plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x = np.array([40,60,80,100])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(10,6))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

plt.plot(x,A,"*",color="green",label="SA",linewidth=1.5,linestyle="-")

plt.plot(x,B,"p",color="blue",label="K-Medoids",linewidth=1.5,linestyle="--")

plt.plot(x,C,"s",color="red",label="GB",linewidth=1.5,linestyle="-")

plt.plot(x,D,"<",color="black",label="PageRank",linewidth=1.5,linestyle="-.")

group_labels=['40','60','80','100'] #x轴刻度的标识

plt.xticks(x,group_labels,fontsize=12) #默认字体大小为10

plt.yticks(fontsize=12)

plt.title(r"$\epsilon$=%.4f,$\alpha$=%.1f" % (epsilon,alpha),fontsize=12) #默认字体大小为12

plt.xlabel("the size of k",fontsize=13)

plt.ylabel(r"Representativeness($R$(S))",fontsize=13)

plt.xlim(40,100) #设置x轴的范围

# plt.ylim(23000,46000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

plt.legend(loc=0,ncol=2,fancybox=True)

# plt.legend(loc=0, numpoints=1)

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12) #设置图例字体的大小和粗细

plt.savefig('%s-%s.pdf' % (str(epsilon).replace(".","-"),str(alpha).replace(".","-")),format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x = np.array([1,2,3])

# x= np.array([1,2,3,4])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(10,6))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

plt.plot(x,A,"*",color="green",label="SA",linewidth=1.5,linestyle="-")

plt.plot(x,B,"p",color="blue",label="K-Medoids",linewidth=1.5,linestyle="--")

plt.plot(x,C,"s",color="red",label="GB",linewidth=1.5,linestyle="-")

plt.plot(x,D,"<",color="black",label="PageRank",linewidth=1.5,linestyle="-.")

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

# group_labels=['40','60','80','100'] #x轴刻度的标识

group_labels=['0.1555','0.1556','0.1560'] #x轴刻度的标识

plt.xticks(x,group_labels,fontsize=12) #默认字体大小为10

plt.yticks(fontsize=12)

# plt.title(r"$\epsilon$=%.4f" % epsilon,fontsize=12,fontweight='bold') #默认字体大小为12

plt.title(r"Accuracy of Classifying($\alpha=%.1f,k=100$)" % alpha,fontsize=13) #默认字体大小为12

plt.xlabel(r"$\epsilon$",fontsize=20)

plt.ylabel("Accuracy",fontsize=13)

# plt.xlim(1,4) #设置x轴的范围

plt.xlim(1,3) #设置x轴的范围

# plt.ylim(23000,46000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

plt.legend(loc=0,ncol=2,fancybox=True)

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12) #设置图例字体的大小和粗细

# plt.savefig('%.4fAccuracy.pdf' % epsilon,format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.savefig('Accuracy%.1f.pdf' % alpha,format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x= np.array([1,2,3,4])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(10,6))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

# ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

plt.plot(x,A,"*",color="green",label="SA",linewidth=1.5,linestyle="-")

plt.plot(x,B,"p",color="blue",label="K-Medoids",linewidth=1.5,linestyle="--")

plt.plot(x,C,"s",color="red",label="GB",linewidth=1.5,linestyle="-")

plt.plot(x,D,"<",color="black",label="PageRank",linewidth=1.5,linestyle="-.")

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

group_labels=['40','60','80','100'] #x轴刻度的标识

plt.xticks(x,group_labels,fontsize=12) #默认字体大小为10

plt.yticks(fontsize=12)

plt.title(r"$\epsilon=%.4f,\alpha=%.1f$" % (epsilon,alpha),fontsize=13)

plt.xlabel(r"the number of $k$",fontsize=13)

plt.ylabel("Accuracy",fontsize=13)

plt.xlim(1,4) #设置x轴的范围

# plt.ylim(23000,46000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

plt.legend(loc=0,ncol=2,fancybox=True)

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12) #设置图例字体的大小和粗细

plt.savefig('%.4f-%.1fAccuracy.pdf' % (epsilon,alpha),format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x = np.array([1,2,3,4])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(10,6))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

plt.plot(x,A,"v",color="blue",label=r"$%s(\epsilon=0.1555)$" % method,linewidth=1.5,linestyle="-.")

# plt.plot(x,A)

plt.plot(x,B,">",color="green",label="$%s(\epsilon=0.1556)$" % method,linewidth=1.5,linestyle="--")

# plt.plot(x,B)

plt.plot(x,C,"s",color="red",label="$%s(\epsilon=0.1560)$" % method,linewidth=1.5,linestyle="-")

# plt.plot(x,C)

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

group_labels=['40','60','80','100'] #x轴刻度的标识

plt.xticks(x,group_labels,fontsize=12) #默认字体大小为10

plt.yticks(fontsize=12)

plt.title(r"$\alpha$=0.6,%s under different $\epsilon$" % method,fontsize=12) #默认字体大小为12

plt.xlabel("the size of k",fontsize=13)

plt.ylabel(r"Representativeness($R$(S))",fontsize=13)

plt.xlim(1,4) #设置x轴的范围

plt.legend(loc=0, numpoints=1)

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12,fontweight='bold') #设置图例字体的大小和粗细

plt.savefig('%s.pdf' % method,format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x = np.array([1,2,3,4,5,6,7,8,9])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(12,5))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

width = 0.25

rects1 = ax.bar(x+ width,categories.values(),width,color="g")

# plt.bar(x,A,color="blue",label="original domain distirbution")

# # plt.plot(x,A)

# plt.bar(x,B,color="red",label="domain distribution of GB\K-Medoids\SA")

# # plt.plot(x,B)

# plt.bar(x,C,color="green",label="domain distribution of PageRank")

# plt.plot(x,C)

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

group_labels = categories.keys()

plt.xticks(x + width * 1.5,group_labels,fontsize=10) #默认字体大小为10

plt.yticks(fontsize=12)

plt.title("Domain Distribution",fontsize=12) #默认字体大小为12

for a,b in zip([c for c in x],categories.values()):

plt.text(a + width * 1.5,b + 0.08,b,ha="center",va="bottom",fontsize=12)

plt.xlabel("domains",fontsize=13)

plt.ylabel("the number of users",fontsize=13)

# plt.xlim(1,9) #设置x轴的范围

# plt.ylim(43800,48000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

# leg = plt.gca().get_legend()

# ltext = leg.get_texts()

# plt.setp(ltext, fontsize=12,fontweight='bold') #设置图例字体的大小和粗细

plt.savefig('OriginalDistribution.pdf',format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

x = np.array([1,2,3,4,5,6,7,8,9])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(12,5))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

width = 0.25

rects1 = ax.bar(x,data[0,:],width,color="y",hatch='/')

rects2 = ax.bar(x + width,data[1,:],width,color="r",hatch="+")

rects3 = ax.bar(x + 2 * width,data[2,:],width,color="b",hatch="-")

# plt.bar(x,A,color="blue",label="original domain distirbution")

# # plt.plot(x,A)

# plt.bar(x,B,color="red",label="domain distribution of GB\K-Medoids\SA")

# # plt.plot(x,B)

# plt.bar(x,C,color="green",label="domain distribution of PageRank")

# plt.plot(x,C)

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

group_labels=['Politics','Religion','Military','Education','Economy','Technology','Agriculture','Sports','Entertainment'] #x轴刻度的标识

plt.xticks(x + 3 * width / 2,group_labels,fontsize=10) #默认字体大小为10

plt.yticks(fontsize=12)

plt.title("Domain Distribution",fontsize=12) #默认字体大小为12

plt.xlabel("domains",fontsize=13)

plt.ylabel("proportion",fontsize=13)

# plt.xlim(1,9) #设置x轴的范围

# plt.ylim(43800,48000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

plt.legend((rects1[0],rects2[0],rects3[0]),('original','GB\K-Medoids\SA','PageRank'))

