Reading from file and saving to struct

[aram4]

I have read so many different opinions, I have gotten myself confused.
I am trying to read from a file (of the "user's" specification) and store the data in a struct named sample.

Here's where I'm stuck:
After using the getline function and reading the data in as a string, how is the best way to assign each character a value in my struct?
For example, if getline returns 54A (I cleaned all spacing and excess characters out of the line already), then how do I manage to assign the first int (5) to the variable X, the second int (4) to the variable Y, and the char (A) to the variable class_type. I have tried using a for loop and I am just STUCK.

Any advice is appreciated.

[Chervil]

Before attempting to answer, I'd like to know more about the original input. What was the contents of the line before it was tidied up. In particular, can the numbers contain multiple digits?

What does the contents of the file look like please.

I'm assuming your struct looks like this, but some clarification would be useful too:

struct abcd {
    int X;
    int Y;
    char class_type;
}

[aram4]

Yes, my struct is very similar to that. Here it is:

struct sample
{
int F1;
int F2;
char char_class;
};

The data input file is as follows:
5, 4, A
6, 3, B
4, 3, A
2, 7, B
6, 1, A
5, 7, B
I cleaned out the white spaces and commas. There is only one digit per variable, so I tried using something like this (I deleted it out of my code already):

string current_line;

current_line[0 ] = sample.F1;
current_line[1] = sample.F2;
current_line[2] = sample.class_type;

[aram4]

I suppose another piece of valuable info is that I have 6 samples, so I have to either create each struct dynamically in the for loop, or I have to assign a sample object to each line that is read in.

[aram4]

Here is what I have done:

#include <iostream>
#include <string>
#include <fstream>
#include <cstdlib>
#include <cstring>

using namespace std;

void clean_data(string &str);
//double distance(double x1, double x2, double y1, double y2);

string file_name;
string current_line;
string converted_line;

int m,n;
int sample_count = 0;

ifstream input(file_name);

struct sample
{
	int F1;
	int F2;
	char class_type;
}samples[10];


int main()
{
	cout << "Enter the number of samples in the data set:\n";
	cin >> m;
	cout << "Enter the number of features in the data set:\n";
	cin >> n;
	sample samples[10];
	cout << "Enter the name of the data file (append\".txt\"):\n";
	cin >> file_name;
	input.open(file_name.c_str());
	do
    {
		dcom:
		getline(input,current_line);
		clean_data(current_line);
		
		cout << current_line << endl;
		
		// store data into struct sample
		//array to hold structs
		
		if(input.eof())
			break;
		else
			goto dcom;
		
		
			

	}while(!input.eof());


	
	system("pause");
	return 0;
}

void clean_data(string &str)
{
    for (int i = 0; i < str.length(); i++)
        if ((str[i]==',')||(str[i]== ' '))
        {
            str.erase(i,1);
            i--;
        }
	
}
/*double distance(double x1,double y1,double x2 ,double y2)
{
	return sqrt((x2-x1)*(x2-x1) + (y2-y1)*(y2-y1));

}*/

[Chervil]

I've omitted the struct here. But here's a way to get the data into int, int, char.
My version of clean_data() replaces any comma with a space.

#include <iostream>
#include <fstream>
#include <iomanip>
#include <string>
#include <sstream>

using namespace std;

void clean_data(string &str);

int main() 
{
    ifstream fin("input.txt");
    string line;
    
    int a, b;
    char ch;
    
    while (getline(fin, line))
    {
        clean_data(line);
        istringstream ss(line);
        ss >> a >> b >> ch;
        cout << "a: " << a << "  b: " << b << "  ch: " << ch << endl;
    }
    
    return 0;
}

void clean_data(string &str)
{
    size_t pos=0;
    while ((pos = str.find(',', pos)) != string::npos )
        str[pos] = ' ';
}

Output:

a: 5  b: 4  ch: A
a: 6  b: 3  ch: B
a: 4  b: 3  ch: A
a: 2  b: 7  ch: B
a: 6  b: 1  ch: A
a: 5  b: 7  ch: B

[aram4]

Thank you! That helped me get started in the right direction. I think I'm going to take a break and get back to it tomorrow!

