Memory leak without allocating any memory? - c++

I'm working on a coding assignment for a C++ class. When I run my program I seem to be dealing with a memory leakage issue, which is weird since I am NOT explicitly allocating any memory in my code. I ran the program under gdb, and it seems as though the program crashes when running the destructor for a Deck object. I tried stepping through the code, but I when I do so I end up in a host of .h files related to vectors. Then suddenly, it stops. I tried going to a TA for some help, but they seem to be as perplexed as I am on the issue.
# include <stdlib.h>
# include <time.h>
# include <iostream>
# include <vector>
# include <stdio.h>
using namespace std;
//function signatures
float bustProbability (const int);
class Deck
{
public:
//data members
vector <int> cardArray;
vector <int> wasteCards;
//constructor
Deck();
//methods
void shuffleDeck();
void populateDeckWithCards();
void removeCopyCards();
int dealCard();
int remainingCards();
void showCards();
};
void Deck::removeCopyCards() {
for (unsigned int i = 0; i < wasteCards.size(); i++) {
bool removedCopy = false;
for (unsigned int j = 0; j < cardArray.size() && removedCopy == false; j++) {
if (cardArray[j] == wasteCards[i]) {
cardArray.erase (cardArray.begin() + j - 1);
removedCopy = true;
}
}
}
}
int Deck::dealCard() {
if (remainingCards() > 0) {
int tmp = cardArray.back();
wasteCards.push_back(tmp);
cardArray.pop_back();
return tmp;
}
else {
populateDeckWithCards();
removeCopyCards();
shuffleDeck();
//shuffle method
int tmp = cardArray.back();
cardArray.pop_back();
return tmp;
}
}
void Deck::populateDeckWithCards() {
//populate regular cards into array
for (int i = 2; i <= 10; i++) {
for (int j = 0; j < 4; j++) {
cardArray.push_back(i);
}
}
//populate J, Q, K into array
for (int i = 0; i < 3; i++) {
for (int j = 0; j < 4; j++) {
cardArray.push_back(10);
}
}
//populating array with Aces... treating them as special case '100'
for (int i = 0; i < 4; i++) {
cardArray.push_back(100);
}
return;
}
void Deck::showCards() {
for (unsigned int i = 0; i < cardArray.size(); i++) {
cout << cardArray[i] << endl;
}
}
Deck::Deck() {
wasteCards.clear();
cardArray.clear();
populateDeckWithCards();
shuffleDeck();
}
void Deck::shuffleDeck() {
int n = cardArray.size();
for(int a = n-1; a > 0; a--) {
int min = 0;
int max = a;
int j = min + rand() / (RAND_MAX / (max-min + 1) + 1);
int tmp = cardArray[a];
cardArray[a] = cardArray[j];
cardArray[j] = tmp;
}
return;
}
int Deck::remainingCards() {
return cardArray.size();
}
class Player {
public:
//data members
vector <int> playerHand;
//constructor
Player();
//methods
bool isBust();
int count();
void hit(Deck&);
void stand();
bool muckHand();
void showHand();
};
Player::Player() {
playerHand.clear();
}
void Player::showHand() {
for (unsigned int i = 0; i < playerHand.size(); i++) {
cout << playerHand[i] << endl;
}
return;
}
int Player::count() {
int handCount = 0;
for (unsigned int i = 0; i < playerHand.size(); i++) {
if (playerHand[i] != 100)
handCount += playerHand[i];
else {
if (playerHand[i] == 100) {
if ((handCount) > 11) {
handCount += 1;
}
else
handCount += 10;
}
}
}
return handCount;
}
bool Player::isBust() {
if (count() > 21)
return true;
else
return false;
}
void Player::hit(Deck& d) {
playerHand.push_back(d.dealCard());
}
void Player::stand() {
return;
}
bool Player::muckHand() {
playerHand.clear();
return true;
}
float bustProbability (const int threshHold) {
int threshHoldReached = 0;
Deck myDeck;
Player myPlayer;
Player dealer;
for (int i = 0; i < 10000; i++) {
myPlayer.hit(myDeck);
dealer.hit(myDeck);
myPlayer.hit(myDeck);
dealer.hit(myDeck);
while (myPlayer.count() < threshHold) {
myPlayer.hit(myDeck);
}
if (!(myPlayer.isBust())) {
++threshHoldReached;
}
myDeck.wasteCards.clear();
myPlayer.muckHand();
dealer.muckHand();
}
float bustFraction = float(threshHoldReached)/float(10000);
return bustFraction;
}
int main () {
cout << "blackjack simulation" << endl;
srand((unsigned int)time(NULL));
cout << bustProbability(19);
return 0;
}
I'm incredibly sorry for just posting my code, but I've spend 4 days on this issue, and I can't even begin to figure out what the problem is.

