SEMtree/tree_sim.R
In fernandoPalluzzi/SEMgraph: Network Analysis and Causal Inference Through Structural Equation Modeling
#load libraries
rm(list = ls(all = TRUE)); graphics.off(); cat("\014")
#setwd("SEMtree_code_data")
library(SEMgraph)
library(org.Hs.eg.db)
library(EnrichmentBrowser)
library(COSINE)
library(BioNet)
library(pathfindR)
library(diffusr)
library(reshape)
library(ggplot2)
library(viridis)
library(dplyr)
source("./code/help_tree.R")
options(warn = -1)
# COSINE simulation
network<- as.undirected(kegg)
seed_vec <- c(1:100)
set_vec <- c(1:4)
nset <- 4; nmet <- 7
results <- data.frame(iter = rep(1:100, each = nset*nmet),
set = rep(rep(1:nset, each=nmet),100),
FG = rep(c(rep(50,3*nmet), rep(40, nmet)), 100),
method = rep(c("BioNet", "COSINE", "pathfindeR",
"ST", "STr2z", "WalktrapGM", "WalktrapGMl"),400),
size = rep(NA,nset*nmet*100), precision = rep(NA,nset*nmet*100),
recall = rep(NA,nset*nmet*100), f1 = rep(NA,nset*nmet*100))
n = 1
for (j in seed_vec){
set.seed(j)
simulated_data <- generate_sim(mu = 0.75)
for (set in set_vec){
cat("\n Iteration", j, "- Set", set, "\n")
if(set == 1) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[2]]))
if(set == 2) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[3]]))
if(set == 3) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[5]]))
if(set == 4) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[6]]))
colnames(SIMdata) <- c(rep("c",20), rep("d", 20))
rownames(SIMdata) <- V(network)$name[1:nrow(SIMdata)]
group <- c(rep(0,20), rep(1,20))
if(set == 4) FG<- rownames(SIMdata)[1:40]
if(set != 4) FG<- rownames(SIMdata)[1:50]
y<- group; x<- SIMdata
write.table(x, file="xdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
write.table(cbind(colnames(x),y), file="cdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
write.table(rownames(x), file="rdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
exprs.file <- file.path(getwd(), "xdat.tab")
cdat.file <- file.path(getwd(), "cdat.tab")
rdat.file <- file.path(getwd(), "rdat.tab")
SIMdata <- EnrichmentBrowser::readSE(exprs.file, cdat.file, rdat.file)
SIMdata <- deAna(SIMdata)
data<- t(assay(SIMdata))
# network <- subNetwork(rownames(SIMdata), gHumanNet)
cat("Bionet...")
pval <- rowData(SIMdata)$ADJ.PVAL
names(pval) <- rownames(SIMdata)
fb <- fitBumModel(pval, plot=FALSE)
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
scores<- ifelse(names(pval) %in% term, 1, -1)
names(scores)<- names(pval)
module <- runFastHeinz(network, scores)
seed <- V(module)$name
if (is.null(seed) | length(seed) == 1){
results$size[n] <- 0
results$precision[n] <- NA
results$recall[n] <- NA
results$f1[n] <- NA
} else{
results$size[n] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n] <- round(pre,3)
results$recall[n] <- round(rec,3)
results$f1[n] <- round(f1,3)
}
cat("COSINE...")
