The elements of a stream are only visited once during the life of a stream. Like an Iterator, a new stream must be generated to revisit the same elements of the source.
(Retrieving elements besides the current element being operated on would imply they could be visited more than once.)
We could also technically do it in a couple other ways:
Statefully (very meh).
Using a stream's iterator is technically still using the stream.

That's not a pure Java8, but recently I've published a small library called StreamEx which has a method exactly for this task:
// Find all numbers where the integer preceded a larger value.
Collection<Integer> numbers = Arrays.asList(10, 1, 15, 30, 2, 6);
List<Integer> res = StreamEx.of(numbers).pairMap((a, b) -> a < b ? a : null)
.nonNull().toList();
assertEquals(Arrays.asList(1, 15, 2), res);
The pairMap operation internally implemented using custom spliterator. As a result you have quite clean code which does not depend on whether the source is List or anything else. Of course it works fine with parallel stream as well.
Committed a testcase for this task.

It's not a one-liner (it's a two-liner), but this works:
List<Integer> result = new ArrayList<>();
values.stream().reduce((a,b) -> {if (a < b) result.add(a); return b;});
Rather than solving it by "looking at the next element", this solves it by "looking at the previous element, which reduce() give you for free. I have bent its intended usage by injecting a code fragment that populates the list based on the comparison of previous and current elements, then returns the current so the next iteration will see it as its previous element.
Some test code:
List<Integer> result = new ArrayList<>();
IntStream.of(10, 1, 15, 30, 2, 6).reduce((a,b) -> {if (a < b) result.add(a); return b;});
System.out.println(result);
Output:
[1, 15, 2]

The accepted answer works fine if either the stream is sequential or parallel but can suffer if the underlying List is not random access, due to multiple calls to get.
If your stream is sequential, you might roll this collector:
public static Collector<Integer, ?, List<Integer>> collectPrecedingValues() {
int[] holder = {Integer.MAX_VALUE};
return Collector.of(ArrayList::new,
(l, elem) -> {
if (holder[0] < elem) l.add(holder[0]);
holder[0] = elem;
},
(l1, l2) -> {
throw new UnsupportedOperationException("Don't run in parallel");
});
}
and a usage:
List<Integer> precedingValues = list.stream().collect(collectPrecedingValues());
Nevertheless you could also implement a collector so thats works for sequential and parallel streams. The only thing is that you need to apply a final transformation, but here you have control over the List implementation so you won't suffer from the get performance.
The idea is to generate first a list of pairs (represented by a int[] array of size 2) which contains the values in the stream sliced by a window of size two with a gap of one. When we need to merge two lists, we check the emptiness and merge the gap of the last element of the first list with the first element of the second list. Then we apply a final transformation to filter only desired values and map them to have the desired output.
It might not be as simple as the accepted answer, but well it can be an alternative solution.
public static Collector<Integer, ?, List<Integer>> collectPrecedingValues() {
return Collectors.collectingAndThen(
Collector.of(() -> new ArrayList<int[]>(),
(l, elem) -> {
if (l.isEmpty()) l.add(new int[]{Integer.MAX_VALUE, elem});
else l.add(new int[]{l.get(l.size() - 1)[1], elem});
},
(l1, l2) -> {
if (l1.isEmpty()) return l2;
if (l2.isEmpty()) return l1;
l2.get(0)[0] = l1.get(l1.size() - 1)[1];
l1.addAll(l2);
return l1;
}), l -> l.stream().filter(arr -> arr[0] < arr[1]).map(arr -> arr[0]).collect(Collectors.toList()));
}
You can then wrap these two collectors in a utility collector method, check if the stream is parallel with isParallel an then decide which collector to return.

If you're willing to use a third party library and don't need parallelism, then jOOλ offers SQL-style window functions as follows
System.out.println(
Seq.of(10, 1, 15, 30, 2, 6)
.window()
.filter(w -> w.lead().isPresent() && w.value() < w.lead().get())
.map(w -> w.value())
.toList()
);
Yielding
[1, 15, 2]
The lead() function accesses the next value in traversal order from the window.
Disclaimer: I work for the company behind jOOλ

You can achieve that by using a bounded queue to store elements which flows through the stream (which is basing on the idea which I described in detail here: Is it possible to get next element in the Stream?
Belows example first defines instance of BoundedQueue class which will store elements going through the stream (if you don't like idea of extending the LinkedList, refer to link mentioned above for alternative and more generic approach). Later you just examine the two subsequent elements - thanks to the helper class:
public class Kata {
public static void main(String[] args) {
List<Integer> input = new ArrayList<Integer>(asList(10, 1, 15, 30, 2, 6));
class BoundedQueue<T> extends LinkedList<T> {
public BoundedQueue<T> save(T curElem) {
if (size() == 2) { // we need to know only two subsequent elements
pollLast(); // remove last to keep only requested number of elements
}
offerFirst(curElem);
return this;
}
public T getPrevious() {
return (size() < 2) ? null : getLast();
}
public T getCurrent() {
return (size() == 0) ? null : getFirst();
}
}
BoundedQueue<Integer> streamHistory = new BoundedQueue<Integer>();
final List<Integer> answer = input.stream()
.map(i -> streamHistory.save(i))
.filter(e -> e.getPrevious() != null)
.filter(e -> e.getCurrent() > e.getPrevious())
.map(e -> e.getPrevious())
.collect(Collectors.toList());
answer.forEach(System.out::println);
}
}

How to force max to return ALL maximum values in a Java Stream?

