Txt file and data storage

Hey guys i need some help. I have a txt file which looks like this:
234 56 123 34 456 56 7 45.....
123 3 34 56 56 67 67 2.....
.
.
.
Where each line represents a series of points in coordinate system, and the first number in each line represents the x-point, second y-point,third x-point and so on. I need to read the file, and store the data. I would like to store it in a structure:
struct path {

int x[];
int y[];
};
I would also need some kind of automatic variable naming for struct path, so when i begin to store second line of points i would have struct name path2. These are just my ideas, so if you can, please help with every approach you can think of. And when the data storing is done i should be able to know the total number of paths and total number of point pairs (x,y) for every path. If you have any ideas please help, i'm a begginer and really need to solve this problem. Thank you in advance!!

If the length of each line and number of lines is unknown, you can't use stack arrays, but need dynamically allocated arrays. In C++, dynamically allocated arrays are managed by the std::vector class.

It might also be wise to keep the x and y value together, rather than in different arrays.

Code:

---

struct Point
{
    int x, y;
};

typedef std::vector<Point> Path;
typedef std::vector<Path> PathCollection;

int main()
{
    PathCollection paths;
    while (...) {
          Path p;
          //fill up Path
          paths.push_back(p);
    }
    //...
}

---

well my way is proberly the most basic way to do it (its not even in binary).

first of all set it so only 1 number is on 1 line i.e
272 //x
729 //y
383 //x

Code:

---

int main()
{
string strHoldData; // uses to put fileinput in here
int  intHoldData; // used so you can convert the string into a integer
int holdData[10]; // holds the x and y points
int arrayIncrease = 0; //used to put points in respected array slots

ifstream readCord(file name here);
 
do{
        getline(readCord,strHoldData;
        intHoldData atoi(strHoldData.c_str())
        holdData[arrayIncrease] = intHoldData;
        holdData++;
}(while !strHoldData = "")

}

---

this remains untested the result will be that Holddata[0] will hold the first x value and HoldData[1] will hold the first y value. E.T.C.

I don't reccomend using this as I am unexperieanced and this is proberly the worst way to go about it.

Thanks for the reply anon. I'm really not familiar with the vector class and i'm new to c++ so could you please tell me how can i access single item stored in vector.I think that this solution isn't good because i need to know which point is x and which y in the coordinate system. Thats why i proposed structure with two array elements. I'm probably wrong so please explain it to me. Thanks

x is Point.x and y is Point.y. Could it be any simpler?

Vector elements are accessed with operator[] just like arrays, but unlike arrays you can safely add elements to the end of the array with push_back.

Ok , that indeed is simple. So for example point[0] would represent first two pairs of x and y. I could then access them with point[0].x. Am i right? There's just one more problem. When i read the file first line is for the first path, second line is for the second path etc. Are there any means by which i could name my vectors automatically (like path1, path2...) because i don't know how many lines are there in txt file?
Maybe it will be clearer to you when i explain the whole problem. As i already mentioned lines in txt file represent paths, and numbers in lines represent dots in coordinate system (x,y). I need to read the values and store them. I need this because i will later use the coordinates of every path for robot motion. So when i'm finished with file reading and storing i should be able to get every coordinate (pair (x,y)) of every path to move my robot. I hope that this clears my problem to you, and i'm really sorry for bugging you.

You can't and you don't need to "create names" on runtime. Instead (and that is every time you wish for a way to create names with numbers appended to them) you use an array (or vector).

In my example I showed some typedefs. The second one is for a vector of vectors of Points. The point of the typedefs is that they can shorted otherwise very long typenames. Instead you could also type:

Code:

---

std::vector<Point> one_path;
std::vector<std::vector<Point> > lots_of_paths;

---

You are right about how you would access the coordinates in a single path. In a collection of paths the syntax would be similar to two-dimensional arrays:

Code:

---

unsigned path_no = 10;
unsigned point_no = 12;

lots_of_paths[path_no][point_no].x = 42;

---

Or create a reference to a single path if we need to do lots of things with it:

Code:

---

std::vector<Point>& interesting_path = lots_of_paths[42];
interesting_path[10].y = 3;

//because interesting_path is a reference the last line affects
//lots_of_paths[42][10].y

---

Thank you for everything anon. You've been a great help.

It's me again:(. lots_of_paths[0] stands for my first path, lots_of_paths[1] for second and so on (if i'm not mistaking). How can i find out the number of complete paths in lots_of_paths. Is there a way to find out number of elements in vector? And what function should i use for reading file and storing numbers? get(ch), or what? I need something that ignores spaces.

size().

