数据科学与机器学习案例之客户的信用风险与预测

数据科学与机器学习之信用卡欺诈识别（严重类失衡数据建模）

数据科学与机器学习案例之汽车目标客户销售策略研究

数据简介

可供建模的训练集与测试集共有140多万条记录，我们对文本数据进行了转换，开发了文本处理的函数，并进行了程序调试。
汇总之后的数据共有以下字段：时间、WiFi地址、X、Y、方向、移动设备地址、信号值、信号频率、距离

详细的数据处理、数据汇总、可视化、模型、预测已在思维导图中说明。以下就不在单独详细介绍。

可视化图集

不同位置的记录数

可视化图集1

可视化图集2

可视化图集3

可视化图集4

代码汇总

> library(ggplot2)
> library(caret)
> library(data.table)
> library(parallel)
> library(lubridate)
> 
> txt = readLines('训练.txt')
> 
> head(txt,6)
> 
> process.data <- function(x){
+        a = strsplit(x,'[,;=]')[[1]]
+        mat = matrix(a[-(1:10)],ncol = 4,byrow = T)
+        result = cbind(matrix(a[c(2,4,6,7,8,10)],nrow = nrow(mat),ncol = 6
+                       ,byrow = T),mat)
+        result}
> 
> process.data(txt[14])
> 
> a1 = lapply(txt[4:20],process.data)
> 
> results = lapply(txt[substr(txt,1,1) != '#'],process.data)
> 
> options(error = recover,warn = 2)
> 
> process.data <- function(x){
+        a = strsplit(x,'[,;=]')[[1]]
+        if(length(a) == 10) return(NULL)
+        mat = matrix(a[-(1:10)],ncol = 4,byrow = T)
+        result = cbind(matrix(a[c(2,4,6,7,8,10)],nrow = nrow(mat),
+                   ncol = 6,byrow = T),mat)
+        result}
> 
> results = lapply(txt[substr(txt,1,1) != '#'],process.data)
> 
> options(error = recover,warn = 1)
> 
> results = as.data.frame(do.call('rbind',results),stringsAsFactors = FALSE)
> 
> colnames(results) = c('时间','WiFi地址','X','Y','Z','方向',
+                        '移动设备地址','信号值','信号频率','模式')
> 
> results[c('时间','X','Y','Z','方向','信号值')] = lapply(results[c('时间','X','Y','Z','方向','信号值')],as.numeric)
> 
# 删除类型变量
> 
> results = results[results$模式 == '3',]
> results  = results[,'模式' != names(results)]
> 
> results$时间 = results$时间 / 1000
> class(results$时间) = c('POSIXT','POSIXct')
> 
> start = c(seq(1,length(results$时间),by = 10000))
> end = c(start[2:length(start)]-1,length(results$时间))
> 
> for(i in 1:length(start)){
+      sink('Time.txt',append = T)
+      print(results$时间[start[i]:end[i]])
+      sink() }
> 
> Time = readLines('Time.txt')
> strsplit(Time[1],'[\\"]')[[1]][c(2,4)]
> 
> BI = function(x){
+         a = strsplit(x,'[\\"]')[[1]][c(2,4)]
+         a = gsub(' CST','',a)
+         a}
> time = unlist(lapply(Time,BI))
> time = time[-length(time)]
> 
> results$时间 = time
>
> results$时间 = as.POSIXct(results$时间)
> results$时间 <- update(results$时间, year=2021)
> 
> summary(results[,c('时间','X','Y','Z','方向','信号值')])
> 
> 
> table(results$WiFi地址)
> 
> sort(table(results$移动设备地址),T)[1:6]
> names(sort(table(results$移动设备地址),T)[1:6])
> 
> histogram(results$方向)
> 
> opar = par(mfrow = c(2,1))
> plot(density(results$方向))
> plot(ecdf(results$方向))
> par(opar)
> 
> angle.res <- function(x){
+         angles = seq(0,by = 45,length = 9)
+         a = sapply(x,function(y){
+                         which.min(abs(y - angles))})
+         c(angles[1:8],0)[a]
+         }
> 
> results$方向 = angle.res(results$方向)
> 
> barplot(table(results$方向),names.arg = names(table(results$方向)),
+         col = 'orange',  # 填充颜色
+         border = '#ffffff')
> 
> c(length(unique(results$移动设备地址)),length(unique(results$信号频率)))
> 
> t = sort(table(results$移动设备地址),T)
> 
> results = results[which(results$移动设备地址 %in% names(t)[1:8]),]
> 
# 检查移动设备地址与频率之间是否是一一对应关系
> 
> data = with(results,table(移动设备地址,信号频率))
> 
> apply(data,2,function(x) sum(x > 0))
> 
> temp = temp[temp$移动设备地址 == '00:0f:a3:39:e0:4b' & temp$信号频率 == '2462000000',]
> 
> results = results[results$移动设备地址 != '00:0f:a3:39:e0:4b' & results$信号频率 != '2462000000',]
> 
> location = with(results,by(results,list(X,Y),function(x) x))
> 
> (sapply(location,class))
> sum(sapply(location,class) == 'NULL')
> location = location[! (sapply(location,class) == 'NULL')]
> length(location)
> 
> sapply(location,class)
> sapply(location,nrow) # 每个坐标下的记录
> 
> 
> lp = sapply(location,function(x) c(unique(x$X),unique(x$Y)))
> lp <- rbind(lp,sapply(location,nrow))