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12) #设置图例字体的大小和粗细

plt.savefig('Distribution.pdf',format='pdf') #建议保存为svg格式，再用inkscape转为矢量图emf后插入word中

features = datapre.Features()

categories = ['Politics','Religion','Military','Education','Economy','Technology','Agriculture','Sports','Entertainment']

number = [0 for i in range(len(categories))]

for profile in profiles:

for i in range(len(categories)):

if features[profile][5] == categories[i]:

number[i] += 1

plt.rcParams['font.sans-serif']=['Arial'] #如果要显示中文字体，则在此处设为：SimHei

plt.rcParams['axes.unicode_minus']=False #显示负号

# x = np.array([1,2,3])

x= np.array([1,2,3,4])

#label在图示(legend)中显示。若为数学公式，则最好在字符串前后添加"$"符号

#color：b:blue、g:green、r:red、c:cyan、m:magenta、y:yellow、k:black、w:white、、、

#线型：- -- -. : ,

#marker：. , o v < * + 1

plt.figure(figsize=(8,5))

# plt.grid(linestyle = "--") #设置背景网格线为虚线

ax = plt.gca()

# ax.spines['top'].set_visible(False) #去掉上边框

# ax.spines['right'].set_visible(False) #去掉右边框

plt.plot(x,A,"*",color="green",label="SA",linewidth=1.5,linestyle="-")

plt.plot(x,B,"p",color="blue",label="K-Medoids",linewidth=1.5,linestyle="--")

plt.plot(x,C,"s",color="red",label="GB",linewidth=1.5,linestyle="-")

plt.plot(x,D,"<",color="black",label="PageRank",linewidth=1.5,linestyle="-.")

# plt.plot(x,D,"--",label="PageRank",linewidth=1.5)

# plt.plot(x,C,"g^")

group_labels=['40','60','80','100'] #x轴刻度的标识

# group_labels=['0.1555','0.1556','0.1560'] #x轴刻度的标识

plt.xticks(x,group_labels,fontsize=12,fontweight='bold') #默认字体大小为10

plt.yticks(fontsize=12,fontweight='bold')

plt.title(r"$\epsilon$=%.4f" % epsilon,fontsize=12,fontweight='bold') #默认字体大小为12

# plt.title("Accuracy of Classifying",fontsize=12,fontweight='bold') #默认字体大小为12

plt.xlabel("the number of k",fontsize=13,fontweight='bold')

plt.ylabel("Time(s)",fontsize=13,fontweight='bold')

plt.xlim(1,4) #设置x轴的范围

# plt.ylim(23000,46000)

#plt.ylim(0.5,1)

#plt.legend() #显示各曲线的图例

plt.legend(loc=0, numpoints=1)

leg = plt.gca().get_legend()

ltext = leg.get_texts()

plt.setp(ltext, fontsize=12,fontweight='bold') #设置图例字体的大小和粗细

plt.savefig('Time-%.4fAccuracy.pdf' % epsilon,format='pdf')

# alpha = 0.4

# 绘制代表性结果

# A = [[39317.670,39790.252,40366.054,40831.374],[38832.139,39350.993,40399.873,40845.153],[38832.139,39405.829,40409.946,40873.602]]

# B = [[44592.917,46209.091,46786.843,47914.210],[44368.051,46369.859,47042.774,47801.642],[44982.317,46167.736,47300.916,48172.758]]

# C = [[46097.602,48321.697,49677.016,50906.232],[46313.745,48547.211,50096.669,51234.068],[46390.040,48623.506,50181.684,51303.911]]

# D = [[26283.041,27082.409,31429.368,32147.902],[26283.041,27082.409,31429.368,32147.902],[26283.041,27082.409,31429.368,32147.902]]

