Elegant way of counting occurrences in a java collection
Given a collection of objects with possible duplicates, I'd like end up with a count of occurrences per object. I do it by initializing an empty Map
, then iterating through the Collection
and mapping the object to its count (incrementing the count each time the map already contains the object).
public Map<Object, Integer> countOccurrences(Collection<Object> list) {
Map<Object, Integer> occurrenceMap = new HashMap<Object, Integer>();
for (Object obj : list) {
Integer numOccurrence = occurrenceMap.get(obj);
if (numOccurrence == null) {
//first count
occurrenceMap.put(obj, 1);
} else {
occurrenceMap.put(obj, numOccurrence++);
}
}
return occurrenceMap;
}
This looks too verbose for a simple logic of counting occurrences. Is there a more elegant/shorter way of doing this? I'm open to a completely different algorithm or a java language specific feature that allows for a shorter code.
Solution 1:
Check out Guava's Multiset. Pretty much exactly what you're looking for.
Unfortunately it doesn't have an addAll(Iterable iterable) function, but a simple loop over your collection calling add(E e) is easy enough.
EDIT
My mistake, it does indeed have an addAll method - as it must, since it implements Collection.
Solution 2:
Now let's try some Java 8 code:
static public Map<String, Integer> toMap(List<String> lst) {
return lst.stream()
.collect(HashMap<String, Integer>::new,
(map, str) -> {
if (!map.containsKey(str)) {
map.put(str, 1);
} else {
map.put(str, map.get(str) + 1);
}
},
HashMap<String, Integer>::putAll);
}
static public Map<String, Integer> toMap(List<String> lst) {
return lst.stream().collect(Collectors.groupingBy(s -> s,
Collectors.counting()));
}
I think this code is more elegant.