> 
> plot(t(lp),type = 'n',xlab = '',ylab = '')
> text(t(lp),labels = t(lp)[,3],cex = .8,srt = 45)
> 
> lattice::bwplot(信号值 ~ factor(方向) | 移动设备地址,data = results,
 +                subset = X == 2 & Y == 12 & 移动设备地址 != '00:0f:a3:39:dd:cd',
 +                layout = c(2,3)) # 信号强度的分布
> 
> lattice::densityplot(~信号值 | 移动设备地址 + factor(方向),data = results,
+                     subset = X == 24 & Y == 4 & 
+                          移动设备地址 != "00:0f:a3:39:dd:cd",
+              bw = 0.5, plot.points = FALSE)
> 
> results$XY = paste(results$X,results$Y,sep = '-')
> 
> XYanglelocation = with(results,by(results,list(XY,方向,移动设备地址),function(x)x))
> 
> 
> temp1 = lapply(XYanglelocation,function(x){
+                  and = x[1,]
+                  and$Mean = mean(x$信号值)
+                  and$Median = median(x$信号值)
+                  and$IQR = IQR(x$信号值)
+                  and$num = length(x$信号值)
+                  and$sd = sd(x$信号值)
+                  and}) # 汇总
> temp1 = do.call('rbind',temp1)
> 
> 
> breaks = seq(-90,-25,by = 5)
> lattice::bwplot(sd ~ cut(Mean,breaks = breaks),data = temp1,
+                 xlab = 'Mean',ylab = 'Sd')
> 
> with(temp1,smoothScatter((Mean - Median) ~ num,
+                        xlab = 'Num',ylab = 'Mean - Median'))
> abline(h = 0,col = '#984ea3',lwd = 2)
> 
> loess.and = with(temp1,loess(diff ~ num ,data = data.frame(diff = Mean - Median,
+                                       num = num)))
> 
> loess.pre = predict(loess.and,newdata = data.frame(num = 60:120))
> 
> with(temp1,smoothScatter((Mean - Median) ~ num,
+                xlab = 'Num',ylab = 'Mean - Median'))
> abline(h = 0,col = '#984ea3',lwd = 2)
> lines(x = 60:120,y = loess.pre)
> points(x = 60:120,y = loess.pre)
> 
> onelocationangle = subset(temp1,
+      移动设备地址 == '00:14:bf:b1:97:8d' & 方向 == 0)
> 
> smooth1 = fields::Tps(onelocationangle[,c('X','Y')],onelocationangle$Mean)
> 
> smooth1 = fields::predictSurface(smooth1)
> fields::plot.surface(smooth1,type = 'C')
> points(x = onelocationangle[,'X'],y = onelocationangle[,'Y'],
+        pch = 19,cex = .5) # 地形图
> 
# 构造函数
> 
> surface = function(data,location,angle = 45){
+              library(fields)
+        onelocationangle = subset(data,
+            data$移动设备地址 == location & data$方向 == angle)
+        smooth1 = Tps(onelocationangle[,c('X','Y')],onelocationangle$Mean)
+        smooth1 = predictSurface(smooth1)
+        plot.surface(smooth1,type = 'C')
+        points(x = onelocationangle[,'X'],y = onelocationangle[,'Y'],
+        pch = 19,cex = .5)    }
> 
> surface(temp1,'00:14:bf:b1:97:90')
> 
> opar = par(mfrow = c(2,2),mar = rep(1,4))
> mapply(surface,data = list(temp1),
+        location = sample(unique(temp1$移动设备地址),4),
+        angle = rep(c(45,135),each = 2)) # 地形图
> par(opar)
> 
> WiFi.matrix = matrix(c( 7.5, 6.3, 2.5, -.8, 12.8, -2.8,  
+                 1, 14, 33.5, 9.3,  33.5, 2.8),
+                ncol = 2,byrow = T,
+         dimnames = list(unique(results$移动设备地址),c('x','y')))
> 
> diffs = results[,c('X','Y')] - WiFi.matrix[results$移动设备地址,]
> 
> results$distance = sqrt(diffs[,1] ^2 + diffs[,2]^2)
> 
> lattice::xyplot(信号值 ~ distance | factor(移动设备地址) + factor(方向),
+                 data = results,pch = 19,cex = .5,xlab = 'Distance')
> 
> lattice::xyplot(信号值 ~ distance | factor(移动设备地址) + factor(方向),
+                 data = results,
+              subset = 移动设备地址 == c("00:14:bf:b1:97:8a","00:14:bf:b1:97:90")&
+                       方向 == c(45,135),pch = 19,cex = .5)
> 
> txt1 = readLines('测试.txt')
> 
> options(error = recover,warn = 2)
> process.data = function(x){
+        a = strsplit(x,'[,;=]')[[1]]
+        if(length(a) <= 10) return(NULL)
+        mat = matrix(a[-(1:10)],ncol = 4,byrow = T)
+        result = cbind(matrix(a[c(2,4,6,7,8,10)],nrow = nrow(mat),
+                   ncol = 6,byrow = T),mat)
+        result}

