Exploring Implementation Possibilities - 1

• List < pair < KeyType , DataType > >
• use List<pair> as container
• Insert(), Remove(), Retrieve() use sequential search
• Insert(): search, then PushBack() if not found
• Insert(), Remove(), Retrieve() runtimes are O(size)
• Sorted vector < pair < KeyType , DataType > >
• Use Vector <pair> as container
• Maintain order by key (first element of pair)
• Use binary search for lookup
• Insert() and Remove() runtime O(size)
• Retrieve() runtime O(log size)
• Vector < DataType >
• assumes key is unsigned int
• random access: V[key] = data
• runspace = O(2^s), where s = size of largest key
• Set < KeyType , DataType >
• No special assumptions on KeyType
• All operations Θ(log n)
• runspace overhead reasonable
• ordered traversal included!