This looks very familiar.

Quote:

---

Is there a way to find out number of elements in vector?

---

Yes, by using the size() member function.

I disagree with the use of a typedef here. You should create a Path class, as noted in my reply at CodeGuru.

Hey laserlight, i didn't use classes because i'm not to familiar with them and i really don't have time to study them now. Here's what i've done so far:

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
using namespace std;

class Koord {                //varijabla vrijednosti koordinatnih tocaka x i y
                int x,y;
        };
        typedef vector<Koord>Tocke;                //skup tocaka jednog puta
        typedef vector<Tocke>Putovi;                //svi putovi mape

int main(int argc, char* argv[])
{
        Putovi trajektorije;
        Tocke t;
        int i,j;
       
        if (argc == 2) {                //ako je pri pozivu naveden samo jedan parametar stvara se ulazni tok dat i pridruzuje datoteci zadanoj kao parametar
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Nije moguce otvoriti zadanu datoteku"<<argv[1]<<endl;
        }
        int br;
        while (dat.get(br)) {
                if (br == '\n')
                trajektorije.push_back(t);
                       
                t.push_back(br);
        }
       
        for(i=0; i < trajektorije.size(); i++) {
                for (j=0; j < t.size(); j++)
        cout << "Put"<<i<<" se sastoji od slijedecih tocaka"<<trajektorije[i][j]<<endl;
        }
}

---

I wanted to compile it to test the output, but i got this error:
eg_decomposition.cpp: In function 'int main(int, char**)':
eg_decomposition.cpp:48: error: no matching function for call to 'std::basic_ifstream<char, std::char_traits<char> >::get(int&)'.

Sorry for Croatian comments in the code but i think that you will get it.

Quote:

---

Hey laserlight, i didn't use classes because i'm not to familiar with them and i really don't have time to study them now.

---

This is a good opportunity to learn how to use classes properly. If you do not have the time to study them now, what do you have time for?

Quote:

---

I wanted to compile it to test the output, but i got this error:
eg_decomposition.cpp: In function 'int main(int, char**)':
eg_decomposition.cpp:48: error: no matching function for call to 'std::basic_ifstream<char, std::char_traits<char> >::get(int&)'.

---

It looks like get() is not what you want to use. You want to use formatted input with dat >> br instead.

Ok laserlight, i will do it the right way and try to do it with classes. I just don't understand how do you suggest me to store complete paths in the class you suggested.

I have one more problem with printing vector values. I get the following error when compiling:
eg_decomposition.cpp: In function 'int main(int, char**)':
eg_decomposition.cpp:66: error: no match for 'operator<<' in 'std::operator<<............

Code:

---

for(i=0; i < trajektorije.size(); i++) {
        for (j=0; j < t.size(); j++)
        cout << "Put #" << j+1 << " se sastoji od slijedecih tocaka:" << '\t' << trajektorije[i][j] << endl;
}

---

Quote:

---

I just don't understand how do you suggest me to store complete paths in the class you suggested.

---

Each Path object represents one complete path (just as each std::vector<Point> represents one complete path). The difference here is that you will provide an interface from which one can create and access this complete path. The actual implementation (i.e., whether you store the Point objects in a vector or a linked list) can vary since it is private.

I figured out that one object represents on complete path, but i don't understand how should i implement collection of object (full collection of complete paths) since i don't know the number of complete paths. How should i store points of one path in list of points? I know that i'm bugging you, but this is just one part of my project which i must do in c++, the rest of project is done in specialised language for robotics (and as you can see i'm not to familiar with c++ unfortunately). Do you know what could be the problem with printing that i mentioned in the post before?

Quote:

---

but i don't understand how should i implement collection of object (full collection of complete paths) since i don't know the number of complete paths.

---

That would still be a vector<Path>, or you could create a full blown PathCollection class instead of a typedef (but in this case a vector<Path> will probably suffice).

Quote:

---

How should i store points of one path in list of points?

---

In a sequential container such as a vector or linked list, since a path is a sequence of points.

Quote:

---

Do you know what could be the problem with printing that i mentioned in the post before?

---

It looks correct to me, so you may need to post a little more code.