There is at least the line
cardArray.erase (cardArray.begin() + j - 1);
which seems to be dubious in case of j = 0

Related

'this' cannot be used in a constant expression error

The error is on line 76 int res[mSize]; the problem is on mSize. It seems like a simple fix but I can't figure it out. If someone can figure it out or point me in the right direction that would be greatly appreciated.
Also, the deconstructor ~MyContainer(), I am not sure if I am using it right or if there is a correct place to put it.
Here is my code:
#include <iostream>
using namespace std;
class MyContainer
{
private:
int* mHead; // head of the member array
int mSize; // size of the member array
public:
MyContainer();
MyContainer(int*, int);
//~MyContainer();
void Add(int);
void Delete(int);
int GetSize();
void DisplayAll();
int FindMissing();
~MyContainer() {}
};
MyContainer::MyContainer()
{
mHead = NULL;
mSize = 0;
}
MyContainer::MyContainer(int* a, int b)
{
mHead = a;
mSize = b;
}
void MyContainer::Add(int a)
{
*(mHead + mSize) = a;
mSize++;
}
void MyContainer::Delete(int a)
{
int index;
for (int i = 0; i < mSize; i++)
{
if (*(mHead + i) == a)
{
index = i;
break;
}
}
for (int i = index; i < mSize; i++)
{
*(mHead + i) = *(mHead + i + 1);
}
mSize--;
}
int MyContainer::GetSize()
{
return mSize;
}
void MyContainer::DisplayAll()
{
cout << "\n";
for (int i = 0; i < mSize; i++)
{
cout << *(mHead + i) << " ";
}
}
int MyContainer::FindMissing()
{
int res[mSize];
int temp;
int flag = 0;
for (int i = 1; i <= mSize; i++)
{
flag = 0;
for (int j = 0; j < mSize; j++)
{
if (*(mHead + j) == i)
{
flag = 1;
break;
}
}
if (flag == 0)
{
temp = i;
break;
}
}
return temp;
}
int main()
{
const int cSize = 5;
int lArray[cSize] = { 2, 3, 7, 6, 8 };
MyContainer lContainer(lArray, cSize);
lContainer.DisplayAll();
lContainer.Delete(7);
lContainer.DisplayAll();
cout << "Size now is: " << lContainer.GetSize() << endl; lContainer.Add(-1);
lContainer.Add(-10);
lContainer.Add(15);
lContainer.DisplayAll();
cout << "Size now is: " << lContainer.GetSize() << endl;
cout << "First missing positive is: " << lContainer.FindMissing() << endl;
system("PAUSE"); return 0;
}
int res[mSize];
The size of the array mSize must be known at compile time. You cannot use a variable here. An option may be to define a macro with an largish value that will not exceeded.
static const int kLargeSize =100;
int res[kLargeSize];
Edited in response to the comments - const and constexpr are a better option than a macro.
