Add files via upload

dbscan/dbscan.py

@@ -0,0 +1,94 @@
+import random
+import matplotlib.pyplot as plot
+import math
+
+# dbscan algorithm - O(n^2)
+
+def euclidean_distance(point1, point2):
+    distance = math.sqrt((point2[0] - point1[0])**2 + (point2[1] - point1[1])**2)
+    return distance
+
+def dbscan(points,eps,min_points):
+	cluster_id = 1
+	# -1 means unclassified point
+	clusters = [-1 for i in range(len(points))]
+	for i,point in enumerate(points):
+		if clusters[i] == -1:
+			if expand_cluster(points,(i,point),cluster_id,eps,min_points,clusters):
+				cluster_id=cluster_id+1
+
+	return clusters
+
+
+def expand_cluster(points,current_point,cluster_id,eps,min_points,clusters):
+	seeds = region_query(points,current_point,eps,clusters)
+
+	if len(seeds) < min_points:
+		# 0 means noise point
+		clusters[point[0]] = 0
+		return False
+
+	else:
+		for i,point in seeds:
+			clusters[i]=cluster_id
+		seeds.remove(current_point)
+
+		while len(seeds)>0:
+			current_point = seeds.pop(0)
+			result = region_query(points,current_point,eps,clusters)
+
+			if len(result) >= min_points:
+				for i,point in result:
+					if clusters[i] == -1:
+						seeds.append((i,point))
+					clusters[i]=cluster_id
+
+		return True
+
+
+def region_query(points,current_point,eps,clusters):
+	seeds = []
+	for i,point in enumerate(points):
+		if euclidean_distance(current_point[1],point) < eps:
+			seeds.append((i,point))
+
+	return seeds;
+
+
+# for testing
+if __name__ == "__main__":
+    points = []
+
+    for i in range(50):
+        x = random.randint(10,20)
+        y = random.randint(10,20)
+        points.append((x,y))
+
+    for i in range(50):
+        x = random.randint(40,50)
+        y = random.randint(40,50)
+        points.append((x,y))
+
+
+    for i in range(50):
+        x = random.randint(50,60)
+        y = random.randint(20,30)
+        points.append((x,y))
+
+
+    eps=5
+    min_points=10
+
+    result = dbscan(points,eps,min_points)
+
+    colors = ["yellow","green","red","blue"]
+
+    for i,point in enumerate(points):
+        plot.scatter(point[0],point[1], color=colors[result[i]])
+
+    plot.xlabel('X-axis')
+    plot.ylabel('Y-axis')
+    plot.title('K Mean')
+
+    plot.show()
+