questionsolved  Initialising 2D vectors

j3lps (11)   Link to this post
Hi All,

I wonder if anyone can help me with this - I'm trying to create a 2D vector.

So in a header file I have:
std::vector< std::vector< float > > item;
What would I then put in the cpp file to make sure it is initialised, before changing the size to what I need using resize, so that it becomes, for example, a 2x2 vector of 0s?

Any suggestions?
J
Last edited on
Bazzy (4111)   Link to this post
std::vector< std::vector< float > > item ( 2, std::vector<float> ( 2, 0 ) );
The first argument for the vector constructor is the number of elements, the second, the value used to fill the vector
http://www.cplusplus.com/reference/stl/vector/vector/
j3lps (11)   Link to this post
Thanks, but I'm not sure how to use that for what I want... Is it possible just to define the vector as a 2-D vector in the header file and then to initialise it in the cpp file? I could I suppose define it like that in the header file (could I?) and then modify it as needed in the cpp file, but I'd rather not. I've tried doing:

item (2, std::vector<float> (2,0))

but it doesn't like that. So if anyone knows how to do this it'd be really helpful.
J
Bazzy (4111)   Link to this post
When you define a size for a vector you are initializing it, a vector can have any size, it isn't fixed as an array is
j3lps (11)   Link to this post
OK... but for some reason I still can't get this to work... If I'm doing something really stupid (which is likely) please tell me so...

I defined the vector in the header file as:
std::vector< std::vector< float > > item;
And the used this (inside a function) to initialise it:
1
2
3
4
 
int height = 4;
int width = 2;
item.resize(height);
for(int i = 0; i < height; i++) item[i].resize(width, 1.0f);


This compiles fine, but when I run the program the command prompt crashes with:
Debug Assertion Failed!
Expression: vector subscript out of range.

Any ideas?
J
Last edited on
j3lps (11)   Link to this post
Well I've found a way of doing it:

1
2
3
4
5
6
7
8
 
int height = 4;
for (int i=0;i<height;i++)
{
std::vector<float> row;
row.push_back(0.0);
row.push_back(1.0);
C_N.push_back(row);
}


I have no idea how this works, but as it does I'm not going to complain.
Thanks for your help Bazzy.
Duoas (3501)   Link to this post
You were close before on your resizing, but line 3 was off since it didn't assign the new values. Try something like this:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
 
typedef std::vector <float> row_t;
typedef std::vector <row_t> matrix_t;

void resize( matrix_t& m, unsigned rows, unsigned cols, float value = 0.0 )
  {
  // first, resize all the available columns
  unsigned min = (m.size() < rows) ? m.size() : rows;

  for (unsigned row = 0; row < min; row++)
    {
    m[ row ].resize( cols, value );
    }

  // next, resize the rows -- adding complete new columns if necessary
  m.resize( rows, row_t( cols, value ) );
  }

Hope this helps.
gi0rgi0ne (1)   Link to this post
This is the way that I work with vector of vectors:
1
2
3
4
5
6
7
8
 
int num_of_col = 5;
int num_of_row = 9;
double init_value = 3.14;

vector< vector<double> > matrix;
//now we have an empty 2D-matrix of size (0,0). Resizing it with one single command:
matrix.resize( num_of col , vector<double>( num_of_row , init_value ) );
// and we are good to go ...


simple and short.
Cheers,
G

This topic is archived - New replies not allowed.