Jump to content

  • Log In with Google      Sign In   
  • Create Account


I need help with vectors of vectors.


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
14 replies to this topic

#1 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 16 December 2012 - 08:43 PM

Hello. I am trying to something like a 2d array of bools (ex: bool var[11][11]). But using vectors instead.



I think that this doesn't work because you cannot put a vector in data type bool.
#include <iostream>
#include <vector>
using namespace std;
int main()
{
vector<vector<bool>> mv;
int x = 11;
int y = 11;
for(int i=0;i<x;i++) {
  mv.push_back(vector<bool>());
  cout<<"Creating x["<<mv.size()-1<<"]\n";
  for(int j=0;j<y;j++) {
   mv[i].push_back(vector<bool>());
  }
}
return 0;
}

But in order to create a dynamic function that takes x and y and creates a 2d vector of bools I have no idea where to begin.

Sponsor:

#2 Álvaro   Crossbones+   -  Reputation: 11906

Like
0Likes
Like

Posted 16 December 2012 - 09:10 PM

What do you mean by "doesn't work"? What error message do you get?

#3 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 16 December 2012 - 09:13 PM

What do you mean by "doesn't work"? What error message do you get?


I'm getting this error
error C2664: 'std::vector<_Ty,_Ax>::push_back' : cannot convert parameter 1 from 'std::vector<_Ty,_Ax>' to 'bool'

Btw that error corresponds to line 13.

Edited by black_darkness, 16 December 2012 - 09:13 PM.


#4 JTippetts   Moderators   -  Reputation: 8159

Like
2Likes
Like

Posted 16 December 2012 - 09:16 PM

In line 13, you should just be pushing a bool, not a vector of bools. You would only push a vector in that inner loop if you were implementing a vector of vectors of vectors of bools (a 3D array, in other words).

Edited by JTippetts, 16 December 2012 - 09:20 PM.


#5 nobodynews   Crossbones+   -  Reputation: 1774

Like
3Likes
Like

Posted 16 December 2012 - 09:31 PM

The error is saying exactly the issue. You're pushing back a vector of bools and that code is expecting a bool. Try changing line 13 to this:

mv[i].push_back(bool());

Of course a better solution is to make this function:

std::vector<std::vector<bool>> Create2DVectorOfBools(int x, int y)
{
  std::vector<bool> temp(y, false);
  std::vector<std::vector<bool>> mv(x, temp);

  return mv;
}

What this does is first creates a 1d vector of bools of length y and initializes every element in it to 'false' (see here for info on every constructor of std::vector http://msdn.microsoft.com/en-us/library/zzw4bwhd.aspx )
Then it creates a vector of vector of bools of length x and initializes every element in it to 'temp' (which, as stated, was initialized to length y).
Then it returns a copy of the resulting 2d vector.

C++: A Dialog | C++0x Features: Part1 (lambdas, auto, static_assert) , Part 2 (rvalue references) , Part 3 (decltype) | Write Games | Fix Your Timestep!


#6 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 16 December 2012 - 09:43 PM

Thank you both. I see the problem now. I was having trouble conceptualizing this. Now I got it. Thank you.

@nobodynews that is a good function. I was planning on creating a function later. Thanks for the good function.

Edited by black_darkness, 16 December 2012 - 09:43 PM.


#7 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 16 December 2012 - 10:24 PM

Here is the finished product.
vector<vector<bool>> mv;
int x = 11;
int y = 11;
for(int i=0;i<x;i++) {
  mv.push_back(vector<bool>());
  for(int j=0;j<y;j++) {
   mv[i].push_back(new bool);
  }
}

Will this handle the garbage collecting?
for (unsigned int i=0;i<mv.size();i++) {
  mv[i].erase(mv[i].begin());
}


#8 JTippetts   Moderators   -  Reputation: 8159

Like
3Likes
Like

Posted 16 December 2012 - 10:27 PM

mv.clear() would be sufficient to clear everything.

#9 black_darkness   Members   -  Reputation: 280

Like
2Likes
Like

Posted 16 December 2012 - 10:34 PM

mv.clear() would be sufficient to clear everything.


Thank you.

BTW. I found this code just now for visual studio compilers. src http://www.david-amador.com/2010/10/tracking-memory-leaks-in-visual-studio/

It detects memory leaks. I had a ton.
Do you recommend this as a good way for me to test for errors in the future?

#ifdef _DEBUG
int flag = _CrtSetDbgFlag(_CRTDBG_REPORT_FLAG);
flag |= _CRTDBG_LEAK_CHECK_DF; // Turn on leak-checking bit
_CrtSetDbgFlag(flag);
#endif


#10 iMalc   Crossbones+   -  Reputation: 2260

Like
2Likes
Like

Posted 17 December 2012 - 01:42 AM

Here is the finished product.

vector<vector<bool>> mv;
int x = 11;
int y = 11;
for(int i=0;i<x;i++) {
  mv.push_back(vector<bool>());
  for(int j=0;j<y;j++) {
   mv[i].push_back(new bool);
  }
}

Whoa hold on a sec there. You're dynamically allocating each bool. That is not a "finished product", and is not right at all. You would simply need to push_back the value, i.e. false. It's only compiling due to a pointer being able to be implicity converted to a bool, with the value 'true' in this case since the pointer isn't NULL. So it leaks like a sieve. It's also much longer, and much less efficient in both memory and time taken*, even if you ignore the masses of leaks it causes. Nope, far from being finished, what you actually have there is downright awful.

