DBSCAN.py

import numpy as np
import collections
import matplotlib.pyplot as plt
import queue
import scipy.io as spio

#Define label for differnt point group
NOISE = 0
UNASSIGNED = 0
core=-1
edge=-2



#function to find all neigbor points in radius
def neighbor_points(data, pointId, radius):
    points = []
    for i in range(len(data)):
        #Euclidian distance using L2 Norm
        if np.linalg.norm(data[i] - data[pointId]) <= radius:
            points.append(i)
    return points

#DB Scan algorithom
def dbscan(data, Eps, MinPt):
    #initilize all pointlable to unassign
    pointlabel  = [UNASSIGNED] * len(data)
    pointcount = []
    #initilize list for core/noncore point
    corepoint=[]
    noncore=[]
    
    #Find all neigbor for all point
    for i in range(len(data)):
        pointcount.append(neighbor_points(train,i,Eps))
    
    #Find all core point, edgepoint and noise
    for i in range(len(pointcount)):
        if (len(pointcount[i])>=MinPt):
            pointlabel[i]=core
            corepoint.append(i)
        else:
            noncore.append(i)

    for i in noncore:
        for j in pointcount[i]:
            if j in corepoint:
                pointlabel[i]=edge

                break
            
    #start assigning point to luster
    cl = 1
    #Using a Queue to put all neigbor core point in queue and find neigboir's neigbor
    for i in range(len(pointlabel)):
        q = queue.Queue()
        if (pointlabel[i] == core):
            pointlabel[i] = cl
            for x in pointcount[i]:
                if(pointlabel[x]==core):
                    q.put(x)
                    pointlabel[x]=cl
                elif(pointlabel[x]==edge):
                    pointlabel[x]=cl
            #Stop when all point in Queue has been checked   
            while not q.empty():
                neighbors = pointcount[q.get()]
                for y in neighbors:
                    if (pointlabel[y]==core):
                        pointlabel[y]=cl
                        q.put(y)
                    if (pointlabel[y]==edge):
                        pointlabel[y]=cl            
            cl=cl+1 #move to next cluster
           
    return pointlabel,cl
    
#Function to plot final result
def plotRes(data, clusterRes, clusterNum):
    nPoints = len(data)
    scatterColors = ['black', 'green', 'brown', 'red', 'purple', 'orange', 'yellow']
    for i in range(clusterNum):
        if (i==0):
            #Plot all noise point as blue
            color='blue'
        else:
            color = scatterColors[i % len(scatterColors)]
        x1 = [];  y1 = []
        for j in range(nPoints):
            if clusterRes[j] == i:
                x1.append(data[j, 0])
                y1.append(data[j, 1])
        plt.scatter(x1, y1, c=color, alpha=1, marker='.')


#Load Data
raw = spio.loadmat('DBSCAN.mat')
train = raw['Points']

#Set EPS and Minpoint
epss = [5,10]
minptss = [5,10]
# Find ALl cluster, outliers in different setting and print resultsw
for eps in epss:
    for minpts in minptss:
        print('Set eps = ' +str(eps)+ ', Minpoints = '+str(minpts))
        pointlabel,cl = dbscan(train,eps,minpts)
        plotRes(train, pointlabel, cl)
        plt.show()
        print('number of cluster found: ' + str(cl-1))
        counter=collections.Counter(pointlabel)
        print(counter)
        outliers  = pointlabel.count(0)
        print('numbrer of outliers found: '+str(outliers) +'\n')

Hello, can you share the DBSCAN.mat file? it is missing from the source files. Thank you

Please share the DBSCAN.mat