Or even better, you can use std::vector - https://en.cppreference.com/w/cpp/container/vector

can't figure out where the heap overwriting is

#include<iostream>
using namespace std;
class Text{
public:
~Text(){
delete data;
}
char* data{};
int mSize{};
void fill(char* stringInput) {
mSize = strlen(stringInput);
data = new char [mSize];
for (int i = 0; i < mSize; i++){
data[i] = stringInput[i];
}
}
};
class myString{
public:
explicit myString(int size){ // constructor
strAmount = size;
strings = new Text [size];
}
~myString(){ // destructor
delete[] strings;
}
void addString(char* input){
strings[filledAmount].fill(input);
filledAmount++;
}
void delString(int pos){
for ( int i = pos; i < filledAmount; i++){
swap(strings[i], strings[i+1]);
}
strings[filledAmount].data = nullptr;
strings[filledAmount].mSize = 0;
filledAmount--;
}
void eraseEverything(){
for ( int i = 0; i < filledAmount; i++){
strings[i].data = {};
strings[i].mSize = 0;
}
filledAmount = 0;
}
int maxString() const {
int index{};
for ( int i = 0 ; i < filledAmount; i++){
if (strings[i].mSize > strings[index].mSize){
index = i;
}
}
return index;
}
int charAmount(){
int counter{};
for(int i = 0 ; i < filledAmount; i++){
counter+=strings[i].mSize;
}
return counter;
}
double digitPercentage(){
int digitsAmount{};
for(int i = 0; i < filledAmount; i++){
for ( int j = 0; j < strings[i].mSize; j++){
if (isdigit(strings[i].data[j])){
digitsAmount++;
}
}
}
double digitPercent = (digitsAmount/(double)charAmount())*100;
return digitPercent;
}
int filledAmount{};
int strAmount{};
Text* strings;
};
void render_text(myString& obj) {
for (int k = 0; k < obj.filledAmount; k++) {
for (int i = 0; i < obj.strings[k].mSize; i++)
cout << obj.strings[k].data[i];
cout << endl;
}
cout << endl;
}
int main(){
myString a(5);
a.addString((char *) "zxc 1v1 forever shadow fiend");
a.addString((char *) "This is a string");
a.addString((char *) "12345");
a.addString((char *) "Hello");
a.addString((char *) "A1oha Dance");
render_text(a);
a.delString(1);
render_text(a);
int maxInd = a.maxString();
cout << "Max string :\n";
for (int i = 0; i < a.strings[maxInd].mSize; i++) {
cout << a.strings[maxInd].data[i];
}
cout << "\n\n";
}
Please help me find the crash point. I suppose it crashes in the destructor pole, but I still can't figure it out.
This is something like a self-written string class, the problem is that I can't find the place where the problems start.
I also have a thought that the destructor tries to delete too much memory from the heap so the compiler prevents it from doing that. Can I somehow change the size of the strings array?
I have fixed all your memory errors, with the help of address sanitizers
The main change here is:
Add copy constructors and copy assignment operators
Manually delete the last element in the function delString
Though that the code now works, it's not a true modern c++ style code. I highly recommend that your using std::string to replace your Text class. Use std::vector to replace the dynamic array. Then you will stay away from the pain of memory errors. The delString should be replaced with vector::erase,which is much neat than your hand write algorithm.
https://en.cppreference.com/w/cpp/string/basic_string
https://en.cppreference.com/w/cpp/container/vector
I strongly recommend rewriting the code with std::string and std::vector
#include <cstring>
#include <iostream>
using namespace std;
class Text {
public:
~Text() { delete[] data; }
char* data{};
int mSize{};
Text() = default;
Text(const Text& oth) {
mSize = oth.mSize;
data = new char[mSize];
std::copy(oth.data, oth.data + oth.mSize, data);
}
Text& operator=(const Text& oth) {
delete[] data;
mSize = oth.mSize;
data = new char[mSize];
std::copy(oth.data, oth.data + oth.mSize, data);