> test = lapply(txt1,process.data)
> test = as.data.frame(do.call('rbind',test))
> 
> 
> colnames(test) = c('时间','WiFi地址','X','Y','Z','方向',
+                        '移动设备地址','信号值','信号频率','模式')
> 
> test[c('时间','X','Y','Z','方向','信号值')] = lapply(test[c('时间','X','Y','Z','方向','信号值')],as.numeric)
> 
# 删除类型变量
> 
> test = test[test$模式 == '3',]
> test = test[,'模式' != names(test)]
> 
> test$时间 = test$时间 / 1000
> class(test$时间) = c('POSIXT','POSIXct')
> 
> start = c(seq(1,length(test$时间),by = 10000))
> end = c(start[2:length(start)]-1,length(test$时间))
> 
> for(i in 1:length(start)){
+      sink('Time1.txt',append = T)
+      print(results$时间[start[i]:end[i]])
+      sink() }
> 
> Time1 = readLines('Time1.txt')
> strsplit(Time[1],'[\\"]')[[1]][c(2,4)]
> 
> BI = function(x){
+         a = strsplit(x,'[\\"]')[[1]][c(2,4)]
+         a = gsub(' CST','',a)
+         a}
> time1 = unlist(lapply(Time1,BI))
> time1 = time1[-length(time1)]
> 
> test$时间 = time1
> 
> test$时间 = as.POSIXct(test$时间)
> test$时间 <- update(test$时间, year=2021)
> 
> summary(test[,c('时间','X','Y','Z','方向','信号值')])
> 
> 
> angle.res <- function(x){
+         angles = seq(0,by = 45,length = 9)
+         a = sapply(x,function(y){
+                         which.min(abs(y - angles))})
+         c(angles[1:8],0)[a]
+         }
> 
> test$方向 = angle.res(test$方向)
> 
> class(with(test,table(移动设备地址,信号频率)))
> 
> apply(with(test,table(移动设备地址,信号频率)),2,function(x) sum(x > 0))
> 
# 修改
>  
> 
> test$移动设备地址[which(test$移动设备地址 == '00:0f:a3:39:e0:4b' &
+                  test$信号频率 == '2462000000')] = '00:0f:a3:39:e1:c0'
> test$移动设备地址[which(test$移动设备地址 == '00:0f:a3:39:e2:10' &
+                   test$信号频率 == '2437000000')] = '00:14:bf:b1:97:8a'
> 
> test = test[which(test$移动设备地址 %in% unique(results$移动设备地址)),]
> 
> test$XY = paste(test$X,test$Y,sep = '-')
> 
> with(test,table(XY,方向))
> 
> variables = c('X','Y','方向','XY')
> test.XY = with(test,by(test,list(XY),function(x) {
+                  as = x[1,variables]
+                  Mean = tapply(x$信号值,x$移动设备地址,mean)
+                  x = matrix(Mean,nrow = 1,ncol = 6,dimnames = list(
+                          as$XY,names(Mean)))
+                  cbind(as,x)  }))
> test.XY = do.call('rbind',test.XY)
> 
# 根据角度以及m选择训练数据集的数据
> 
> names(test.XY)
> m = 3
> angleNewobs = 230
> rfs = seq(0,by = 45,length = 8)
> nearangle = angle.res(angleNewobs)
> 
> if(m %% 2 == 1){# 选择奇数个近邻
+     angles = seq(-45 * (m-1) / 2,45 * (m-1) / 2,length = m)
+ } else { # 选择偶数个近邻
+     m = m+1
+     angles = seq(-45 * (m-1) / 2,45 * (m-1) / 2,length = m)
+     if(sign(angleNewobs-nearangle) > -1){
+      angles = angles[-1]}
+      else {
+       angles = angles[-m]}
+                  }
> angles = angles + nearangle
> angles[angles < 0] = angles[angles < 0] + 360
> angles[angles > 360] = angles[angles > 360] - 360
> 
> 
> resize = function(data,var = 'Mean',keep = c('X','Y','XY')){
+            ans = with(data,by(data,list(XY),function(x){
+                         as = x[1,keep]
+                         y = tapply(x[,var],x$移动设备地址,mean)
