-----------------------------------------------------------------------
G2 datasets creation
-----------------------------------------------------------------------

The datasets include two Gaussian normal distributions:

Dataset name:    G2-dim-sd
Centroid 1:      [500,500, ...]
Centroid 2:      [600,600, ...]
Dimensions:      dim = 1,2,4,8,16, ... 1024
St.Dev:          sd  = 10,20,30,40 ... 100

They have been created using the following C-language code:

Calculate random value in (0,1]:

  U = (double)(rand()+1)/(double)(RAND_MAX+1);
  V = (double)(rand()+1)/(double)(RAND_MAX+1);

Box-Muller method to create two independent standard 
one-dimensional Gaussian samples:

  X = sqrt(-2*log(U))*cos(2*3.14159*V);  /* pi = 3.14159 */  
  Y = sqrt(-2*log(U))*sin(2*3.14159*V);

Adjust mean and deviation:

  X_final = 500 + s * X;    /* mean + deviation * X */
  Y_final = 600 + s * Y;

The points are stored in the files so that:
- First 1024 points are from the cluster 1
- Rest  1024 points are from the cluster 2

-----------------------------------------------------------------------