[aram4]

Okay, I have used what you gave me, and I know my code is a little sloppy with comments. I was just trying to get all my thoughts out.
I'm not really sure how to proceed. I have my data in the struct, and I can access it. I have to calculate the distance for each data set, find the nearest hit (closest distance of same class type) and the nearest miss(closest distance of opposite class type), then use the relief algorithm to rank the features, which is: W(F1) = (0+[-(diff(point1, near hit)^2+(diff(point1,near miss)^2]+[-(diff(point2, near hit)^2+(diff(point2, near miss)^2] +...[-(diff(pointn, near hit)^2]+(diff(pointn, near miss)))/number of samples. Then I will do the same with F2 and output the features in order of highest score to lowest score.
I'm telling you all this, so you can see where I'm going. What is the best way to procede? My initial idea to store the differences in an array isn't going to work because it's impossible to tell which points the data originated from. The function doesn't work right, anyways. I just didn't want to invest more time in it, if it's not going to do what I need it to do.
I thought to make another struct named distance to hold the distance info, but I can't seem to figure out how to pack all the data I need into the rest of this program. Help, please! I just need ideas of how to pack all this into a neat little package. It seems so messy and all over the place to me.

Here is what I have done so far: As I said above, the distance function is wrong, and I know it. Just left it as is so you could see what I'm working with.

#include <iostream> 
#include <string>
#include <fstream>
#include <sstream>
#include <iomanip>
#include <cmath>

using namespace std;

string file_name;
string current_line;

int m,n;
double F1, F2;
char class_type;

ifstream input(file_name);
ofstream output;

//struct for each sample
struct sample
{
	double F1;
	double F2;
	char class_type;
};


sample p[6]; //array of 6 samples
sample data;//each individual sample
double q[100]; //array to hold distances


void print(sample Array[], ofstream& output);
void clean_data(string &str);
void distance(sample p[], double q[], ofstream& output);

int main()
{
	cout << "Enter the number of samples in the data set:\n";
	cin >> m;
	cout << "Enter the number of features in the data set:\n";
	cin >> n;
	
	cout << "Enter the name of the data file (append\".txt\"):\n";
	cin >> file_name;
	
	input.open(file_name.c_str());
	output.open("output.txt");
	
	for (int i = 0; i < m && !input.eof(); i++)
		{
			getline(input, current_line);
			clean_data(current_line);
			istringstream ss(current_line);
			ss >> F1 >> F2 >> class_type;

			p[i].F1 = F1;				//assigns the values from file into the array of structs
			p[i].F2 = F2;
			p[i].class_type = class_type;

		}
    

	print(p,output);
	distance(p,q,output);
	

	output.close();
	input.close();
	
	system("pause");
	return 0;
}
//output to text file
void print(sample Array[], ofstream& output)
{
	for ( int i=0; i<m; i++ )
		output << setw(15) << Array[i].F1 <<setw(15)
			<< Array[i].F2 << setw(40)
			<< Array[i].class_type << endl;
}
//removes commas
void clean_data(string &str)
{
    size_t pos=0;
    while ((pos = str.find(',', pos)) != string::npos )
        str[pos] = ' ';
}

void distance(sample p[], double q[], ofstream& output)
{
	double distance;
	for (int i = 0; i < m; i++)
	{
		for (int j = 1; j < m; j++)
			distance = sqrt(((p[i].F1 - p[j].F1)*(p[i].F1 - p[j].F1))-((p[i].F2 - p[j].F2)*(p[i].F2 - p[j].F2)));
		q[i] = distance;
		output << q[i] << endl;
	}	
	
}

//function to find near miss and near hit and store in array for each data point 
	//check class type 
	//for same class, return smallest difference as near hit
	//for diff class, return smallest difference as near miss
	//store each value in an array
//function to calculate releif algorithm
	//FOR EACH FEATURE
	//make near hit a negative value
	//add near miss to near hit
	//square the sum and add to the next instance
	//
	

[Chervil]

Wow - this question became a lot more complicated since yesterday!
That's fine, but I don't have any immediate response to make, I need to try to understand exactly what it all means before saying anything useful.

In the meantime of course, anyone else is welcome to contribute if there are ideas out there.