The code is almost the same as the one before:

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
using namespace std;

class Koord {                //varijabla vrijednosti koordinatnih tocaka x i y

        public:
                int x,y;
        };
        typedef vector<Koord>Tocke;                //skup tocaka jednog puta
        typedef vector<Tocke>Putovi;                //svi putovi mape

int main(int argc, char* argv[])
{
        Putovi trajektorije;
        Tocke t;
        int i,j;
       
        if (argc == 2) {                //ako je pri pozivu naveden samo jedan parametar stvara se ulazni tok dat i pridruzuje datoteci zadanoj kao parametar
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Nije moguce otvoriti zadanu datoteku"<<argv[1]<<endl;
        }
       
        while (!dat.eof()) {
                int br, p=0;
                dat >> br;
               
                if (br == '\n') {
                trajektorije.push_back(t);
                p++;
        }               
               
                else if ((p%2) == 0) {       
                t[p].x=br;
                p++;
        }
               
                else {
                t[p].y=br;
                p++;
        }
        }
       
        for(i=0; i < trajektorije.size(); i++)
                for (j=0; j < t.size(); j++)
        cout << "Put #" << j+1 << " se sastoji od slijedecih tocaka:" << '\t' << trajektorije[i][j] << endl;

}               
               
        return 0;
}

---

Oh, of course. trajektorije[i][j] is a Point object, and you have not overloaded operator<< for ostream for Point objects.

Ok you lost me completely there. I don't know anything about operator overloading. How can i do that? (this c++ is complicated as hell). I've started something on the class approach so i will post my code for class and constructor. I used constructor for file opening (as you can see). Can i do that and is this ok so far? And as you can see i declared public function unos_tocaka. With it i would like to read numbers and store collection of paths in class. Is this possible and if it is how? I've done something for storing points for one path (if that is even correct) but i don't know how to go on. I'm a real pain in the ass ain't i?? ;)

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
using namespace std;

struct Koord {
       
        int x;
        int y;
}

class Putovi {
       
        public:
        int i;
        Putovi (int argc, char* argv[]);
        void unos_tocaka();
       
        private:
        vector<Koord>put;
       
};
       
Putovi::Putovi(int argc, char* argv[]) {
               
                if (argc == 2) {
                        ifstream dat(argv[1]);
                        if (!dat) {
                                cerr << "Zadanu datoteku nije moguce otvorit" << argv[1] << endl;
                                exit (1);
                        }
                }
                else if (argc < 2) {
                        cout << "Greska pri pozivanju programa" <<endl;
                        cout << "Koristenje: eg_decomposition <ime_datoteke>" <<endl;
                }
                else {
                        cerr << "Zadano je previse parametara u pozivu programa" << endl;
                        exit (1);
                }
        }

void Putovi::unos_tocaka() {
       
        while (!dat.eof()) {
                int br, p=0;
                dat >> br;
               
                if (br == '\n')
            break;
               
                else if ((p&#37;2) == 0) {       
                put[p].x=br;
                p++;
        }
               
                else {
                put[p].y=br;
                p++;
        }
               
        }


int main(int argc, char* argv[])
{
        return 0;
}

---

Quote:

---

I don't know anything about operator overloading. How can i do that?

---

If you have not learnt how to do operator overloading, then at this point just write:

Code:

---

cout << trajektorije[i][j].x << ' ' << trajektorije[i][j].y << endl;

---

However, overloading the operator is easy:

Code:

---

std::ostream& operator<<(std::ostream& out, const Point& point)
{
    return out << point.x << ' ' << point.y;
}

---

Basically, operator<< is just a function with an unusual name and some other limitations.

Quote:

---

I used constructor for file opening (as you can see). Can i do that and is this ok so far?

---

Yes, you can do that, but of course you should test if it works. However, although you can do that, it may be better to open the file in main() (or some other function), and then either pass the data read to the constructor, or pass the std::istream reference to the constructor. Passing the command line arguments to the constructor of an object that should have no knowledge of whether the command line or a graphical user interface is used certainly is not very good practice.

Quote:

---

And as you can see i declared public function unos_tocaka. With it i would like to read numbers and store collection of paths in class. Is this possible and if it is how?

---

Of course it is possible. You can overload operator>> to read into a Point. Then you can overload operator>> to read Point objects into a Path. Then you can write the code to read Path objects from a file into a vector<Path> or PathCollection. To start with, use what you know of reading integers provided by a user at the command line (i.e., cin >> num).

WHre in the code should i put the operator overloading main() or...? Because i got this error on compiling:
eg_decomposition.cpp:34: error: 'std::ostream& Koord::operator<<(std::ostream&, Koord)' must take exactly one argument

Code:

---

eg_decomposition.cpp:34: error: 'std::ostream& Koord::operator<<(std::ostream&, Koord)' must take exactly one argument

---

How can i pass the std::istream reference to the constructor as an argument?