return *this;
}
void fill(char* stringInput) {
mSize = strlen(stringInput) + 1;
data = new char[mSize];
for (int i = 0; i < mSize; i++) {
data[i] = stringInput[i];
}
}
};
class myString {
public:
explicit myString(int size) { // constructor
strAmount = size;
strings = new Text[size];
}
myString(const myString& oth) {
strAmount = oth.strAmount;
strings = new Text[oth.strAmount];
for (size_t i = 0; i < strAmount; ++i) {
strings[i] = oth.strings[i];
}
}
myString& operator=(const myString& oth) {
delete[] strings;
strAmount = oth.strAmount;
strings = new Text[oth.strAmount];
for (size_t i = 0; i < strAmount; ++i) {
strings[i] = oth.strings[i];
}
return *this;
}
~myString() { // destructor
delete[] strings;
}
void addString(char* input) {
strings[filledAmount].fill(input);
filledAmount++;
}
void delString(int pos) {
for (int i = pos; i < filledAmount; i++) {
swap(strings[i], strings[i + 1]);
}
delete[] strings[filledAmount].data;
strings[filledAmount].data = nullptr;
strings[filledAmount].mSize = 0;
filledAmount--;
}
void eraseEverything() {
for (int i = 0; i < filledAmount; i++) {
strings[i].data = {};
strings[i].mSize = 0;
}
filledAmount = 0;
}
int maxString() const {
int index{};
for (int i = 0; i < filledAmount; i++) {
if (strings[i].mSize > strings[index].mSize) {
index = i;
}
}
return index;
}
int charAmount() {
int counter{};
for (int i = 0; i < filledAmount; i++) {
counter += strings[i].mSize;
}
return counter;
}
double digitPercentage() {
int digitsAmount{};
for (int i = 0; i < filledAmount; i++) {
for (int j = 0; j < strings[i].mSize; j++) {
if (isdigit(strings[i].data[j])) {
digitsAmount++;
}
}
}
double digitPercent = (digitsAmount / (double)charAmount()) * 100;
return digitPercent;
}
int filledAmount{};
int strAmount{};
Text* strings = nullptr;
};
void render_text(myString& obj) {
for (int k = 0; k < obj.filledAmount; k++) {
for (int i = 0; i < obj.strings[k].mSize; i++)
cout << obj.strings[k].data[i];
cout << endl;
}
cout << endl;
}
int main() {
myString a(6);
a.addString((char*)"zxc 1v1 forever shadow fiend");
a.addString((char*)"This is a string");
a.addString((char*)"12345");
a.addString((char*)"Hello");
a.addString((char*)"A1oha Dance");
render_text(a);
a.delString(1);
render_text(a);
int maxInd = a.maxString();
cout << "Max string :\n";
for (int i = 0; i < a.strings[maxInd].mSize; i++) {
cout << a.strings[maxInd].data[i];
}
cout << "\n\n";
return 0;
}

Can't convert memory address to value

thanks for visiting my question! Currently, when I run this code (an implementation of the Set data structure in c++), the memory address of each element in the set is printed instead of the values in the set. To help with debugging, here's my code:
#include <iostream>
#include <algorithm>
#include <string>
#include <vector>
#include <cassert>
#define PB push_back
typedef std::vector<int> vint;
class SetBase
{
public:
SetBase() {}
~SetBase() {}
void insert(int x)
{
if (!this->isInSet(x))
{
this->set.PB(x);
}
}
int size()
{
return this->set.size();
}
bool empty()
{
return this->size() == 0;
}
int operator[](int index)
{
if (index >= 0 && index < this->size())
{
return this->set[index];
}
else
{
return -1;
}
}
bool find(int target)
{
sort(this->set.begin(), this->set.end());
int low = 0, high = this->size();
while (low <= high)
{
long long mid = low + (high - low) / 2;
long long guess = this->set[mid];
if (guess == target)
return true;
else if (guess < target)
low = mid + 1;
else
high = mid - 1;
}
return false;
}
int count(int target)
{
int counter = 0;
for (int i = 0; i < this->set.size(); i++)
{
if (this->set[i] == target)
counter++;
}
return counter;
}
bool operator=(SetBase &other)
{
if (other.size() != this->size())
return false;
for (int i = 0; i < other.size(); i++)