PPI<- igraph::as_data_frame(network)
genes<- rownames(assay(SIMdata)) %in% V(network)$name
SIMdatat<- t(assay(SIMdata)[genes,])
control_data <- SIMdatat[1:20,]
case_data <- SIMdatat[21:40,]
scaled_diff <- quiet(diff_gen_PPI(case_data,control_data,PPI))
scaled_node_score <- scaled_diff[[1]]
scaled_edge_score <- scaled_diff[[2]]
lambda<- 0.5
num_iter<- 100 #300
GA_result<- quiet(GA_search_PPI(lambda,scaled_node_score,scaled_edge_score, PPI,
num_iter, muCh=0.05, zToR=10, minsize=10))
results$size[n+1] <- GA_result$Subnet_size
seedID <- GA_result$Subnet
seed<- V(network)$name[V(network) %in% seedID]
if (is.null(seed) | length(seed) ==1){
results$size[n+1] <- 0
results$precision[n+1] <- NA
results$recall[n+1] <- NA
results$f1[n+1] <- NA
} else{
results$size[n+1] <- length(seed)
int<- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+1] <- round(pre,3)
results$recall[n+1] <- round(rec,3)
results$f1[n+1] <- round(f1,3)
}
cat("pathfindR...")
RA_input <- rowData(SIMdata)[, c(1,2,5)]
Symbol<- unlist(mget(RA_input[,1], org.Hs.egSYMBOL, ifnotfound=NA))
RA_input[,1] <- Symbol
RA_input <- data.frame(RA_input)
colnames(RA_input) <- c("GENE", "CHANGE", "P_VALUE")
# greedy search (GR)
snws_file <- paste0("active_snws_GR", n+2)
snws_tot <- quiet(active_snw_search(input_for_search = RA_input,
pin_name_path = "KEGG",
snws_file = snws_file,
score_quan_thr = 0.5,
sig_gene_thr = 0.01,
search_method = "GR",
silent_option = TRUE))
if(length(snws_tot) == 0){
results$size[n+2] <- 0
results$precision[n+2] <- NA
results$recall[n+2] <- NA
results$f1[n+2] <- NA
} else{
if(length(snws_tot[[1]]) == 1){
results$size[n+2] <- 0
results$precision[n+2] <- NA
results$recall[n+2] <- NA
results$f1[n+2] <- NA
} else{
seed <- snws_tot[[1]]
results$size[n+2] <- length(seed)
seed <- quiet(mapIds(org.Hs.eg.db, seed, column = 'ENTREZID', keytype = 'SYMBOL'))
seed<- as.character(na.omit(seed))
int<- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+2] <- round(pre,3)
results$recall[n+2] <- round(rec,3)
results$f1[n+2] <- round(f1,3)
}
}
cat("SEMtree...")
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
if (identical(term, character(0)) | length(term) == 1){
results$size[n+3] <- 0
results$precision[n+3] <- NA
results$recall[n+3] <- NA
results$f1[n+3] <- NA
} else{
tree<- SEMtree(network, data, seed=term, type="ST")
seed <- V(tree)$name
results$size[n+3] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+3] <- round(pre,3)
results$recall[n+3] <- round(rec,3)
results$f1[n+3] <- round(f1,3)
}
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
if (identical(term, character(0)) | length(term) == 1){
results$size[n+4] <- 0
results$precision[n+4] <- NA
results$recall[n+4] <- NA
results$f1[n+4] <- NA
} else{
if (length(term) > 5){
graph <- quiet(weightGraph(graph=network, data=data, group=group, method="r2z"))
tree<- SEMtree(graph, data=data, seed=term, type="ST", eweight = "pvalue")
seed <- V(tree)$name
results$size[n+4] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+4] <- round(pre,3)
results$recall[n+4] <- round(rec,3)
results$f1[n+4] <- round(f1,3)
} else{
results$size[n+4] <- 0
results$precision[n+4] <- NA
results$recall[n+4] <- NA
results$f1[n+4] <- NA
}
}
cat("WalktrapGM...")