+                         A = matrix(y,nrow = 1,ncol = 6,
+                                 dimnames = list(as$XY,
+                                   names(y)))
+                         cbind(as,A)}))
+             newdata = do.call('rbind',ans)
+             return(newdata)}
> 
# 根据角度选择训练集
> 
> select.train = function(angleNewobs,data,m = 1){ 
+          rfs = seq(0,by = 45,length = 8)
+        nearangle = angle.res(angleNewobs)
+ 
+       if(m %% 2 == 1){# 选择奇数个近邻
+        angles = seq(-45 * (m-1) / 2,45 * (m-1) / 2,length = m)
+        } else { # 选择偶数个近邻
+          m = m+1
+         angles = seq(-45 * (m-1) / 2,45 * (m-1) / 2,length = m)
+        if(sign(angleNewobs-nearangle) > -1){
+           angles = angles[-1]}
+        else {
+           angles = angles[-m]}
+                  }
+        angles = angles + nearangle
+        angles[angles < 0] = angles[angles < 0] + 360
+        angles[angles > 360] = angles[angles > 360] - 360
+         
+        angles = sort(angles)
+        train.subset = data[data$方向 %in% angles,]
+        resize(train.subset)  }
>       
> train.1 = select.train(130,temp1,m=1)
> train.2 = select.train(130,temp1,m=2)
> 
> findnn = function(newsignal,train.subset){
+           diffs = apply(train.subset[,5:10],1,function(x)
+                               x - newsignal)
+           distance = apply(diffs,2,function(x) sqrt(sum(x)^2))
+           closest = order(distance)
+           return(train.subset[closest,1:3])    } # 按照距离的长短排列(k近邻)
> 
> pred.XY = function(newSignals, newAngles, trainData, 
+                   numAngles = 1, k = 3){
+   
+    closeXY = list(length = nrow(newSignals))
+   
+   for (i in 1:nrow(newSignals)) {
+     trainSS = select.train(newAngles[i], trainData, m = numAngles) # 筛选不同的训练集
+     closeXY[[i]] = 
+       findnn(newsignal = as.numeric(newSignals[i, ]), trainSS) # 储存数据框
+                                  }
+    estXY = lapply(closeXY, 
+                  function(x) sapply(x[ , 1:2], 
+                                     function(x) mean(x[1:k]))) # 按照第k个近邻计算
+   estXY = do.call("rbind", estXY)
+   return(estXY)
+                            }
> 
> final1.1 = pred.XY(test.XY[,5:10],test.XY[,3],temp1) # 角度数1,3近邻
> 
> plot(x = test.XY[,1],y = test.XY[,2],type = 'n',xlim = range(test.XY[,1]),
+      ylim = range(test.XY[,2]),xlab = 'X',ylab = 'Y')
> points(x = test.XY[,1],y = test.XY[,2],col = 'black',pch = 16)
> points(x = final1.1[,1],y = final1.1[,2],col = 'red',pch = 16)
> 
> 
> A = vector(mode = 'list',length = 25)
> df = vector(mode = 'numeric',length = 25)
> error.cal = function(Matrix){
+                res = sum(rowSums((Matrix - test.XY[,c('X','Y')])^2))
+                res}
> 
> knn.grid = expand.grid(m = 1:5,k = 1:5)
> 
> for(i in 1:length(A)){
+        A[[i]] = pred.XY(test.XY[,5:10],test.XY[,3],temp1,knn.grid[i,1],
+                             knn.grid[i,2]) 
+        df[i] = error.cal(A[[i]]) }
>  
> a1 = pred.XY(test.XY[,5:10],test.XY[,3],temp1,knn.grid[1,1],knn.grid[1,2])
> 
> barplot(df)
> knn.grid[which.min(df),]
> 
# 可视化
> 