Quote:

---

Originally Posted by radnik

WHre in the code should i put the operator overloading main() or...? Because i got this error on compiling:
eg_decomposition.cpp:34: error: 'std::ostream& Koord::operator<<(std::ostream&, Koord)' must take exactly one argument

Code:

---

eg_decomposition.cpp:34: error: 'std::ostream& Koord::operator<<(std::ostream&, Koord)' must take exactly one argument

---

How can i pass the std::istream reference to the constructor as an argument?

---

If you make the stream << or >> operator a member function (rather than a friend function), all you pass is the stream - the extra argument is the "Koord" one, since that should be implied with the member function. It is generally considered better to make it a friend function, and make it take a const reference.

--
Mats

Thanks matsp. In the first solution i didn't use the class approach but typedefs. The compiling went well but when i run the program with correct file path to open i get a segmentation fault. Can you please check whats the problem. I first need one working solution to continue my project and i will then try to finish the class approach. Anyway here is my code:

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
using namespace std;

class Koord {                //varijabla vrijednosti koordinatnih tocaka x i y

        public:
                int x;
                int y;
               
        };
        typedef vector<Koord>Tocke;                //skup tocaka jednog puta
        typedef vector<Tocke>Putovi;                //svi putovi mape

int main(int argc, char* argv[])
{
       
        Putovi trajektorije;
        Tocke t;
        int i,j;
       
        if (argc == 2) {                //ako je pri pozivu naveden samo jedan parametar stvara se ulazni tok dat i pridruzuje datoteci zadanoj kao parametar
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Nije moguce otvoriti zadanu datoteku "<< argv[1]<<endl;
        }
       
        while (!dat.eof()) {
                int br, p=0;
                dat >> br;
               
                if (br == '\n') {
                trajektorije.push_back(t);
                p++;
        }               
               
                else if ((p%2) == 0) {       
                t[p].x=br;
                p++;
        }
               
                else {
                t[p].y=br;
                p++;
        }
        }
       
        for(i=0; i < trajektorije.size(); i++)
                for (j=0; j < t.size(); j++)
        cout << "Put #" << j+1 << " se sastoji od slijedecih tocaka:" << '\t' << trajektorije[i][j].x <<' '<<trajektorije[i][j].y << endl;

}       
else if (argc < 2) {
                        cout << "Greska pri pozivanju programa" <<endl;
                        cout << "Koristenje: eg_decomposition <ime_datoteke>" <<endl;
                }
                else {
                        cerr << "Zadano je previse parametara u pozivu programa" << endl;
                        exit (1);
                }       
               
        return 0;
}

---

Can you post the input file. I'm suspecting that you are trying to access the vector before you have given it any size (before you do "put_back()").

--
Mats

Here's the file. This is just one example. The file can have unknown number of lines (paths) and unknown number of point per path (where first number is x, second y, third x and so on):

Code:

---

12 14 15 17 987 234 23 45
1 2 65 34 56 34 12 78 23 987

---

As I suspected, the problem is that you are accessing the t[p] before it has been assigned to anything. You will either have to collect a coordinate pair and add that to the vector when it's complete, or give the vector a size before you access elements in the vector.

Also this:

Code:

---

                if (br == '\n') {
                trajektorije.push_back(t);
                p++;
        }

---

is completely broken - br == '\n' will be true when you see the value 10 in the input stream, not when you hit a newline, since br is an integer, and not a char, the input filtering will ignore newlines.

If you need to know about newlines, I would suggest you read the input into a string using getline, then use stringstream to parse out the integer values.
--
Mats

Ok, thank you. Where can i read about usage of getline and stringstream because i don't have any books on c++?

Can someone please help me with this stringstream. Just give me a good link if you know it or explain it to me how to parse integers from string with it please.

Quote:

---

Originally Posted by radnik

Can someone please help me with this stringstream. Just give me a good link if you know it or explain it to me how to parse integers from string with it please.

---

A stringstream works just like an input or output stream. So we could do something like this:

Code:

---

#include <sstream>

int main()
{
    std::stringstream ss;

    ss << "100 200 300";

    int x;
    while(ss >> x)
        std::cout >> x >> std::endl;
    return 0;
}

---

This would give you an output of
100
200
300

--
Mats

stringstreams are a kind of streams, therefore the interface and usage is practically the same as cout, cin and fstreams.

The difference from file streams is that they don't work on a file, but on a string. So, in the constructor you don't specify a file name but a string.

Code:

---

#include <iostream>
#include <sstream>
#include <string>

using namespace std;

int main()
{
    string s("10 23 -3 18 9");
    stringstream ss(s);
    int n;
    while (ss >> n) {
        std::cout << n << '\n';
    }
}

---

Oh, i see. Thank you guys very much.