* The code nobodynews posted performs a copy-construction of each inner vector which results in never having to grow the vector because it starts out at the ideal size. Your code on the other hand causes the vector to grow when it reaches a certain size, and it will grow the capacity to probably larger than you need it for.

Do it the right way. Taking the important bits from nobodynews' code and inserting them into yours produces this:
int x = 11, y = 11;
std::vector< std::vector<bool> > mv(x, std::vector<bool>(y, false));
Now that's how you do it properly!

Edited by iMalc, 17 December 2012 - 01:43 AM.

"In order to understand recursion, you must first understand recursion."
My website dedicated to sorting algorithms

#11 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 17 December 2012 - 03:33 AM


Here is the finished product.

vector<vector<bool>> mv;
int x = 11;
int y = 11;
for(int i=0;i<x;i++) {
  mv.push_back(vector<bool>());
  for(int j=0;j<y;j++) {
   mv[i].push_back(new bool);
  }
}

Whoa hold on a sec there. You're dynamically allocating each bool. That is not a "finished product", and is not right at all. You would simply need to push_back the value, i.e. false. It's only compiling due to a pointer being able to be implicity converted to a bool, with the value 'true' in this case since the pointer isn't NULL. So it leaks like a sieve. It's also much longer, and much less efficient in both memory and time taken*, even if you ignore the masses of leaks it causes. Nope, far from being finished, what you actually have there is downright awful.

* The code nobodynews posted performs a copy-construction of each inner vector which results in never having to grow the vector because it starts out at the ideal size. Your code on the other hand causes the vector to grow when it reaches a certain size, and it will grow the capacity to probably larger than you need it for.

Do it the right way. Taking the important bits from nobodynews' code and inserting them into yours produces this:
int x = 11, y = 11;
std::vector< std::vector<bool> > mv(x, std::vector<bool>(y, false));
Now that's how you do it properly!


I feel evil now. Thanks. I was wondering why the program wouldn't work. I got distracted trying to solve my memory leaks and couldn't fix them. Then I came back and saw your post. I appreciate it. It doesn't help that cplusplus.com is down, which is my reference of choice for problems.

#12 Álvaro   Crossbones+   -  Reputation: 11906

Like
2Likes
Like

Posted 17 December 2012 - 03:39 AM

Once you are using vectors, you shouldn't write `new' anywhere: The vector is doing the memory allocations for you. If you feel tempted to use `new' again, ask here to make sure you really have a good reason (hint: unless you are trying to get polymorphic behavior, you won't need it).

#13 Nanook   Members   -  Reputation: 471

Like
0Likes
Like

Posted 17 December 2012 - 05:03 AM

Are you sure you actually want to use a std::vector<bool> ?

http://www.gotw.ca/publications/N1185.pdf
http://www.gotw.ca/p...tions/N1211.pdf

Edited by Nanook, 17 December 2012 - 05:06 AM.


#14 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 17 December 2012 - 10:05 PM

My test I wrote is compiling properly but It crashes.

When it crashes it tells me
Expression: vector<bool> iterator not dereferencable.

And
Standard C++ libraries are out of range &&0

What I am trying to get the test to do is make a class where I can input x,y, and a string. X and Y will determine width and height. And the string will determine if the coordinate is passable or impassable. I keep thinking I solve the problem but I keep getting an error. I am starting to get lost because I keep having to correct mistakes and I forget what mistakes I made. I am getting lost I feel like I am in a vicious cycle. Any help would be appreciated.

vector<vector<bool>> make_map(int x, int y, string grid)
{
vector<vector<bool>> mv(x, vector<bool>(y, false));
int counter = 0;
for(int i=0;i<x;i++) {
  for(int j=0;i<y;j++) {
   if(grid[counter]=='X') {
    mv[i][j]=false;
   }
   else if(grid[counter]=='O') {
    mv[i][j] = true;
   }
   counter++;
  }
}
return mv;
}

int main()
{
int mx = 3;
int my = 3;
string grid = "XOXOOOOXO";
vector<vector<bool>> mv = make_map(mx,my,grid);
for(int i=0;i<mx;i++) {
  for(int j=0;j<my;j++) {
   cout<<"X: "<<i<<"Y: "<<j<<"= "<<mv[i][j]<<"\n";
  }
}
return 0;
}


Edited by black_darkness, 17 December 2012 - 10:07 PM.


#15 black_darkness   Members   -  Reputation: 280

Like
0Likes
Like

Posted 17 December 2012 - 10:15 PM

Well I solved this one myself. The problem was on line 6. I had an i instead of a j. Now everything works. I laughed like a maniac. I am very pleased with this.

Edited by black_darkness, 17 December 2012 - 10:16 PM.





Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS