Using STL functions with a struct? - c++

Question

I've recently read up on STL functions in C++. I understand the basic uses of the functions, but I am struggling getting them to use member variables of a struct.
I have this struct:
struct Apples
{
double weight; // oz
string color; // red or green
void print() const { cout << color << ", " << weight << endl; }
};
Basically, I insert Apples into a vector storing random weights and random color. Now, I want to use a count_if function to determine how many apples are greater than a given weight. I want to convert a function like this:
int cnt = 0;
for(auto it = crate.cbegin(); it != crate.cend(); ++it)
if(it->weight > toFind)
cnt++;
to a count_if() version (this does not work):
int cnt = count_if(crate.begin(), crate,end(), isGreater())
With isGreater() being like this:
void isGreater()
{
if(it->weight > toFind)
return it->weight > toFind;
}
What I don't understand about STL functions and a struct is how to use the member variables inside of the struct with the STL functions. I'm not sure what to pass inside of the STL function, either. Would it be better to use a lambda function in this case? If so, why?
Here is all the current code, if it doesn't make sense:
#include <iostream>
#include <ctime>
#include <cstdlib>
#include <vector>
#include <deque>
#include <string>
using namespace std;
struct Apples
{
double weight; // oz
string color; // red or green
void print() const { cout << color << ", " << weight << endl; }
};
void isGreater()
{
if(it->weight > toFind)
return it->weight > toFind;
}
int main()
{
srand(time(nullptr));
const double minWeight = 8.;
const double maxWeight = 3.;
cout << "Input crate size: ";
int size;
cin >> size;
vector <Apples> crate(size);
for(auto it = crate.begin(); it != crate.end(); ++it)
{
it->weight = minWeight + static_cast<double>(rand())/RAND_MAX*(maxWeight - minWeight);
it->color = rand() % 2 == 1 ? "green" : "red";
}
cout << "Enter weight to find: ";
double toFind;
cin >> toFind;
//this is what I want to convert to count if
int cnt = 0;
for(auto it = crate.cbegin(); it != crate.cend(); ++it)
if(it->weight > toFind)
cnt++;

Answer 1

std::count_if takes unary predicate as the third argument. In this case unary predicate is a function taking one object and returning true if object matches find criterion or false if not.
Since your criterion depends on toFind, it seems more laconic to use lambda capturing toFind:
int cnt = count_if(crate.begin(), crate.end(), [toFind](const Apple& apple) {
return it->weight > toFind;
});
If you want a standalone function, you can use:
bool isGreater(double toFind, const Apple& apple) {
return it->weight > toFind;
}
...
int cnt = count_if(crate.begin(), crate.end(),
std::bind(&isGreater, toFind, std::placeholders::_1));
Note, that you don't need to call function, you need to pass it:
int cnt = count_if(crate.begin(), crate,end(), isGreater())
// ^^ remove parentheses

Answer 2

you are not storing the apples in the vector.
you have to initialize inside a loop each apple and then store them in the vector.
crate.push_back(newApple).
so run a loop from 0 to size.
inside that loop initialize new apples and give them weights and colors
then push_back in vector:
for(int i = 0; i < size ++i)
{
apples newApple;
newApple.weight = ...;
newApple.color = ...;
crate.push_back(newApple);
}

Answer 3

This is usually accomplished by creating a "functor" class, a class whose objects can be called like a function. Each instance call hold the reference weight:
struct IsGreater {
double w;
IsGreater(double weight) : w{weight} {}
bool operator()(const Apples& A) const {
return A.weight > w;
}
};
Then we just need to create an instance of the class holding the reference weight and pass it to count_if:
const int count = std::count_if(crate.begin(), crate.end(), IsGreater(toFind));
You can avoid creating an explicit class using a lambda:
const int count = std::count_if(crate.begin(), crate.end(),
[=](const Apples& A) -> bool {
return A.weight > toFind;
});
Here the reference value toFind is captured by value.