I'm stuck again. Now when i do something like:

Code:

---

string tmp;
               
getline(dat, tmp);
stringstream stmp(tmp);

---

i have stringstream stmp with one line of numbers from txt file. But how do i read stringstream and store data in x and y coordinates. I understand the simple example that you gave me, but have no idea on how to do this :(
Is there a way to test end of stringstream in wile loop or something similar??

This line in both mine and matsp's example tests for end of stringstream in exactly the same way you'd check for end of user input with std::cout (Ctrl + Z) and end of file if simply reading integers from a file.

Code:

---

while (ss >> n)

---

You can read two variable at a time:

Code:

---

#include <iostream>
#include <sstream>
#include <string>

using namespace std;

int main()
{
    string s("10 23 -3 18 9 100 83 29 1 -1");
    stringstream ss(s);
    int n, m;
    while (ss >> n >> m) {
        std::cout << n << ' ' << m << '\n';
    }
}

---

However, it is somewhat hard to detect problems while parsing the line (e.g odd number of values), so it just assumes original input is good.

Thanks, that helped a lot. What do you suggest me to do on writing numbers of one path in vector Path. Since i don't set it's size (for logical reasons) i can't use something like:

Code:

---

while (stmp >> x >> y) {
                       
                        //cout <<"broj is smtp"<<br<<"referenca od p"<<pp<<endl;
                       
                                t[pp].x=x;
                                t[pp].y=y;
                                pp++;
        }

---

What would be my other options? I know how i can count lines in text and in that way set the size of vector Collection of paths. But how can i count numbers per line?? As a matter of fact even if i could count them not every line has the same amount of numbers so this approach is no good. Suggestions please.

The good thing about vectors is that you don't need to know how much data you are going to store beforehand: push_back.

I know that, but when i read two values x and y how can i do bush back. How can i know that it will read the x value in vetor.x and y in vector.y?So when i read the values from file like this:

Code:

---

while (stmp >> x >> y) {
                       
                                //cout <<"broj x        "<<x<<"        broj y        "<<y<<"        broj pp        "<<pp<<endl;

---

how can i store it in vector of classes Koord?:

Code:

---

class Koord {                //varijabla vrijednosti koordinatnih tocaka x i y

        public:
                double x;
                double y;
               
        };
        typedef vector<Koord>Tocke;                //skup tocaka jednog puta
        typedef vector<Tocke>Putovi;                //svi putovi mape

---

I know that i will fill "collectionofpaths" with trajektorije.push_back(t); (Putovi trajektorije;, Tocke t), as you said me to do it earlier in this topic.

Guys i'm really desperate here. I'm obviously doing something wrong and i don't know what. I post the code and you compile it and look at the output. I have no clue what is the problem and i'm sure that it's something trivial but i don't have the experience to notice it. When i try to print the whole trajektorije vector which is collectionofpaths i get some random memory junk. And is still have the problem i mentioned in the earlier post above. So please if anyone could help me i would be grateful.Here's the code:

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
#include <string>
#include <sstream>
using namespace std;

class Koord {                //varijabla vrijednosti koordinatnih tocaka x i y

        public:
                double x;
                double y;
               
        };
        typedef vector<Koord>Tocke;                //skup tocaka jednog puta
        typedef vector<Tocke>Putovi;                //svi putovi mape

int main(int argc, char* argv[])
{
       
        Putovi trajektorije(15);
        Tocke t(50);
       
        if (argc == 2) {                //ako je pri pozivu naveden samo jedan parametar stvara se ulazni tok dat i pridruzuje datoteci zadanoj kao parametar
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Nije moguce otvoriti zadanu datoteku "<< argv[1]<<endl;
        }
       
        while (!dat.eof()) {
               
                string tmp;
                double x,y;
               
                getline(dat, tmp);
                stringstream stmp(tmp);
                //cout << "Linija:"<<tmp<<endl;
                int p=-1;
                int &pp = p;
               
                while (stmp >> x >> y) {
                               
                                p++;
                                cout <<"brojac pp="<<pp<<"\n"<<"broj x        "<<x<<"        broj y        "<<y<<endl;
                       
                                t[pp].x=x;
                                t[pp].y=y;
                                cout <<"broj x unutar vektora        "<<t[pp].x<<"        broj y unutar vektora        "<<t[pp].y<<endl;
        }

               
                trajektorije.push_back(t);
                //t.clear();
        }       

       

