Question

Say I have 10 points all on a sphere (randomly distributed) and I want to rotate the entire system to make sure one point sits at the north pole. How would I do this using c++?

I went about it by looking at 3D rotation matrices:

http://en.wikipedia.org/wiki/Rotation_matrix

I rotate my point around the x-axis until the y component is zero and then rotate around the y-axis until the x component is zero. This should leave the point in question at either the north or south pole right?

My code looks like so:

#include <stdio.h> 
#include <string.h>
#include <math.h>
#include <iostream>
#include <iomanip>
#include <fstream>
#include <time.h>
#include <stdlib.h>
#include <sstream>
using namespace std;
#define PI 3.14159265358979323846

int main()
{

  int a,b,c,f,i,j,k,m,n,s;
  double p,Time,Averagetime,Energy,energy,Distance,Length,DotProdForce,Forcemagnitude,
         ForceMagnitude[101],Force[101][4],E[1000001],En[501],x[101][4],y[101][4],
         Dist[101][101],Epsilon,z[101],theta,phi;

    /*  set the number of points */
    n=10;

    /* check that there are no more than 100 points */
    if(n>100){
      cout << n << " is too many points for me :-( \n";
      exit(0);
    }

    /* reset the random number generator */
    srand((unsigned)time(0));  

    for (i=1;i<=n;i++){
      x[i][1]=((rand()*1.0)/(1.0*RAND_MAX)-0.5)*2.0;
      x[i][2]=((rand()*1.0)/(1.0*RAND_MAX)-0.5)*2.0;
      x[i][3]=((rand()*1.0)/(1.0*RAND_MAX)-0.5)*2.0;

      Length=sqrt(pow(x[i][1],2)+pow(x[i][2],2)+pow(x[i][3],2));

      for (k=1;k<=3;k++){
        x[i][k]=x[i][k]/Length;
      }
    }

    /* calculate the energy */
    Energy=0.0;

    for(i=1;i<=n;i++){
      for(j=i+1;j<=n;j++){
        Distance=sqrt(pow(x[i][1]-x[j][1],2)+pow(x[i][2]-x[j][2],2)
                    +pow(x[i][3]-x[j][3],2));

        Energy=Energy+1.0/Distance;
      }
    }

   cout << fixed << setprecision(10) << "energy=" << Energy << "\n";  

  /* Save Values so far */

  for(i=1;i<=n;i++){
    for(j=1;j<=3;j++){
      y[i][j]=x[i][j];
    }
  }

  /* Choose each point in turn and make it the north pole note this is what the while loop os for, but have set it to a<2 to just look at first point */

  a=1;
  b=0;
  c=0;

  while(a<2){

  /* Find theta and phi to rotate points by */

  cout << fixed << setprecision(5) << "x[" << a << "][1]=" << x[a][1] << 
  " x[" << a << "][2]=" << x[a][2] << " x[" << a << "][3]=" << x[a][3] << "\n";

  theta=x[a][2]/x[a][3];
  theta=b*PI+atan(theta);

  /* Rotate Points by theta around x axis and then by phi around y axis */

  for(i=1;i<=n;i++){
    x[i][1]=x[i][1];
    x[i][2]=x[i][2]*cos(theta)-x[i][3]*sin(theta);
    x[i][3]=x[i][2]*sin(theta)+x[i][3]*cos(theta);
  }

  phi=x[a][1]/x[a][3];
  phi=c*PI+atan(phi);

  for(i=1;i<=n;i++){
    x[i][1]=x[i][1]*cos(phi)-x[i][3]*sin(phi);
    x[i][2]=x[i][2];
    x[i][3]=x[i][1]*sin(phi)+x[i][3]*cos(phi);
  }

  cout << fixed << setprecision(5) << "x[" << a << "][1]=" << x[a][1] << 
  " x[" << a << "][2]=" << x[a][2] << " x[" << a << "][3]=" << x[a][3] << "\n";

   if(x[a][3]==1.0 && x[a][2]==x[a][3]==0)
    a=a+1;
  else if(b==0 && c==0)
    for(i=1;i<=n;i++){
      for(j=1;j<=3;j++){
        x[i][j]=y[i][j];
        c=1;
      }
    }
  else if(b==0 && c==1)
    for(i=1;i<=n;i++){
      for(j=1;j<=3;j++){
        x[i][j]=y[i][j];
        b=1;
        c=0;
      }
    }
  else if(b==1 && c==0)
    for(i=1;i<=n;i++){
      for(j=1;j<=3;j++){
        x[i][j]=y[i][j];
        c=1;
      }
    }
  else if(b==1 && c==1)
    break;

  }

  energy=0.0;

  for(i=1;i<=n;i++){
    for(j=i+1;j<=n;j++){

      Distance=sqrt(pow(x[i][1]-x[j][1],2)+pow(x[i][2]-x[j][2],2)
                    +pow(x[i][3]-x[j][3],2));
      energy=energy+1.0/Distance;
    }
  }  

  cout << fixed << setprecision(10) << "ENERGY=" << energy << "\n";  
  cout << fixed << setprecision(5) << "x[" << a << "][1]=" << x[a][1] << 
  " x[" << a << "][2]=" << x[a][2] << " x[" << a << "][3]=" << x[a][3] << "\n";

  /* Output to help with gnuin.txt */
  ofstream File4 ("mypoints");
  for(i=1;i<=n;i++){
    File4 << x[i][1] << " " <<   x[i][2] << " " << x[i][3] << "\n";
  }
  File4.close(); 

  return 0;

}

Ok, so there are loads of issues here, like the if statement (line 103) shouldn’t really have a condition for equal to a double as it will never work, but I can sort that out later using indirect comparison (some epsilon stuff). My real query is why does the rotation even though it is acting on all the points take the points off the sphere? (As you can see the values have been normalised to make them all on the unit sphere in line 38).

Note: the b, c stuff is to check whether the point is at the north or south pole.

Answer 1

You have a problem with your rotation code. For example:

x[i][1]=x[i][1];
x[i][2]=x[i][2]*cos(theta)-x[i][3]*sin(theta);
x[i][3]=x[i][2]*sin(theta)+x[i][3]*cos(theta);

You modify x[i][2] in the 2nd line, then use it in the 3rd. You should use temporary storage for intermediate results, to avoid modifying values before you’re finished referencing them.

The first line is fairly redundant, the rest should look more like:

double new_y, new_z;
new_y=x[i][2]*cos(theta)-x[i][3]*sin(theta);
new_z=x[i][2]*sin(theta)+x[i][3]*cos(theta);
x[i][2] = new_y;
x[i][3] = new_z;

(and obviously do that where you perform such a calculation)

A better way to orient your sphere might be to calculate a transformation matrix in the same fashion as a ‘look-at’ matrix. In a look-at matrix, the frame is rotated to align some vector with the z-axis. In your case you probably want to align along the y-axis, but the principle is the same.

I’d also comment that you seem to be ignoring the 0th element in your arrays… IMHO this is a bad habit – you should get used to the fact that arrays start at zero. Sooner or later you’ll either get the indexing wrong, or you will have to interface to someone else’s code.