public IList <Cluster> SearchClusters(double radius, IList <Dot> dots)
{
int i = 0;
double range = int.MaxValue;//radius * radius;
Dot[] copy = new Dot[dots.Count];
dots.CopyTo(copy, 0);
var dotsToCluster = copy.ToList();
List <Cluster> clusters = new List <Cluster>(dots.Count);
Dot middlePoint = default;
while (dotsToCluster.Count > 0)
{
middlePoint = dotsToCluster.OrderBy(d => d.GetDistance(middlePoint)).First();
var dotsInRange = dotsToCluster
.Where(d => d.GetDistance(middlePoint) <= range)
.OrderBy(d => d.GetDistance(middlePoint))
.ToList();
List <Dot> dotsInCluster = new List <Dot>(dotsInRange.Count());
if (dotsInRange.Any())
{
foreach (var inRangeDot in dotsInRange)
{
if (dotsInCluster.Count > 0)
{
middlePoint = default;
foreach (var dot in dotsInCluster)
{
middlePoint += dot;
}
middlePoint /= dotsInCluster.Count;
}
Dot displaced = (middlePoint + inRangeDot) / 2;
dotsInCluster.Add(inRangeDot);
if (dotsInCluster.Any(d => d.GetDistance(displaced) >= radius))
{
dotsInCluster.Remove(inRangeDot);
break;
}
else
{
middlePoint = displaced;
}
}
}
clusters.Add(new Cluster()
{
Dots = ColoredDots(dotsInCluster, i),
Number = i,
Radius = radius,
RadiusDot = middlePoint
});
dotsToCluster = dotsToCluster.Except(dotsInCluster).ToList();
i += 1;
}
clusters = ClusterOptimizer.RemoveSharedClusters(clusters);
return(clusters);
}