I've tested a bit the max function on Java 8 lambdas and streams, and it seems that in case max is executed, even if more than one object compares to 0, it returns an arbitrary element within the tied candidates without further consideration.
Is there an evident trick or function for such a max expected behavior, so that all max values are returned? I don't see anything in the API but I am sure it must exist something better than comparing manually.
For instance:
// myComparator is an IntegerComparator
Stream.of(1, 3, 5, 3, 2, 3, 5)
.max(myComparator)
.forEach(System.out::println);
// Would print 5, 5 in any order.

I believe the OP is using a Comparator to partition the input into equivalence classes, and the desired result is a list of members of the equivalence class that is the maximum according to that Comparator.
Unfortunately, using int values as a sample problem is a terrible example. All equal int values are fungible, so there is no notion of preserving the ordering of equivalent values. Perhaps a better example is using string lengths, where the desired result is to return a list of strings from an input that all have the longest length within that input.
I don't know of any way to do this without storing at least partial results in a collection.
Given an input collection, say
List<String> list = ... ;
...it's simple enough to do this in two passes, the first to get the longest length, and the second to filter the strings that have that length:
int longest = list.stream()
.mapToInt(String::length)
.max()
.orElse(-1);
List<String> result = list.stream()
.filter(s -> s.length() == longest)
.collect(toList());
If the input is a stream, which cannot be traversed more than once, it is possible to compute the result in only a single pass using a collector. Writing such a collector isn't difficult, but it is a bit tedious as there are several cases to be handled. A helper function that generates such a collector, given a Comparator, is as follows:
static <T> Collector<T,?,List<T>> maxList(Comparator<? super T> comp) {
return Collector.of(
ArrayList::new,
(list, t) -> {
int c;
if (list.isEmpty() || (c = comp.compare(t, list.get(0))) == 0) {
list.add(t);
} else if (c > 0) {
list.clear();
list.add(t);
}
},
(list1, list2) -> {
if (list1.isEmpty()) {
return list2;
}
if (list2.isEmpty()) {
return list1;
}
int r = comp.compare(list1.get(0), list2.get(0));
if (r < 0) {
return list2;
} else if (r > 0) {
return list1;
} else {
list1.addAll(list2);
return list1;
}
});
}
This stores intermediate results in an ArrayList. The invariant is that all elements within any such list are equivalent in terms of the Comparator. When adding an element, if it's less than the elements in the list, it's ignored; if it's equal, it's added; and if it's greater, the list is emptied and the new element is added. Merging isn't too difficult either: the list with the greater elements is returned, but if their elements are equal the lists are appended.
Given an input stream, this is pretty easy to use:
Stream<String> input = ... ;
List<String> result = input.collect(maxList(comparing(String::length)));

I would group by value and store the values into a TreeMap in order to have my values sorted, then I would get the max value by getting the last entry as next:
Stream.of(1, 3, 5, 3, 2, 3, 5)
.collect(groupingBy(Function.identity(), TreeMap::new, toList()))
.lastEntry()
.getValue()
.forEach(System.out::println);
Output:
5
5

I implemented more generic collector solution with custom downstream collector. Probably some readers might find it useful:
public static <T, A, D> Collector<T, ?, D> maxAll(Comparator<? super T> comparator,
Collector<? super T, A, D> downstream) {
Supplier<A> downstreamSupplier = downstream.supplier();
BiConsumer<A, ? super T> downstreamAccumulator = downstream.accumulator();
BinaryOperator<A> downstreamCombiner = downstream.combiner();
class Container {
A acc;
T obj;
boolean hasAny;
Container(A acc) {
this.acc = acc;
}
}
Supplier<Container> supplier = () -> new Container(downstreamSupplier.get());
BiConsumer<Container, T> accumulator = (acc, t) -> {
if(!acc.hasAny) {
downstreamAccumulator.accept(acc.acc, t);
acc.obj = t;
acc.hasAny = true;
} else {
int cmp = comparator.compare(t, acc.obj);
if (cmp > 0) {
acc.acc = downstreamSupplier.get();
acc.obj = t;
}
if (cmp >= 0)
downstreamAccumulator.accept(acc.acc, t);
}
};
BinaryOperator<Container> combiner = (acc1, acc2) -> {
if (!acc2.hasAny) {
return acc1;
}
if (!acc1.hasAny) {
return acc2;
}
int cmp = comparator.compare(acc1.obj, acc2.obj);
if (cmp > 0) {
return acc1;
}
if (cmp < 0) {
return acc2;
}
acc1.acc = downstreamCombiner.apply(acc1.acc, acc2.acc);
return acc1;
};
Function<Container, D> finisher = acc -> downstream.finisher().apply(acc.acc);
return Collector.of(supplier, accumulator, combiner, finisher);
}
So by default it can be collected to a list using:
public static <T> Collector<T, ?, List<T>> maxAll(Comparator<? super T> comparator) {
return maxAll(comparator, Collectors.toList());
}
But you can use other downstream collectors as well:
public static String joinLongestStrings(Collection<String> input) {
return input.stream().collect(
maxAll(Comparator.comparingInt(String::length), Collectors.joining(","))));
}