{
if (other[i] != this->set[i])
return false;
}
return true;
}
private:
vint set;
bool isInSet(int target)
{
for (int i = 0; i < this->size(); i++)
{
if (set[i] == target)
{
return true;
}
}
return false;
}
};
class Set : public SetBase
{
public:
void set_union(Set *set1, Set *set2, Set &back_insertor)
{
for (int i = 0; i < set1->size(); i++)
{
if (this->isInSet(back_insertor, i))
{
back_insertor.insert(i);
}
}
}
void set_difference(Set set1, Set set2, Set &back_insertor)
{
// set_difference = set1 - set2
}
void set_intersection(Set set1, Set set2, Set &back_insertor)
{
// set_difference = set1 U set2
for (int i = 0; i < set1.size(); i++)
{
for (int j = 0; j < set2.size(); j++)
{
if (set1[i] == set2[j])
{
back_insertor.insert(set1[i]);
}
}
}
}
void printSet(Set *in)
{
for (int i = 0; i < in->size(); i++)
{
std::cout << &in[i] << "\n";
}
}
private:
bool isInSet(SetBase set1, int target)
{
for (int i = 0; i < set1.size(); i++)
{
if (target == set1[i])
{
return true;
}
}
return false;
}
};
int main()
{
Set *set_1 = new Set();
Set *set_2 = new Set();
Set *back = new Set();
for (int i = 1; i <= 10; i++)
set_1->insert(i);
for (int i = 1; i <= 10; i++)
set_2->insert(i);
set_2->insert(11);
set_1->set_union(set_1, set_2, *back);
set_1->printSet(set_1);
delete set_1;
delete set_2;
delete back;
}
When running the set_1->printSet(set_1); line, here's what I get:
0x7fb498c05a20
0x7fb498c05a38
0x7fb498c05a50
0x7fb498c05a68
0x7fb498c05a80
0x7fb498c05a98
0x7fb498c05ab0
0x7fb498c05ac8
0x7fb498c05ae0
0x7fb498c05af8
Even though this works, I would like to print out the values (integer) instead. Any help would be appreciated! Thanks! :)
Inside printSet(), you use &in[i] to print each element.
The & operator returns the address of the object you are referencing. So, instead of getting the value, you are getting its address. You should remove it, eg:
void printSet(Set *in)
{
for (int i = 0; i < in->size(); i++)
{
std::cout << (*in)[i] << "\n";
}
}

Segmentation Fault when dealing with 2D Arrays [closed]

Closed. This question needs debugging details. It is not currently accepting answers.
Edit the question to include desired behavior, a specific problem or error, and the shortest code necessary to reproduce the problem. This will help others answer the question.
Closed 3 years ago.
Improve this question
I'm am trying to dynamically make 2d arrays that are then supposed to be iterated through to check their contents. Whenever I try to use a function that indexes the array I get a segmentation fault. The two functions that are creating the problems are the printg() and get() functions. I'm not sure exactly what I'm doing wrong, but neither of them will work properly for me.
Any help would be great. Thank you.
#ifndef _GRID_H
#define _GRID_H
#include <iostream>
using namespace std;
class Grid
{
public:
Grid();
Grid(const Grid& g2);
Grid(int x, int y, double density);
Grid(string file);
~Grid();
bool check(int x, int y); //check if a cell is inhabited or not
bool isEmpty();//check if a grid is living
bool equals(const Grid& g2);//checks if two grids are equal
void kill(int x, int y);//kill a cell
void grow(int x, int y);//grow a cell
int getSize();
int getNumRows();
int getNumCol();
int getNumLiving();
void printg(int r, int c);
char get(int x, int y) const;
private:
int size; //number of cells in grid
int row; //row length (number of columns)
int column; //column length (number of rows)
int num_living; //number of X's in the grid
char** myGrid;
};
#endif
#include "Grid.h"
#ifndef _GRID_C
#define _GRID_C
#include <iostream>
#include <cmath>
#include <fstream>
#include <cstdlib>
//compile with g++ -I /home/cpsc350/GameOfLife Grid.cpp
using namespace std;
Grid::Grid() //do i need a default constructor????