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
WTC<- quiet(WalktrapGM(network, data, group))
ID <- WTC$scores$ID[1]
seed <- names(WTC$membership[WTC$membership == ID]) #length(seed) #166
if (is.null(seed) | length(seed) == 1){
results$size[n+5] <- 0
results$precision[n+5] <- NA
results$recall[n+5] <- NA
results$f1[n+5] <- NA
} else{
results$size[n+5] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+5] <- round(pre,3)
results$recall[n+5] <- round(rec,3)
results$f1[n+5] <- round(f1,3)
}
WTC<- WalktrapGMl(network, rowData(SIMdata))
ID <- WTC$scores$ID[1]
seed <- names(WTC$membership[WTC$membership == ID]) #length(seed) #166
if (is.null(seed) | length(seed) == 1){
results$size[n+6] <- 0
results$precision[n+6] <- NA
results$recall[n+6] <- NA
results$f1[n+6] <- NA
} else{
results$size[n+6] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+6] <- round(pre,3)
results$recall[n+6] <- round(rec,3)
results$f1[n+6] <- round(f1,3)
}
n = n + 7
}
}
# Summarise results
res <- results %>%
group_by(method, set) %>%
summarise(across(c(size, precision, recall, f1), .f = list(mean = mean), na.rm = TRUE))
res <- as.data.frame(res)
res_long <- melt(res, id.vars = c("method", "set"))
res_long$method[res_long$method == "WalktrapGM"] <- "WGM_RWR"
res_long$method[res_long$method == "WalktrapGMl"] <- "WGM_FC"
cols <- c("recall_mean", "precision_mean", "f1_mean")
file=paste0('all_meas.tiff')
tiff(file=file,width=8, height=5, units="in", res=300)
setEPS()
postscript(width=8, height = 6, "fig4.eps")
res_long %>%
dplyr::filter(variable %in% cols) %>%
ggplot(aes(x=method, y=value, group=as.factor(variable), fill=as.factor(variable)))+
geom_bar(position='dodge', stat='identity') +
ggtitle("") +
theme_minimal() +
ylab("") +
xlab("") +
facet_wrap(~ set, ncol = 2, scales = "free_y") +
theme(text = element_text(size = 13)) +
theme(axis.text.x = element_text(angle = 90)) +
theme(legend.title = element_text(size=9)) +
theme(legend.title=element_blank())+
scale_fill_discrete(labels=c('Precision', 'Recall', 'F-measure')) +
theme(legend.position="bottom") +
theme(legend.key.size = unit(0.3, "cm"))
dev.off()
cols <- c("size_mean")
file=paste0('size_meas.tiff')
tiff(file=file,width=8, height=5, units="in", res=300)
# setEPS()
# postscript(width=8, height = 5, "size.eps")
res_long %>%
dplyr::filter(variable %in% cols) %>%
ggplot(aes(x=method, y=value, group=as.factor(set), fill=as.factor(set)))+
geom_bar(position='dodge', stat='identity') +
ggtitle("") +
theme_minimal() +
ylab("") +
xlab("") +
theme(text = element_text(size = 16)) +
theme(axis.text.x = element_text(angle = 90)) +
scale_colour_discrete("Set") +
theme(legend.title=element_blank())+
theme(legend.position="bottom") +
theme(legend.key.size = unit(0.3, "cm"))
dev.off()
fernandoPalluzzi/SEMgraph documentation built on Feb. 20, 2025, 4:27 p.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.