# PaintTotalRepre(A,B,C,D,0.4)

# 绘制准确率结果

# A = [[0.315,0.307,0.303,0.296],[0.332,0.338,0.320,0.320],[0.332,0.336,0.327,0.327]]

# B = [[0.321,0.342,0.319,0.336],[0.331,0.355,0.358,0.361],[0.327,0.351,0.366,0.362]]

# C = [[0.351,0.362,0.379,0.389],[0.365,0.369,0.390,0.402],[0.365,0.369,0.390,0.403]]

# D = [[0.176,0.215,0.230,0.239],[0.176,0.215,0.230,0.239],[0.176,0.215,0.230,0.239]]

# PaintTotalAccuracy(A,B,C,D,0.4)

# A = [0.296,0.320,0.327]

# B = [0.336,0.361,0.362]

# C = [0.389,0.402,0.403]

# D = [0.239,0.239,0.239]

# PaintAccuracyAgainstEp(A,B,C,D,0.4)

# alpha = 0.6

# 0.1560

# A = np.array([37584.537, 38015.656, 38307.095, 38468.860])

# B = np.array([43507.348, 44922.403, 45761.403, 46271.064])

# C = np.array([44219.653, 45963.595, 47108.449, 47929.333])

# D = np.array([25592.255, 25970.675, 29950.690, 30489.681])

# 0.1556

# A = np.array([37584.537,38015.656,38307.095,38468.860])

# B = np.array([43541.673,45055.524,45616.712,46301.794])

# C = np.array([44174.361,45932.558,47044.142,47882.450])

# D = np.array([25592.255, 25970.675, 29950.690, 30489.681])

# 0.1555

# A = np.array([37583.323,37975.945,38289.485,38436.601])

# B = np.array([43229.790,44540.790,45283.024,46057.038])

# C = np.array([43953.424,45642.373,46879.324,47653.545])

# D = np.array([25592.255, 25970.675, 29950.690, 30489.681])

# #

# PaintRepre(A,B,C,D,0.1560,0.6)

# 绘制分类准确性(against epsilon)

# A = [0.319,0.340,0.342]

# B = [0.338,0.368,0.366]

# C = [0.387,0.406,0.407]

# D = [0.239,0.239,0.239]

# PaintAccuracyAgainstEp(A,B,C,D,0.6)

# 绘制GB在不同阈值下的结果

# A = [43953.424, 45642.373, 46879.324, 47653.545]

# B = [44174.361, 45932.558, 47044.142, 47882.450]

# C = [44219.653, 45963.595, 47108.449, 47929.333]

# 绘制K-Medoids在不同阈值下的结果

# A = np.array([43229.790, 44540.790, 45283.024, 46057.038])

# B = np.array([43541.673, 45055.524, 45616.712, 46301.794])

# C = np.array([43507.348, 44922.403, 45761.403, 46271.064])

# 绘制SA在不同阈值下的结果

# A = np.array([37583.323, 37975.945, 38289.485, 38436.601])

# B = np.array([37584.537, 38015.656, 38307.095, 38468.860])

# C = np.array([37584.537, 38015.656, 38307.095, 38468.860])

# PaintEpsilon(A,B,C,"SA")

# 绘制数据集领域分布

# DomainDistribution(datapre.CategoriesDistribution())

# 绘制分类准确性(against k)

# epsilon = 0.1555

# A = [0.320,0.321,0.310,0.319]

# B = [0.318,0.350,0.342,0.339]

# C = [0.334,0.368,0.377,0.387]

# D = [0.176,0.215,0.230,0.239]

# epsilon = 0.1556

# A = [0.332,0.322,0.333,0.340]

# B = [0.343,0.352,0.372,0.368]

# C = [0.360,0.389,0.400,0.406]

# D = [0.176,0.215,0.230,0.239]

#

# # epsilon = 0.1560

# A = [0.332,0.322,0.335,0.342]

# B = [0.349,0.362,0.359,0.374]

# C = [0.360,0.386,0.401,0.407]

# D = [0.176,0.215,0.230,0.239]

#

# PaintAccuracyAgainstK(A,B,C,D,0.1555)

# 绘制原集中领域分布

# DomainDistribution(datapre.CategoriesUsers())

# 绘制领域分布图

# 