> plot(x = test.XY[,'X'],y = test.XY[,'Y'],xlim = range(test.XY[,'X']),
+      ylim = range(test.XY[,'Y']),xlab = 'X',ylab = 'Y',type = 'n')
> box()
> points(x = test.XY[,'X'],y = test.XY[,'Y'],pch = 16,col = 'black')
> points(x = A[[14]][,1],y = A[[14]][,2],pch = 17,col = 'red')
> segments(x0 = test.XY[,'X'],y0 = test.XY[,'Y'],
+          x1 = A[[14]][,1],y1 = A[[14]][,2],lwd = 1,lty = 2,
+          col = 'grey')
> legend('topright',c('原始位置','最优预测位置m=4 k=3'),
+        pch = c(16,17),col = c('black','red'))
> 
# CV交叉验证
> 
> resize = function(data,keep = c('X','Y','XY','方向'),
+                   angles = seq(0,by = 45,length = 8),
+                   sampleangle = FALSE){
+ 
+              res = with(data,by(data,list(XY),function(x){
+                       if(sampleangle){
+                          x = x[x$方向 == sample(angles,1),]}
+                       as = x[1,keep]
+                       y = tapply(x$信号值,x$移动设备地址,mean)
+                       A = matrix(y,nrow = 1,ncol = length(y),
+                                  dimnames = list(as$XY,names(y)))
+                       cbind(as,A)         } ))
+               newdata = do.call('rbind',res)
+               return(newdata) }
> 
> v = 10
> sample.location = sample(unique(temp1$XY))
> sample.location = matrix(sample.location,ncol = v,
+                     nrow = floor(length(sample.location) / v))
> 
> test.cv = resize(results,sampleangle = TRUE)
> 
> knn.grid = expand.grid(m = 1:5,k = 1:10)
> B = matrix(NA,nrow = nrow(knn.grid),ncol = 10)
> error.cal = function(Matrix,data){
+             res = sum(rowSums((Matrix - data)^2))
+              res}
> 
> for(i in 1:v){
+      test.fold = subset(test.cv,XY %in% sample.location[,i])
+      train.fold = subset(temp1,XY %in% sample.location[,-i])
+      actual.fold = test.fold[,c('X','Y')]
+      for(j in 1:nrow(B)){
+           temp = pred.XY(test.fold[,5:10],test.fold[,4],
+                            train.fold,knn.grid[j,1],knn.grid[j,2] )
+           B[j,i] = error.cal(temp,actual.fold)}
+        }
# 可视化
> y = apply(B,1,mean)
> col <- RColorBrewer::brewer.pal(8,'Accent')
> plot(x = 1:10,y = c(1:10),type = 'n',xlim = c(1,10),
       ylim = range(y),xlab = 'k近邻个数',ylab = '误差')
> box()
> axis(side = 1,at = 1:10)
> lines(x = 1:10,y = y[1:10],lty = 2,lwd = 1,col = '#7FC97F')
> lines(x = 1:10,y = y[11:20],lty = 2,lwd = 1,col = '#BEAED4')
> lines(x = 1:10,y = y[21:30],lty = 2,lwd = 1,col = '#FDC086')
> lines(x = 1:10,y = y[31:40],lty = 2,lwd = 1,col = '#FFFF99')
> lines(x = 1:10,y = y[41:50],lty = 2,lwd = 1,col = '#386CB0')
> points(x = 1:10,y = y[1:10],pch = 17,col = 'red')
> points(x = 1:10,y = y[11:20],pch = 17,col = 'red')
> points(x = 1:10,y = y[21:30],pch = 17,col = 'red')
> points(x = 1:10,y = y[31:40],pch = 17,col = 'red')
> points(x = 1:10,y = y[41:50],pch = 17,col = 'red')
> legend('bottomright',c('m=1','m=2','m=3','m=4','m=5'),
+                  lty = rep(2,5),lwd = rep(1,5),
+                  col = col[1:5])
> 
> final1 = pred.XY(newSignals = test.XY[,5:10],newAngles = test.XY[,3],