#load libraries
rm(list = ls(all = TRUE)); graphics.off(); cat("\014")
#setwd("SEMtree_code_data")
library(SEMgraph)
library(org.Hs.eg.db)
library(EnrichmentBrowser)
library(COSINE)
library(BioNet)
library(pathfindR)
library(diffusr)
library(reshape)
library(ggplot2)
library(viridis)
library(dplyr)
source("./code/help_tree.R")
options(warn = -1)
# COSINE simulation
network<- as.undirected(kegg)
seed_vec <- c(1:100)
set_vec <- c(1:4)
nset <- 4; nmet <- 7
results <- data.frame(iter = rep(1:100, each = nset*nmet),
set = rep(rep(1:nset, each=nmet),100),
FG = rep(c(rep(50,3*nmet), rep(40, nmet)), 100),
method = rep(c("BioNet", "COSINE", "pathfindeR",
"ST", "STr2z", "WalktrapGM", "WalktrapGMl"),400),
size = rep(NA,nset*nmet*100), precision = rep(NA,nset*nmet*100),
recall = rep(NA,nset*nmet*100), f1 = rep(NA,nset*nmet*100))
n = 1
for (j in seed_vec){
set.seed(j)
simulated_data <- generate_sim(mu = 0.75)
for (set in set_vec){
cat("\n Iteration", j, "- Set", set, "\n")
if(set == 1) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[2]]))
if(set == 2) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[3]]))
if(set == 3) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[5]]))
if(set == 4) SIMdata <- t(rbind(simulated_data[[7]], simulated_data[[6]]))
colnames(SIMdata) <- c(rep("c",20), rep("d", 20))
rownames(SIMdata) <- V(network)$name[1:nrow(SIMdata)]
group <- c(rep(0,20), rep(1,20))
if(set == 4) FG<- rownames(SIMdata)[1:40]
if(set != 4) FG<- rownames(SIMdata)[1:50]
y<- group; x<- SIMdata
write.table(x, file="xdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
write.table(cbind(colnames(x),y), file="cdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
write.table(rownames(x), file="rdat.tab", sep="\t", quote=FALSE, col.names=FALSE, row.names=FALSE)
exprs.file <- file.path(getwd(), "xdat.tab")
cdat.file <- file.path(getwd(), "cdat.tab")
rdat.file <- file.path(getwd(), "rdat.tab")
SIMdata <- EnrichmentBrowser::readSE(exprs.file, cdat.file, rdat.file)
SIMdata <- deAna(SIMdata)
data<- t(assay(SIMdata))
# network <- subNetwork(rownames(SIMdata), gHumanNet)
cat("Bionet...")
pval <- rowData(SIMdata)$ADJ.PVAL
names(pval) <- rownames(SIMdata)
fb <- fitBumModel(pval, plot=FALSE)
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
scores<- ifelse(names(pval) %in% term, 1, -1)
names(scores)<- names(pval)
module <- runFastHeinz(network, scores)
seed <- V(module)$name
if (is.null(seed) | length(seed) == 1){
results$size[n] <- 0
results$precision[n] <- NA
results$recall[n] <- NA
results$f1[n] <- NA
} else{
results$size[n] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n] <- round(pre,3)
results$recall[n] <- round(rec,3)
results$f1[n] <- round(f1,3)
}
cat("COSINE...")
PPI<- igraph::as_data_frame(network)
genes<- rownames(assay(SIMdata)) %in% V(network)$name
SIMdatat<- t(assay(SIMdata)[genes,])
control_data <- SIMdatat[1:20,]
case_data <- SIMdatat[21:40,]
scaled_diff <- quiet(diff_gen_PPI(case_data,control_data,PPI))
scaled_node_score <- scaled_diff[[1]]
scaled_edge_score <- scaled_diff[[2]]
lambda<- 0.5
num_iter<- 100 #300
GA_result<- quiet(GA_search_PPI(lambda,scaled_node_score,scaled_edge_score, PPI,
num_iter, muCh=0.05, zToR=10, minsize=10))
results$size[n+1] <- GA_result$Subnet_size
seedID <- GA_result$Subnet
seed<- V(network)$name[V(network) %in% seedID]
if (is.null(seed) | length(seed) ==1){
results$size[n+1] <- 0
results$precision[n+1] <- NA
results$recall[n+1] <- NA
results$f1[n+1] <- NA
} else{
results$size[n+1] <- length(seed)
int<- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+1] <- round(pre,3)
results$recall[n+1] <- round(rec,3)
results$f1[n+1] <- round(f1,3)
}
cat("pathfindR...")