}       
else if (argc < 2) {
                        cout << "Greska pri pozivanju programa" <<endl;
                        cout << "Koristenje: eg_decomposition <ime_datoteke>" <<endl;
                }
                else {
                        cerr << "Zadano je previse parametara u pozivu programa" << endl;
                        exit (1);
                }       
               
                ifstream dat (argv[1]);
                dat.clear();
                dat.seekg(0);
               
                string linija;
                int s=0;
                int &size = s;
               
                while (getline (dat,linija)) {
                        size++;
                }
                s=size;
                trajektorije.resize(s);
               
                int i,j;
                for(i=0; i < trajektorije.size(); i++) {
                        for (j=0; j < t.size(); j++)
                                cout << '\t' << trajektorije[i][j].x <<' '<<trajektorije[i][j].y << endl;
        }
       
        return 0;
}

---

Anyone?? :(

I suggest that you format your code properly, and more importantly, use English. You can always convert them to your natural language later, but when asking for help in a message board community that is pre-dominantly English, write in English.

It would also help if you divided your code into well named functions. This would allow you to test a particular function to see if it is working, and thus become confident that if there is a problem, it most likely lies elsewhere. This divide and conquer approach will help you debug your code, and help others understand it. Likewise, use meaningful variable names, not t and p.

I may have gotten lost, but why do you resize trajektorije after all the data has been read in?

And for that matter, why bother with the s and the size -- since size is a reference to s, they will always and forever refer to the same variable, so there's no need for both.

Code:

---

        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Nije moguce otvoriti zadanu datoteku "<< argv[1]<<endl;
        }

---

I think you want "if(!dat.is_open())" there.

Anyway. I'm not too sure what you're trying to do -- I haven't read the whole thread -- but perhaps this will be helpful:

Code:

---

#include <iostream>
#include <vector>

class pos {
public:
    int xp, yp;
    pos() {}
    pos(int x, int y) : xp(x), yp(y) {}
};

int main() {
    std::vector<pos> v(10);
    v.push_back(pos(2, 3));
    v.push_back(pos(5, 4));

    return 0;
}

---

You are right laserlight. Sorry for the Croatian language. I will post English version now. I don't know what you mean by formatting of code because all of the "layout" was done by my editor i didn't touch a thing. I will put all of the code in functions later.
Dwks thank you. That helped a lot. But still while printing values of paths vector i get some memory junk in between my real values. Like this:

Code:

---

       
        12 14
        15 17              PATH 1
        987 234
        23 45

     
        3.60739e-313 4.95672e-270
        1.54972e-312 12
        14 15
        17 987
        234 23                          SOME MEMORY JUNK AND REPEATED VALUES FROM PATH 1
        12 14
        15 17
        987 234
        23 45

       
        1 2
        65 34
        56 34                    PATH 2
        12 78
        23 987

---

I'll post new code with English and deleted unnecessary parts:

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
#include <string>
#include <sstream>
using namespace std;

class Coord {                //values of coordinates (x,y)

        public:
                double xt;
                double yt;
               
                Coord(){};
                Coord(double x, double y) : xt(x), yt(y) {};
               
        };
       
typedef vector<Coord>Dots;                //collection of coordinates for one path (one line in txt file is one path)
typedef vector<Dots>Paths;                //collection of all paths

int main(int argc, char* argv[])
{
       
        Paths trajectories;
        Dots pairs;
        int i,j;
       
        if (argc == 2) {       
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Given file cannot be opened "<< argv[1]<<endl;
        }
       
        while (!dat.eof()) {
               
                string tmp;
                double x,y;
               
                getline(dat, tmp);
                stringstream stmp(tmp);

                int p=-1;
                int &pp = p;
               
                while (stmp >> x >> y) {
                               
                                p++;
                                cout <<"counter pp="<<pp<<"\n"<<"number x        "<<x<<"        number y        "<<y<<endl;
                       
                                pairs.push_back(Coord(x,y));

                                cout <<"number x inside vector        "<<pairs[pp].xt<<"        number y inside vector        "<<pairs[pp].yt<<endl;
                                // the lines for printing are just for testing because i had some problems here
                }
               
                trajectories.push_back(pairs);       
        }       
}       
else if (argc < 2) {
                cout << "Error while running program" <<endl;
                cout << "Usage: eg_decomposition <file_name>" <<endl;
        }
else {
                cerr << "Too much parameters" << endl;
                exit (1);
        }       
               
// printing of all paths               
for(i=0; i < trajectories.size(); i++) {
        for (j=0; j < pairs.size(); j++)
                cout << '\t' << trajectories[i][j].xt <<' '<<trajectories[i][j].yt << endl;
                               