{
char myGrid[10][10] = {{0,1,2,3}, {4,5,6,7}, {8,9,10,11}};
row = 10;
column = 10;
size = 100;
}
Grid::Grid(const Grid& g2)//copy constructor/////////////help
{
size = g2.size;
row = g2.row;
column = g2.column;
num_living = g2.num_living;
char** myGrid = new char*[row];
for(int i = 0; i < row; i++)
myGrid[i] = new char[column];
for(int i1 = 0; i1 < row; i1++)
{
for(int i2 = 0; i2 < column; i2++)
{
//copy(&g2[i1][i2], &g2[i1][i2]+row*column,&myGrid[i1][i2]);
myGrid[i1][i2] = g2.get(i1,i2);
}
}
}
Grid::Grid(int x, int y, double density)
{
char** myGrid = new char*[x];
for(int i = 0; i < x; i++)
myGrid[i] = new char[y];
row = x;
column = y;
size = x*y;
num_living = size * density;
string str = "";
for(int a = 0; a < num_living; a++)//adds the density of X's to a string
{
str += 'X';
}
for(int a = 0; a < size - num_living; a++)//adds the rest to the string
{
str += '-';
}
int randnum;
//randomly generates indicies in the string str and puts them into the array
for(int i1 = 0; i1 < column; i1++)
{
for(int i2 = 0; i2 < row; i2++)
{
//generate random numbers from index 0 to length of string - 1
if(str.length()>1)
{
randnum = (rand()%(str.length()-1))+1;
}
else
{
randnum = 0;
}
myGrid[i1][i2] = str[randnum];
str.erase(randnum);
}
}
}
Grid::Grid(string file)
{
num_living = 0;
//code to create a 2d array from a filepath
ifstream openfile(file);
//error handling
if(! openfile)
{
cout << "Error, file could not be opened" << endl;
exit(0);
}
openfile >> column;//gets number of rows
openfile >> row;//gets number of columns
size = row*column;
char** myGrid = new char*[row];
for(int i = 0; i < row; i++)
myGrid[i] = new char[column];
for(int x = 0; x<column; x++)
{
for(int y = 0; y<row; y++)
{
openfile >> myGrid[x][y];
if(! openfile)//error handling
{
cout << "Error reading file at " << row << "," << column << endl;
}
if(myGrid[x][y] == 'X')
{
num_living++;
}
}
}
openfile.close();
}
Grid::~Grid()
{
if(myGrid)
{
for(int i = 0; i < row; i++)
{
delete []myGrid[i];
}
delete []myGrid;
}
}
void Grid::kill(int x, int y)
{
if(myGrid[x][y] == 'X')
{
num_living--;
}
myGrid[x][y] = '-';
}
void Grid::grow(int x, int y)
{
if(myGrid[x][y] == '-')
{
num_living++;
}
myGrid[x][y] = 'X';
}
bool Grid::check(int x, int y)
{
if(y<0 || x<0)
{
return(false);
}
return (myGrid[x][y] == 'X');
}
bool Grid::isEmpty()
{
return (num_living == 0);
}
bool Grid::equals(const Grid& g2)
{
if(size != g2.size) //checks if sizes are equal
{
return false;
}
if(row != g2.row)//checks if numRows are equal
{
return false;
}
if(column != g2.column)//checks if numCol are equal
{
return false;
}
if(num_living != g2.num_living)//checks if numliving are equal
{
return false;
}
for(int x = 0; x < row; x++)//checks each element
{
for(int y = 0; y < column; y++)
{
if(myGrid[x][y] != g2.get(x,y))
{
return false;
}
}
}
return true;
}
int Grid::getSize()
{
return(size);
}
int Grid::getNumRows()
{
return(column);
}
int Grid::getNumCol()
{
return(row);
}
int Grid::getNumLiving()
{
return(num_living);
}
void Grid::printg(int r, int c)
{
for(int x = 0; x < r; x++)
{
for(int y = 0; y < c; y++)
{
cout << myGrid[x][y];
}
cout << endl;
}
}
char Grid::get(int x, int y) const
{
return myGrid[x][y];
}
#endif
The problem that I see at first is that both your default and copy constructor do not initialize myGrid. what you are doing in them will create an additional array with the same name which 'shadows' myGrid. instead you have to do:
Grid::Grid(const Grid& g2)
{
size = g2.size;
row = g2.row;
column = g2.column;
num_living = g2.num_living;
myGrid = new char*[row]; // removed "char**" at the start of this line
for(int i = 0; i < row; i++)
myGrid[i] = new char[column];
for(int i1 = 0; i1 < row; i1++)
{
for(int i2 = 0; i2 < column; i2++)
{
//copy(&g2[i1][i2], &g2[i1][i2]+row*column,&myGrid[i1][i2]);
myGrid[i1][i2] = g2.get(i1,i2);
}
}
}
your default constructor has the same problem. but note that you can't initialize it with braces. but you don't have to have a default constructor if you are not using it.

