Boost - unordered_set tutorial/examples/ANYTHING?

Question

I'd like to use unordered_set in a project.

However, documentation for it is either incomplete or just a technical reference, no examples.

Can anyone provide links to online resources which deal with it? Books also welcome, preferably free. Google searching returned nothing of value.

Thanks!

Solution 1

There's little docs on it because it behaves exactly like std::set, with the exception that it requires a hashing and equals function instead of a comparison function. Just look up examples for std::set, and replace them with std::unordered_set and you should be fine.

If you need to write a hashing function, there are examples in the docs, i.e. this one.

Solution 2

Code for the most common use case:

#include <boost/unordered_set.hpp>
using boost::unordered_set;
using std::string;
using std::cout;
using std::endl;

int main (void)
{   
    // Initialize set
    unordered_set<string> s;
    s.insert("red");
    s.insert("green");
    s.insert("blue");

    // Search for membership
    if(s.find("red") != s.end())
        cout << "found red" << endl;
    if(s.find("purple") != s.end())
        cout << "found purple" << endl;
    if(s.find("blue") != s.end())
        cout << "found blue" << endl;

    return 0;
}

Output

found red
found blue

More Information

http://www.cplusplus.com/reference/unordered_set/unordered_set/find/

Solution 3

The boost containers are effectively an implementation of the interface first specified by the C++ Standard Library Technical Report (known as TR1), as mentioned in the boost docs. They seem to be part of the new standards working draft by now. Google turns up some more documentation/examples if you search for tr1 and unordered_set. I like the MSDN reference, which also has some samples:

http://msdn.microsoft.com/en-us/library/bb982739.aspx

http://www.google.de/search?q=tr1+unordered_set

Solution 4

I would try using the same methods of access that you use on std::set or other containers, http://www.boost.org/doc/libs/1_37_0/doc/html/unordered.html seems to agree.

View more solutions

Comments

[John Zwinck]

It may have an interface very similar to std::set, but its behavior is different. Aside from the ways in which hashing makes it faster, there are also ways in which Boost's unordered_set can be slower, such as: svn.boost.org/trac/boost/ticket/3693

[Billy ONeal]

@John: It's performance characteristics are different, and it cannot be iterated in an ordered manner (well, it is called unordered_set). Otherwise it behaves exactly as std::set does.

[Sid]

The erase performance issue has since been fixed: Quote from boost 1.61_0: Notes: In older versions this could be inefficient because it had to search through several buckets to find the position of the returned iterator. The data structure has been changed so that this is no longer the case, and the alternative erase methods have been deprecated.