        }
       
return 0;
}

---

If i'm right the repeating of numbers occurs because i never clear my pairs vector so those old coordinates remain in the vector. How can i solve this? I've tried this:

Code:

---

while (stmp >> x >> y) {
                               
                                p++;
                                cout <<"counter pp="<<pp<<"\n"<<"number x        "<<x<<"        number y        "<<y<<endl;
                       
                                tmp_pairs.push_back(Coord(x,y));

                                cout <<"number x inside vector        "<<tmp_pairs[pp].xt<<"        number y inside vector        "<<tmp_pairs[pp].yt<<endl;
                                // the lines for printing are just for testing because i had some problems here
                }
               
                pairs=tmp_pairs;
                trajectories.push_back(pairs);
                tmp_pairs.clear();

---

But then when i tried to print the values i got nothing. Why is this happening if i cleared the temporary vector tmp_pairs? I presume that i can't do something like pairs=tmp_pairs;. How could i do this the other way? I tried the vector copying method with vector<Coord> pairs(tmp_pairs) but still got the same results :( Is there any method of appending values of one vector to another?

Quote:

---

If i'm right the repeating of numbers occurs because i never clear my pairs vector so those old coordinates remain in the vector. How can i solve this?

---

The easiest way that I can think of? Declare pairs at the beginning of the while loop, so that a new variable is constructed for each iteration of the loop. :)

I just noticed that you posted your input file a while back. I'm testing your program now . . . .

[edit] By doing that, and by changing the reference to pairs after the while loop, your program works correctly, at least I think it does. :)

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
#include <string>
#include <sstream>
using namespace std;

class Coord {          //values of coordinates (x,y)

        public:
                double xt;
                double yt;
               
                Coord(){};
                Coord(double x, double y) : xt(x), yt(y) {};
               
        };
       
typedef vector<Coord>Dots;              //collection of coordinates for one path (one line in txt file is one path)
typedef vector<Dots>Paths;              //collection of all paths

int main(int argc, char* argv[])
{
       
        Paths trajectories;
       
        int i,j;
       
        if (argc == 2) {       
       
        ifstream dat (argv[1]);
        if (!dat) {
                cerr << "Given file cannot be opened "<< argv[1]<<endl;
        }
       
        while (!dat.eof()) {
                Dots pairs;
                string tmp;
                double x,y;
               
                getline(dat, tmp);
                stringstream stmp(tmp);

                int p=-1;
               
                while (stmp >> x >> y) {
                               
                                p++;
                                cout <<"counter pp="<<p<<"\n"<<"number x      "<<x<<" number y        "<<y<<endl;
                       
                                pairs.push_back(Coord(x,y));

                                cout <<"number x inside vector  "<<pairs[p].xt<<"      number y inside vector  "<<pairs[p].yt<<endl;
                                // the lines for printing are just for testing because i had some problems here
                }
               
                trajectories.push_back(pairs); 
        }     
}     
else if (argc < 2) {
                cout << "Error while running program" <<endl;
                cout << "Usage: eg_decomposition <file_name>" <<endl;
        }
else {
                cerr << "Too much parameters" << endl;
                exit (1);
        }     
               
// printing of all paths               
for(i=0; i < trajectories.size(); i++) {
        for (j=0; j < trajectories[i].size(); j++)
                cout << '\t' << trajectories[i][j].xt <<' '<<trajectories[i][j].yt << endl;
                               
        }
       
return 0;
}

---

You'll notice I removed pp. It's the same thing as p, after all. [/edit]

Here's the code from my previous post modified slightly. It does exactly the same thing. Consider my changes suggestions. :)

Code:

---

#include <iostream>
#include <cstdlib>
#include <fstream>
#include <vector>
#include <string>
#include <sstream>
using namespace std;

class Coord {          //values of coordinates (x,y)
public:
        double xt;
        double yt;
       
        Coord(){};
        Coord(double x, double y) : xt(x), yt(y) {};
       
};

typedef vector<Coord>Dots;              //collection of coordinates for one path (one line in txt file is one path)
typedef vector<Dots>Paths;              //collection of all paths

int main(int argc, char* argv[])
{
        Paths trajectories;
       
        if (argc == 2) {
                ifstream dat (argv[1]);
                if (!dat) {
                        cerr << "Given file cannot be opened "<< argv[1]<<endl;
                }
       
                string line;
                while(getline(dat, line)) {
                        stringstream stmp(line);
                        Dots pairs;
                        double x,y;
                       