c++ - Segmentation fault for class function of vector of custom class

I am using following code to run kmeans algorithm on Iris flower dataset- https://github.com/marcoscastro/kmeans/blob/master/kmeans.cpp
I have modified the above code to read input from files. Below is my code -
#include <iostream>
#include <vector>
#include <math.h>
#include <stdlib.h>
#include <time.h>
#include <algorithm>
#include <fstream>
using namespace std;
class Point
{
private:
int id_point, id_cluster;
vector<double> values;
int total_values;
string name;
public:
Point(int id_point, vector<double>& values, string name = "")
{
this->id_point = id_point;
total_values = values.size();
for(int i = 0; i < total_values; i++)
this->values.push_back(values[i]);
this->name = name;
this->id_cluster = -1;
}
int getID()
{
return id_point;
}
void setCluster(int id_cluster)
{
this->id_cluster = id_cluster;
}
int getCluster()
{
return id_cluster;
}
double getValue(int index)
{
return values[index];
}
int getTotalValues()
{
return total_values;
}
void addValue(double value)
{
values.push_back(value);
}
string getName()
{
return name;
}
};
class Cluster
{
private:
int id_cluster;
vector<double> central_values;
vector<Point> points;
public:
Cluster(int id_cluster, Point point)
{
this->id_cluster = id_cluster;
int total_values = point.getTotalValues();
for(int i = 0; i < total_values; i++)
central_values.push_back(point.getValue(i));
points.push_back(point);
}
void addPoint(Point point)
{
points.push_back(point);
}
bool removePoint(int id_point)
{
int total_points = points.size();
for(int i = 0; i < total_points; i++)
{
if(points[i].getID() == id_point)
{
points.erase(points.begin() + i);
return true;
}
}
return false;
}
double getCentralValue(int index)
{
return central_values[index];
}
void setCentralValue(int index, double value)
{
central_values[index] = value;
}
Point getPoint(int index)
{
return points[index];
}
int getTotalPoints()
{
return points.size();
}
int getID()
{
return id_cluster;
}
};
class KMeans
{
private:
int K; // number of clusters
int total_values, total_points, max_iterations;
vector<Cluster> clusters;
// return ID of nearest center (uses euclidean distance)
int getIDNearestCenter(Point point)
{
double sum = 0.0, min_dist;
int id_cluster_center = 0;
for(int i = 0; i < total_values; i++)
{
sum += pow(clusters[0].getCentralValue(i) -
point.getValue(i), 2.0);
}
min_dist = sqrt(sum);
for(int i = 1; i < K; i++)
{
double dist;
sum = 0.0;
for(int j = 0; j < total_values; j++)
{
sum += pow(clusters[i].getCentralValue(j) -
point.getValue(j), 2.0);
}
dist = sqrt(sum);
if(dist < min_dist)
{
min_dist = dist;
id_cluster_center = i;
}
}
return id_cluster_center;
}
public:
KMeans(int K, int total_points, int total_values, int max_iterations)
{
this->K = K;
this->total_points = total_points;
this->total_values = total_values;
this->max_iterations = max_iterations;
}
void run(vector<Point> & points)
{
if(K > total_points)
return;
vector<int> prohibited_indexes;
printf("Inside run \n");
// choose K distinct values for the centers of the clusters
printf(" K distinct cluster\n");
for(int i = 0; i < K; i++)
{
while(true)
{
int index_point = rand() % total_points;
if(find(prohibited_indexes.begin(), prohibited_indexes.end(),
index_point) == prohibited_indexes.end())
{
printf("i= %d\n",i);
prohibited_indexes.push_back(index_point);
points[index_point].setCluster(i);
Cluster cluster(i, points[index_point]);
clusters.push_back(cluster);
break;
}
}
}
int iter = 1;
printf(" Each point to nearest cluster\n");
while(true)
{
bool done = true;
// associates each point to the nearest center
for(int i = 0; i < total_points; i++)
{
int id_old_cluster = points[i].getCluster();
int id_nearest_center = getIDNearestCenter(points[i]);
if(id_old_cluster != id_nearest_center)
{
if(id_old_cluster != -1)