'Politics','Religion','Military','Education','Economy','Technology','Agriculture','Sports','Entertainment'

# A = np.array([11934,1888,1299,969,3553,12640,2558,13217,12013])

# A = A * 1.0 / A.sum(axis=0)

# B = A

# with open("/home/duncan/InfluenceTop100","rb") as f:

# lines = f.readlines()

# profiles = set()

# for line in lines:

# profiles.add(line.split(" ")[0])

# C = np.array(ProfileDomainDistribution(profiles))

# C = C * 1.0 / C.sum(axis=0)

# data = np.vstack((A,B,C))

# DomainsDistribution(data)

# 绘制时间复杂度

# epsilon = 0.1555

# A = [8.119,8.418,8.176,8.376]

# B = [224.052,200.843,180.963,139.895]

# C = [182.477,1735.569,1494.784,11457.288]

# D = [3.92,3.92,3.92,3.92]

# epsilon = 0.1556

# A = [8.217,8.417,8.251,9.878]

# B = [289.346,179.231,151.791,127.804]

# C = [40.096,154.016,183.975,483.792]

# D = [3.92,3.92,3.92,3.92]

# epsilon = 0.1560

# A = [8.275,8.970,8.478,8.426]

# B = [196.425,221.924,131.916,138.491]

# C = [40.243,58.980,71.586,88.046]

# D = [3.92,3.92,3.92,3.92]

# TimeComplexityAgainstK(A,B,C,D,0.1560)

# 绘制准确率

# A = [[0.298,0.298,0.291,0.296],[0.336,0.337,0.320,0.326],[0.337,0.341,0.321,0.325]]

# B = [[0.341,0.337,0.355,0.338],[0.318,0.336,0.336,0.361],[0.341,0.361,0.353,0.368]]

# C = [[0.343,0.346,0.377,0.382],[0.347,0.365,0.383,0.400],[0.347,0.365,0.383,0.403]]

# D = [[0.176,0.215,0.230,0.239],[0.176,0.215,0.230,0.239],[0.176,0.215,0.230,0.239]]

# PaintTotalAccuracy(A,B,C,D,0.8)

A = [0.296,0.326,0.325]

B = [0.338,0.361,0.368]

C = [0.382,0.400,0.403]

D = [0.239,0.239,0.239]

features = datapre.Features()

epsilons = [0.1560,0.1556,0.1555]

number = [40,60,80,100]

for epsilon in epsilons:

for n in number:

start_time = time.time()

profile1 = greedy.Greedy(n,features,datapre.CategoriesDistribution(),epsilon).SearchWithReplace()

end_time = time.time()

with open("GB%d_%.4f" % (n,epsilon),"wb") as f:

f.write("cost %f s" % (end_time - start_time))

f.write("Attribute Representativeness is:")

f.write(str(metric.AttributeRepresentative(features,profile1)))

f.write("\n")

for profile in profile1:

f.write(profile + "\t")

start_time = time.time()

profile2 = kmedoids.KMedoids(n,features,datapre.CategoriesDistribution(),epsilon).Search()

end_time = time.time()

with open("kmedoids%d_%.4f" % (n,epsilon),"wb") as f:

f.write("cost %f s" % (end_time - start_time))

f.write("Attribute Representativeness is:")

f.write(str(metric.AttributeRepresentative(features,profile2)))

f.write("\n")

for profile in profile2:

f.write(profile + "\t")

start_time = time.time()

profile3 = sa.SAalgo(n,features,datapre.CategoriesDistribution(),epsilon,0.3,10,0.9).Search()

end_time = time.time()

with open("sa%d_%.4f" % (n,epsilon),"wb") as f:

f.write("cost %f s" % (end_time - start_time))

f.write("Attribute Representativeness is:")

f.write(str(metric.AttributeRepresentative(features,profile3)))

f.write("\n")

for profile in profile3: