Nothing
R/MS.test.clust.R
In MSeasy: Preprocessing of Gas Chromatography-Mass Spectrometry (GC-MS) data
MS.test.clust <-
function (data_tot, nclust)
{
attach(what=NULL, name="e1")
if(missing(data_tot)){
data_tot<-tk_choose.files(default=file.path(getwd(),"*.txt"),caption="Please, select a file for MS.test.clust", multi=FALSE, filters=matrix(c("your file","*.txt"),ncol=2))
data_tot<-read.table(data_tot, header=TRUE, check.names=FALSE)
}
library(fpc)
Dunn.test <- matrix(nrow = 7, ncol = 5)
rownames(Dunn.test) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(Dunn.test) <- c("cor", "mink_1/2", "mink_1/3", "manhattan",
"euclidean")
silhouette.width <- matrix(nrow = 7, ncol = 5)
rownames(silhouette.width) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(silhouette.width) <- c("cor", "mink_1/2", "mink_1/3",
"manhattan", "euclidean")
for (i in c("average", "single", "complete", "ward", "centroid")) {
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
}
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
homogeneite <- matrix(nrow = 7, ncol = 5)
rownames(homogeneite) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(homogeneite) <- c("cor", "mink_1/2", "mink_1/3",
"manhattan", "euclidean")
k <- 1
lis <- c("average", "single", "complete", "ward", "centroid")
lisd <- c(1/2, 1/3, 1, 2)
for (i in 1:length(lis)) for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- hclust(Dis, method = lis[i])
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[i, j + 1] <- nb/nclust
}
for (i in 1:length(lis)) {
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- hclust(Dis, method = lis[i])
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[i, 1] <- nb/nclust
}
for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[6, j + 1] <- nb/nclust
}
for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- pamk(Dis, krange = nclust)
cl <- Hc$pamobject$clustering
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[7, j + 1] <- nb/nclust
}
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]), cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[6, 1] <- nb/nclust
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- pamk(Dis, krange = nclust)
cl <- Hc$pamobject$clustering
res <- as.data.frame(cbind(as.vector(data_tot[, 1]), cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[7, 1] <- nb/nclust
#creation of a directory
st <- strsplit(date(), " ")[[1]]
stBis <- strsplit(st[4], ":")[[1]]
Hour <- paste(stBis[1], stBis[2], stBis[3], sep = "-")
Date <- paste(st[1], st[2], st[3], Hour, sep = "_")
Mypath <- paste("output_MStestclust", "_", "result", Date, sep = "")
dir.create(Mypath)
par(mfrow=c(2,1))
test<-matrix(homogeneite,nrow=7)
ar<-names(homogeneite[,1])
ar2<-names(homogeneite[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Matching coefficient", main = "Matching coefficient", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
test<-matrix(silhouette.width,nrow=7)
ar<-names(silhouette.width[,1])
ar2<-names(silhouette.width[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Silhouette width", main = "Silhouette Width", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
#creation of a pdf file with graphics
pdf(paste(Mypath, "/", "Graph_MStestClust.pdf", sep = ""))
par(mfrow=c(2,1))
test<-matrix(homogeneite,nrow=7)
ar<-names(homogeneite[,1])
ar2<-names(homogeneite[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Matching coefficient", main = "Matching coefficient", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
test<-matrix(silhouette.width,nrow=7)
ar<-names(silhouette.width[,1])
ar2<-names(silhouette.width[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Silhouette width", main = "Silhouette Width", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
dev.off()
par(mfrow=c(1,1))
assign("Dunn.test",Dunn.test ,envir=as.environment("e1"))
assign("silhouette.width", silhouette.width,envir=as.environment("e1"))
assign("matching.coef",homogeneite,envir=as.environment("e1"))
print(paste("A pdf file have been generated in the path:", Mypath, cat("\n")))
}
Try the MSeasy package in your browser
Any scripts or data that you put into this service are public.
MSeasy documentation built on May 2, 2019, 10:02 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
MS.test.clust <-
function (data_tot, nclust)
{
attach(what=NULL, name="e1")
if(missing(data_tot)){
data_tot<-tk_choose.files(default=file.path(getwd(),"*.txt"),caption="Please, select a file for MS.test.clust", multi=FALSE, filters=matrix(c("your file","*.txt"),ncol=2))
data_tot<-read.table(data_tot, header=TRUE, check.names=FALSE)
}
library(fpc)
Dunn.test <- matrix(nrow = 7, ncol = 5)
rownames(Dunn.test) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(Dunn.test) <- c("cor", "mink_1/2", "mink_1/3", "manhattan",
"euclidean")
silhouette.width <- matrix(nrow = 7, ncol = 5)
rownames(silhouette.width) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(silhouette.width) <- c("cor", "mink_1/2", "mink_1/3",
"manhattan", "euclidean")
for (i in c("average", "single", "complete", "ward", "centroid")) {
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- hclust(Dis, method = i)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == i, colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == i, colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
}
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
Test_valid <- cluster.stats(Dis, cl)
Dunn.test[rownames(Dunn.test) == "diana", colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "diana", colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"cor"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"cor"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/2)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"mink_1/2"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"mink_1/2"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1/3)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"mink_1/3"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"mink_1/3"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 1)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"manhattan"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"manhattan"] <- Test_valid$avg.silwidth
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = 2)
Hc <- pamk(Dis, krange = nclust)
Test_valid <- cluster.stats(Dis, Hc$pamobject$clustering)
Dunn.test[rownames(Dunn.test) == "PAM", colnames(Dunn.test) ==
"euclidean"] <- Test_valid$dunn
silhouette.width[rownames(silhouette.width) == "PAM", colnames(silhouette.width) ==
"euclidean"] <- Test_valid$avg.silwidth
homogeneite <- matrix(nrow = 7, ncol = 5)
rownames(homogeneite) <- c("average", "single", "complete",
"ward", "centroid", "diana", "PAM")
colnames(homogeneite) <- c("cor", "mink_1/2", "mink_1/3",
"manhattan", "euclidean")
k <- 1
lis <- c("average", "single", "complete", "ward", "centroid")
lisd <- c(1/2, 1/3, 1, 2)
for (i in 1:length(lis)) for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- hclust(Dis, method = lis[i])
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[i, j + 1] <- nb/nclust
}
for (i in 1:length(lis)) {
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- hclust(Dis, method = lis[i])
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[i, 1] <- nb/nclust
}
for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[6, j + 1] <- nb/nclust
}
for (j in 1:length(lisd)) {
nb <- 0
Dis <- dist(data_tot[, 4:ncol(data_tot)], method = "minkowski",
p = lisd[j])
Hc <- pamk(Dis, krange = nclust)
cl <- Hc$pamobject$clustering
res <- as.data.frame(cbind(as.vector(data_tot[, 1]),
cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[7, j + 1] <- nb/nclust
}
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- diana(Dis, diss = TRUE)
cl <- cutree(Hc, k = nclust)
res <- as.data.frame(cbind(as.vector(data_tot[, 1]), cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[6, 1] <- nb/nclust
nb <- 0
Dis <- as.dist((1 - cor(t(data_tot[, 4:ncol(data_tot)]),
method = "spearman")))
Hc <- pamk(Dis, krange = nclust)
cl <- Hc$pamobject$clustering
res <- as.data.frame(cbind(as.vector(data_tot[, 1]), cl))
clust <- levels(res$cl)
homog_clust <- list()
for (l in 1:length(clust)) {
res_temp <- subset(res, res$cl == clust[l])
temp_mol <- levels(as.factor(as.vector(res_temp$V1)))
homog_clust[[l]] <- temp_mol
}
for (g in (1:length(homog_clust))) {
if (length(homog_clust[[g]]) == 1)
nb <- nb + 1
}
homogeneite[7, 1] <- nb/nclust
#creation of a directory
st <- strsplit(date(), " ")[[1]]
stBis <- strsplit(st[4], ":")[[1]]
Hour <- paste(stBis[1], stBis[2], stBis[3], sep = "-")
Date <- paste(st[1], st[2], st[3], Hour, sep = "_")
Mypath <- paste("output_MStestclust", "_", "result", Date, sep = "")
dir.create(Mypath)
par(mfrow=c(2,1))
test<-matrix(homogeneite,nrow=7)
ar<-names(homogeneite[,1])
ar2<-names(homogeneite[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Matching coefficient", main = "Matching coefficient", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
test<-matrix(silhouette.width,nrow=7)
ar<-names(silhouette.width[,1])
ar2<-names(silhouette.width[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Silhouette width", main = "Silhouette Width", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
#creation of a pdf file with graphics
pdf(paste(Mypath, "/", "Graph_MStestClust.pdf", sep = ""))
par(mfrow=c(2,1))
test<-matrix(homogeneite,nrow=7)
ar<-names(homogeneite[,1])
ar2<-names(homogeneite[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Matching coefficient", main = "Matching coefficient", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
test<-matrix(silhouette.width,nrow=7)
ar<-names(silhouette.width[,1])
ar2<-names(silhouette.width[1,])
matplot( test, type = "b", lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), cex = 0.8, col = c("red", "blue","black","green","orange"), xlab = NULL, ylab = "Silhouette width", main = "Silhouette Width", axes=FALSE)
axis(1,seq(7), labels=ar, las=2)
axis(2)
legend(cex=0.5,"bottomright",legend=ar2, col=c("red", "blue","black","green","orange"), lty = c("dotted", "solid"), lwd = 2, pch = c(1, 2,3), bg="white")
dev.off()
par(mfrow=c(1,1))
assign("Dunn.test",Dunn.test ,envir=as.environment("e1"))
assign("silhouette.width", silhouette.width,envir=as.environment("e1"))
assign("matching.coef",homogeneite,envir=as.environment("e1"))
print(paste("A pdf file have been generated in the path:", Mypath, cat("\n")))
}
Try the MSeasy package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.