[C++] Using std::max_element on a vector of pairs

[Shoob]

I have a vector of pairs that contains an object from another class and an integer. 

vector<pair <Product, int>> cProducts;

One of the methods I have to make is to find the most expensive product that the client has bought. My initial solution was to not use std::max_element and simply iterate through the vector storing the object with the highest price.

Product Client::GetMostExpensiveProduct()
{
	double max = 0.0f;
	Product mostExpensive;
	for (auto& it : cProducts)
	{
		if (it.first.GetPrice() > max)
		{
			mostExpensive = it.first;
		}
	}
	return mostExpensive;
}

However, it's suggested to use methods from <algorithm> for this assignment. Is there an easy way to use std::max_element on this container (possibly using a lambda expression?) or should I switch to a completely different one for the task?

[as96]

56 minutes ago, Pandalf said:

I have a vector of pairs that contains an object from another class and an integer. 

vector<pair <Product, int>> cProducts;

One of the methods I have to make is to find the most expensive product that the client has bought. My initial solution was to not use std::max_element and simply iterate through the vector storing the object with the highest price.

Product Client::GetMostExpensiveProduct()
{
	double max = 0.0f;
	Product mostExpensive;
	for (auto& it : cProducts)
	{
		if (it.first.GetPrice() > max)
		{
			mostExpensive = it.first;
		}
	}
	return mostExpensive;
}

However, it's suggested to use methods from <algorithm> for this assignment. Is there an easy way to use std::max_element on this container (possibly using a lambda expression?) or should I switch to a completely different one for the task?

The code you wrote doesn't look wrong at all, and should work perfectly and as far as performance goes I don't see any major issue, however if you HAVE to use the algorithm library for whatever reason you can use the sort function:

http://www.cplusplus.com/reference/algorithm/sort/

and write a function like this to pass to the sort method:

bool myfunction (int prod1, int prod2) 
{ 
	return (prod1.GetPrice() < prod2.GetPrice()); 
}

Then take the last item in the vector.

 

PS: max_element would work in the same way.

[Shoob]

1 hour ago, as96 said:

The code you wrote doesn't look wrong at all, and should work perfectly and as far as performance goes I don't see any major issue, however if you HAVE to use the algorithm library for whatever reason you can use the sort function:

http://www.cplusplus.com/reference/algorithm/sort/

and write a function like this to pass to the sort method:

bool myfunction (int prod1, int prod2) 
{ 
	return (prod1.GetPrice() < prod2.GetPrice()); 
}

Then take the last item in the vector.

 

PS: max_element would work in the same way.

Thanks! I didn't think of using another method. 

 

After tinkering with it I came to the conclusion that it's practical to use a lambda expression based on the method you suggested.

 

Here's the code I wrote if anyone has a similar problem in the future:

Product Client::GetMostExpensiveProduct()
{
	auto mostExpensive = max_element(cProducts.begin(), cProducts.end(), [](pair<Product, int>& prod1, pair<Product, int>& prod2) 
	{ 
		return (prod1.first.GetPrice() < prod2.first.GetPrice()); 
	});
	return mostExpensive->first;
}