2 replies to this topic

[PGP_Protector]

Right now I'm using this to sort my data points for a quick sort for a collection of points.

   public class MapOptimization

    {

        private double[,] MappingPoints;

        private Point[] MapPoints;

        private int PointCount;

        public int LoadPoints(double[,] DataSet)

        {

            PointCount = DataSet.GetLength(0);

            MapPoints = new Point[PointCount];

            MappingPoints = new double[PointCount, 2];

            for (int i = 0; i < PointCount; i++)

            {

                MapPoints[i].X = Math.Round(DataSet[i, 0], 5);     // Round off to 5 Digits

                MapPoints[i].Y = Math.Round(DataSet[i, 1], 5);

            }

            // Move DataSet to MappingPoints, Checking each Point For Duplicates.

            // Return How Many Valid Points Loaded

            return PointCount;

        }

        public double[,] Optimize()

        {

            var sortedPoints = MapPoints.OrderBy(p => p.X).ThenBy(p => p.Y);


            int i = 0;

            foreach (Point TempPoint in sortedPoints)

            {

                MappingPoints[i, 0] = TempPoint.X;

                MappingPoints[i, 1] = TempPoint.Y;

                i++;

            }

            return MappingPoints;

        }

    }

But I am looking for help in truly optimizing the path.

An Example map would be something like this


What's the best way of determining the optimal path to stop at each point at least once with minimal movement?

[PGP_Protector]

Ok, looks like I need a form of the Traveling Salesman Algorithm, off to Wiki :D

[fayyazlodhi]

yeah but unless i am mistaken about your problem, if your problem can be limited to two dimensions of possible travel, it translates into unidirectional TSP which is solvable using Dynamic programming creating a 2D map of the possible paths and choosing the optimum one.

I am not able to figure out if that is the case with your problem or not.