If I understood well, you want the frequency of the max value in the Stream.
One way to achieve that would be to store the results in a TreeMap<Integer, List<Integer> when you collect elements from the Stream. Then you grab the last key (or first depending on the comparator you give) to get the value which will contains the list of max values.
List<Integer> maxValues = st.collect(toMap(i -> i,
Arrays::asList,
(l1, l2) -> Stream.concat(l1.stream(), l2.stream()).collect(toList()),
TreeMap::new))
.lastEntry()
.getValue();
Collecting it from the Stream(4, 5, -2, 5, 5) will give you a List [5, 5, 5].
Another approach in the same spirit would be to use a group by operation combined with the counting() collector:
Entry<Integer, Long> maxValues = st.collect(groupingBy(i -> i,
TreeMap::new,
counting())).lastEntry(); //5=3 -> 5 appears 3 times
Basically you firstly get a Map<Integer, List<Integer>>. Then the downstream counting() collector will return the number of elements in each list mapped by its key resulting in a Map. From there you grab the max entry.
The first approaches require to store all the elements from the stream. The second one is better (see Holger's comment) as the intermediate List is not built. In both approached, the result is computed in a single pass.
If you get the source from a collection, you may want to use Collections.max one time to find the maximum value followed by Collections.frequency to find how many times this value appears.
It requires two passes but uses less memory as you don't have to build the data-structure.
The stream equivalent would be coll.stream().max(...).get(...) followed by coll.stream().filter(...).count().

I'm not really sure whether you are trying to
(a) find the number of occurrences of the maximum item, or
(b) Find all the maximum values in the case of a Comparator that is not consistent with equals.
An example of (a) would be [1, 5, 4, 5, 1, 1] -> [5, 5].
An example of (b) would be:
Stream.of("Bar", "FOO", "foo", "BAR", "Foo")
.max((s, t) -> s.toLowerCase().compareTo(t.toLowerCase()));
which you want to give [Foo, foo, Foo], rather than just FOO or Optional[FOO].
In both cases, there are clever ways to do it in just one pass. But these approaches are of dubious value because you would need to keep track of unnecessary information along the way. For example, if you start with [2, 0, 2, 2, 1, 6, 2], it would only be when you reach 6 that you would realise it was not necessary to track all the 2s.
I think the best approach is the obvious one; use max, and then iterate the items again putting all the ties into a collection of your choice. This will work for both (a) and (b).

If you'd rather rely on a library than the other answers here, StreamEx has a collector to do this.
Stream.of(1, 3, 5, 3, 2, 3, 5)
.collect(MoreCollectors.maxAll())
.forEach(System.out::println);
There's a version which takes a Comparator too for streams of items which don't have a natural ordering (i.e. don't implement Comparable).

System.out.println(
Stream.of(1, 3, 5, 3, 2, 3, 5)
.map(a->new Integer[]{a})
.reduce((a,b)->
a[0]==b[0]?
Stream.concat(Stream.of(a),Stream.of(b)).toArray() :
a[0]>b[0]? a:b
).get()
)

I was searching for a good answer on this question, but a tad more complex and couldn't find anything until I figured it out myself, which is why I'm posting if this helps anybody.
I have a list of Kittens.
Kitten is an object which has a name, age and gender. I had to return a list of all the youngest kittens.
For example:
So kitten list would contain kitten objects (k1, k2, k3, k4) and their ages would be (1, 2, 3, 1) accordingly. We want to return [k1, k4], because they are both the youngest. If only one youngest exists, the function should return [k1(youngest)].
Find the min value of the list (if it exists):
Optional<Kitten> minKitten = kittens.stream().min(Comparator.comparingInt(Kitten::getAge));
filter the list by the min value
return minKitten.map(value -> kittens.stream().filter(kitten -> kitten.getAge() == value.getAge())
.collect(Collectors.toList())).orElse(Collections.emptyList());

The following two lines will do it without implementing a separate comparator:
List<Integer> list = List.of(1, 3, 5, 3, 2, 3, 5);
list.stream().filter(i -> i == (list.stream().max(Comparator.comparingInt(i2 -> i2))).get()).forEach(System.out::println);