Top 5 most frequent from an array c++

Question

Hi i've an array and im looking to get the top 5 most frequently occuring from this array.

static std::string pickRandomStockSymbol()
{
    static std::string stockSymbols[] = {"SIRI", "INTC", "ZNGA", "BBRY", "MSFT", 
        "QQQ", "CSCO", "FB", "MU", "DELL", "AMAT", "NWSA", "AAPL", "AFFY", "ORCL", 
        "YHOO", "GRPN", "MDLZ", "VOD", "CMCSA" };

    return stockSymbols[rand() % 20];

^^ this is the array i will be using.

the transactions are randomly created using this struct:

struct Transaction
{
string stockSymbol;     // String containing the stock symbol, e.g. "AAPL"
string buyerName;       // String containing the buyer's name e.g. "Mr Brown"
int buyerAccount;       // Integer containing an eight digit account code
int numShares;          // Integer containing the number of sold shares
int pricePerShare;      // Integer containing the buy price per share
};

it is within this function i plan to do this in, i just dont really know what way i approach this:

string* Analyser::topFiveStocks()
{

return new string[5];
}

is there anyone out there willing to show me how i could run through the transactions to get these top 5 occuring elements?

if there would be any more information needed i'll be more than happy to provide.

Thanks in advance, Andrew

Answer 1

You could use a std::unordered_map with the stock symbol as the key, and the transaction count as the value. Then just put the five highest in a std::vector and return that.

As for putting the top N in the vector, you could keep it sorted, and re-sort it after every insert so that the stock with the highest transaction count is first. Then it's easy to see if the current stock when iterating over the map has a higher transaction count than the last item in the vector (which is the item in the vector with the smallest transaction count), then add it to the vector and re-sort it.

---

You could also just add all stocks from the map into a vector, and then sort it using the value in the map, and get the first five entries in the vector.

This can be something like this:

using transaction_map_type = std::unordered_map<std::string, unsigned int>;

transaction_map_type transactions;

// ...

std::vector<std::string> topFiveStocks()
{

    std::vector<transaction_map_type::value_type> all_trans;

    // Copy all transaction into our vector
    std::copy(std::begin(transactions), std::end(transactions),
              std::back_inserter(all_trans));

    // Now sort the transactions
    std::sort(std::begin(all_trans), std::end(all_trans),
              [](const transaction_map_type::value_type& t1,
                 const transaction_map_type::value_type& t2)
              { return t1.second > t2.second; });

    // And get the top five (or less) results into a separate vector
    std::vector<std::string> top_five;

    auto count = std::min(5UL, all_trans.size());
    for (unsigned i = 0; i < count; i++)
        top_five.push_back(all_trans[i].first);

    return top_five;
}

Also, remember to increase the counter for the transactions in the map whenever you do a transaction.

Note: This solution not tested, just written in the browser. May not even compile.

Answer 2

只需对数组进行排序，然后在其上循环以计算相等元素的最长间隔。

Answer 3

Accumulate the stock symbols:

• the counts into a map<string, int>
• the highest 5 symbols into a set<string>
• the lowest frequency of the highest 5 symbols into an int
• the lowest of the highest 5 symbols into a string