                        while (stmp >> x >> y) {
                                pairs.push_back(Coord(x,y));
                        }
                       
                        trajectories.push_back(pairs);
                }
        }
        else if (argc < 2) {
                cout << "Error while running program" <<endl;
                cout << "Usage: eg_decomposition <file_name>" <<endl;
        }
        else {
                cerr << "Too much parameters" << endl;
                exit (1);
        }
       
        // printing of all paths
        for(Paths::size_type i=0; i < trajectories.size(); i++) {
                for (Dots::size_type j=0; j < trajectories[i].size(); j++)
                        cout << '\t' << trajectories[i][j].xt <<' '<<trajectories[i][j].yt << endl;
               
        }
       
        return 0;
}

---

A brief overview of my changes:

• I indented your code a little bit more consistently, trying to stick with your indentation style.
• I removed the p variable, because you don't need it when you're just printing stuff. (Besides, it's the same thing as pairs.size().)
• I used size_types for the loop control variables because otherwise you have a signed/unsigned type mismatch. Basically, vector::size() returns an unsigned integer, and int is a signed integer; they have different ranges, and this could cause problems if you had [i]extremely[i] large vectors. You could also just cast the .size() calls to (int), or ignore this altogether. :)
• I removed the while(!dat.eof()) call. That can cause problems; see this: http://faq.cprogramming.com/cgi-bin/...&id=1043284351

[edit] Oh, and I syntax highlighted the code with codeform. :) [/edit]

Dwks thank you very much, this is great. I have only one more thing to do. Now i must calculate Euclidean distance for all paths in trajectories vector. Sort them and find the shortest path so i can send the robot down that path. I've solved the robot motion part of course but how do you suggest that i do the calculation and finding of the shortest path? Once again thank you very much for all your help.
By the way, you've done very good job on codeform ;)

For the distances, use Pythagoras's theorem.

Code:

---

#include <cmath>

double hypotenuse(double x, double y) {
    return std::sqrt(x*x + y*y);
}

---

(C99 has a hypot() function that does this. C++ may, you might want to try it.)

I'm not sure which pairs of points you want to examine, but finding the shortest path is pretty easy. It should just be a matter of finding the lowest number, no?

Let me know if you have trouble with that . . . .

[edit] Note that you'll have to pass a hypot()-like function with the differences between coordinates, like this:

Code:

---

Coord one(4, 5), two(6, 8);
double distance = hypotenuse(one.xt - two.xt, one.yt - two.yt);

---

[/edit]

You can see the formula i need to use on this link:
http://en.wikipedia.org/wiki/Euclidean_metric.
I thought to calculate distance for all paths and store these values in vector of doubles. Then to sort them in some way. But how could i know to which path does the smallest number (shortest distance) belong to. Should i calculate those values for all paths again and compare them with the one that is the shortest?
I should pass values from the trajectories vector while calculating distances.

If all you're trying to do is figure out the shortest distance, then you don't need to record every distance and then sort them. You can just calculate one distance; if it's the shortest distance found so far, then record the set of points you used to calculate this distance and the actual distance.

Something like this.

Code:

---

        double value, min_value = distance(trajectories[0]);
        Paths::size_type min_index = 0;
       
        for(Paths::size_type i=1; i < trajectories.size(); i++) {
                value = distance(trajectories[x]);
                if(value < min_value) {
                        min_value = value;
                        min_index = i;
                }
        }
       
        // now trajectories[min_index] has the minimum Euclidean distace,
        // the exact value of which is recorded in min_value.

---

You'll have to write the distance() function, of course. :) That should be pretty easy, though.

[edit] Just use a for loop to loop through all the items in a Paths. Calculate (x1-x2)**2 for each pair of coordinates, as your Wiki page indicates, sum those values, and return the square root of the final sum . . . . [/edit]

Best of luck, I'm off. :)

Ok, thanks. I'll try that later today and post you on how it went.

Ok i have problem again right at the start. I don't know how to pass vector with index (vector[x]) as function argument. I did this:

Code:

---

double distance (Paths path[i]) {
        double dist, tmp;
       
        for (Dots::size_type j=0; j<path[i].size[];j++) {
       
                tmp += pow((path[i][j].xt - path[i][++j].xt),2) + pow((path[i][j].yt - path[i][++j].yt),2);
        }
        dist = sqrt(tmp);
        return dist;
}

---

and during compiling i got error that i and path aren't declared in this scope of course. How can i pass it as argument? And i'm not sure about the ++j in the calculation line. In fact i'm quite sure that it is wrong. How can i access the next coordinate if not in this manner?