clusters[id_old_cluster].removePoint(points[i].getID());
points[i].setCluster(id_nearest_center);
clusters[id_nearest_center].addPoint(points[i]);
done = false;
}
}
// recalculating the center of each cluster
for(int i = 0; i < K; i++)
{
for(int j = 0; j < total_values; j++)
{
int total_points_cluster = clusters[i].getTotalPoints();
double sum = 0.0;
if(total_points_cluster > 0)
{
for(int p = 0; p < total_points_cluster; p++)
sum += clusters[i].getPoint(p).getValue(j);
clusters[i].setCentralValue(j, sum / total_points_cluster);
}
}
}
if(done == true || iter >= max_iterations)
{
cout << "Break in iteration " << iter << "\n\n";
break;
}
iter++;
}
// shows elements of clusters
for(int i = 0; i < K; i++)
{
int total_points_cluster = clusters[i].getTotalPoints();
cout << "Cluster " << clusters[i].getID() + 1 << endl;
for(int j = 0; j < total_points_cluster; j++)
{
cout << "Point " << clusters[i].getPoint(j).getID() + 1 << ": ";
for(int p = 0; p < total_values; p++)
cout << clusters[i].getPoint(j).getValue(p) << " ";
string point_name = clusters[i].getPoint(j).getName();
if(point_name != "")
cout << "- " << point_name;
cout << endl;
}
cout << "Cluster values: ";
for(int j = 0; j < total_values; j++)
cout << clusters[i].getCentralValue(j) << " ";
cout << "\n\n";
}
}
};
int main(int argc, char *argv[])
{
srand(time(NULL));
int total_points, total_values, K, max_iterations, has_name;
ifstream inFile("datafile.txt");
if (!inFile) {
cerr << "Unable to open file datafile.txt";
exit(1); // call system to stop
}
inFile >> total_points >> total_values >> K >> max_iterations >> has_name;
cout << "Details- \n";
vector<Point> points;
string point_name,str;
int i=0;
while(inFile.eof())
{
string temp;
vector<double> values;
for(int j = 0; j < total_values; j++)
{
double value;
inFile >> value;
values.push_back(value);
}
if(has_name)
{
inFile >> point_name;
Point p(i, values, point_name);
points.push_back(p);
i++;
}
else
{
inFile >> temp;
Point p(i, values);
points.push_back(p);
i++;
}
}
inFile.close();
KMeans kmeans(K, total_points, total_values, max_iterations);
kmeans.run(points);
return 0;
}
Output of code is -
Details-
15043100000Inside run
K distinct cluster i= 0
Segmentation fault
When I run it in gdb, the error shown is -
Program received signal SIGSEGV, Segmentation fault.
0x0000000000401db6 in Point::setCluster (this=0x540, id_cluster=0)
at kmeans.cpp:41
41 this->id_cluster = id_cluster;
I am stuck at this as I cannot find the cause for this segmentation fault.
My dataset file looks like -
150 4 3 10000 1
5.1,3.5,1.4,0.2,Iris-setosa
4.9,3.0,1.4,0.2,Iris-setosa
4.7,3.2,1.3,0.2,Iris-setosa
. . .
7.0,3.2,4.7,1.4,Iris-versicolor
6.4,3.2,4.5,1.5,Iris-versicolor
6.9,3.1,4.9,1.5,Iris-versicolor
5.5,2.3,4.0,1.3,Iris-versicolor
6.5,2.8,4.6,1.5,Iris-versicolor
. . .
in KMeans::run(vector<Point>&) you call points[index_point].setCluster(i); without any guarantee that index_point is within bounds.
index_point is determined by int index_point = rand() % total_points;, and total_points is retrieved from the input file "datafile.txt" which could be anything. It certainly does not have to match points.size(), but it should. Make sure it does, or just use points.size() instead.
A bit offtopic, but using rand() and only using modulo is almost always wrong. If you use C++11 or newer, please consider using std::uniform_int_distribution.
points[index_point].setCluster(i); could be accessing the vector out of bounds. The code you quoted actually always sets a number of total_points in the vector points before calling run, while your modified code just reads until end of file and has no guarantees that the number of total points passed to the constructor of KMeans matches the value of entries in points. Either fix your file I/O or fix the logic of bounds checking.