+                  trainData = temp1,numAngles = 1, k = 10)
> 
> err1 = error.cal(final1,test.XY[,c('X','Y')])
> 
# 加权k近邻
> 
> findnn = function(newsignal,train.subset){
+           diffs = apply(train.subset[,5:10],1,function(x)
+                               x - newsignal)
+           distance = apply(diffs,2,function(x) sqrt(sum(x)^2))
+           train.subset$distance = round(1/distance,5)
+           closest = order(distance)
+           return(train.subset[closest,c(1:3,11)])    } # 按照距离的长短排列(k近邻)
> 
> trainSS = select.train(130, temp1, m = 2)
> 
> pred.XY = function(newSignals, newAngles, trainData, 
+                   numAngles = 1, k = 3){
+   
+   closeXY = list(length = nrow(newSignals))
+   
+   for (i in 1:nrow(newSignals)) {
+     trainSS = select.train(newAngles[i], trainData, m = numAngles) # 筛选不同的训练集
 +    closeXY[[i]] = 
+       findnn(newsignal = as.numeric(newSignals[i, ]), trainSS) # 储存数据框
+                                  }
+    estXY = lapply(closeXY,function(x) {
+                  weights = x$distance[1:k] / sum(x$distance[1:k])
+                  sapply(x[,1:2],function(x) sum(x[1:k] * weights))}) # 按照第k个近邻计算
+   estXY = do.call("rbind", estXY)
+   return(estXY)
+                            }
> 
> C = matrix(NA,nrow = nrow(knn.grid),ncol = 10)
> error.cal = function(Matrix,data){
+               res = sum(rowSums((Matrix - data)^2))
+               res}
> 
> for(i in 1:v){
+      test.fold = subset(test.cv,XY %in% sample.location[,i])
+      train.fold = subset(temp1,XY %in% sample.location[,-i])
+      actual.fold = test.fold[,c('X','Y')]
+      for(j in 1:nrow(C)){
+           temp = pred.XY(test.fold[,5:10],test.fold[,4],
+                            train.fold,knn.grid[j,1],knn.grid[j,2] )
+           C[j,i] = error.cal(temp,actual.fold)}
+        }
> 
# 可视化
> y = apply(C,1,mean)
> col <- RColorBrewer::brewer.pal(8,'Accent')
> plot(x = 1:10,y = c(1:10),type = 'n',xlim = c(1,10),
+        ylim = range(y),xlab = '加权k近邻个数',ylab = '误差')
> box()
> axis(side = 1,at = 1:10)
> lines(x = 1:10,y = y[1:10],lty = 2,lwd = 1,col = '#7FC97F')
> lines(x = 1:10,y = y[11:20],lty = 2,lwd = 1,col = '#BEAED4')
> lines(x = 1:10,y = y[21:30],lty = 2,lwd = 1,col = '#FDC086')
> lines(x = 1:10,y = y[31:40],lty = 2,lwd = 1,col = '#FFFF99')
> lines(x = 1:10,y = y[41:50],lty = 2,lwd = 1,col = '#386CB0')
> points(x = 1:10,y = y[1:10],pch = 17,col = 'red')
> points(x = 1:10,y = y[11:20],pch = 17,col = 'red')
> points(x = 1:10,y = y[21:30],pch = 17,col = 'red')
> points(x = 1:10,y = y[31:40],pch = 17,col = 'red')
> points(x = 1:10,y = y[41:50],pch = 17,col = 'red')
> legend('topright',c('m=1','m=2','m=3','m=4','m=5'),
+                  lty = rep(2,5),lwd = rep(1,5),
+                  col = col[1:5])
> which.min(rowSums(C))
> final2 = pred.XY(newSignals = test.XY[,5:10],newAngles = test.XY[,3],
+                  trainData = temp1,numAngles = 1, k = 10)
> 
> err2 = error.cal(final2,test.XY[,c('X','Y')])

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576

以上就是本篇博客的全部内容，欢迎各位小伙伴批评指正，您的建议将是我不断更新的动力！谢谢！