RA_input <- rowData(SIMdata)[, c(1,2,5)]
Symbol<- unlist(mget(RA_input[,1], org.Hs.egSYMBOL, ifnotfound=NA))
RA_input[,1] <- Symbol
RA_input <- data.frame(RA_input)
colnames(RA_input) <- c("GENE", "CHANGE", "P_VALUE")
# greedy search (GR)
snws_file <- paste0("active_snws_GR", n+2)
snws_tot <- quiet(active_snw_search(input_for_search = RA_input,
pin_name_path = "KEGG",
snws_file = snws_file,
score_quan_thr = 0.5,
sig_gene_thr = 0.01,
search_method = "GR",
silent_option = TRUE))
if(length(snws_tot) == 0){
results$size[n+2] <- 0
results$precision[n+2] <- NA
results$recall[n+2] <- NA
results$f1[n+2] <- NA
} else{
if(length(snws_tot[[1]]) == 1){
results$size[n+2] <- 0
results$precision[n+2] <- NA
results$recall[n+2] <- NA
results$f1[n+2] <- NA
} else{
seed <- snws_tot[[1]]
results$size[n+2] <- length(seed)
seed <- quiet(mapIds(org.Hs.eg.db, seed, column = 'ENTREZID', keytype = 'SYMBOL'))
seed<- as.character(na.omit(seed))
int<- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+2] <- round(pre,3)
results$recall[n+2] <- round(rec,3)
results$f1[n+2] <- round(f1,3)
}
}
cat("SEMtree...")
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
if (identical(term, character(0)) | length(term) == 1){
results$size[n+3] <- 0
results$precision[n+3] <- NA
results$recall[n+3] <- NA
results$f1[n+3] <- NA
} else{
tree<- SEMtree(network, data, seed=term, type="ST")
seed <- V(tree)$name
results$size[n+3] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+3] <- round(pre,3)
results$recall[n+3] <- round(rec,3)
results$f1[n+3] <- round(f1,3)
}
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
term <- rowData(SIMdata)$ENTREZID[rowData(SIMdata)$ADJ.PVAL < 0.05]
if (identical(term, character(0)) | length(term) == 1){
results$size[n+4] <- 0
results$precision[n+4] <- NA
results$recall[n+4] <- NA
results$f1[n+4] <- NA
} else{
if (length(term) > 5){
graph <- quiet(weightGraph(graph=network, data=data, group=group, method="r2z"))
tree<- SEMtree(graph, data=data, seed=term, type="ST", eweight = "pvalue")
seed <- V(tree)$name
results$size[n+4] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+4] <- round(pre,3)
results$recall[n+4] <- round(rec,3)
results$f1[n+4] <- round(f1,3)
} else{
results$size[n+4] <- 0
results$precision[n+4] <- NA
results$recall[n+4] <- NA
results$f1[n+4] <- NA
}
}
cat("WalktrapGM...")
data<- t(assay(SIMdata));group <- c(rep(0,20), rep(1,20))
WTC<- quiet(WalktrapGM(network, data, group))
ID <- WTC$scores$ID[1]
seed <- names(WTC$membership[WTC$membership == ID]) #length(seed) #166
if (is.null(seed) | length(seed) == 1){
results$size[n+5] <- 0
results$precision[n+5] <- NA
results$recall[n+5] <- NA
results$f1[n+5] <- NA
} else{
results$size[n+5] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+5] <- round(pre,3)
results$recall[n+5] <- round(rec,3)
results$f1[n+5] <- round(f1,3)
}
WTC<- WalktrapGMl(network, rowData(SIMdata))
ID <- WTC$scores$ID[1]
seed <- names(WTC$membership[WTC$membership == ID]) #length(seed) #166
if (is.null(seed) | length(seed) == 1){
results$size[n+6] <- 0
results$precision[n+6] <- NA
results$recall[n+6] <- NA
results$f1[n+6] <- NA
} else{
results$size[n+6] <- length(seed)
int <- length(intersect(seed, FG))
pre <- int/length(seed)
rec <- int/length(FG)
f1 <- 2*((pre*rec)/(pre+rec))
results$precision[n+6] <- round(pre,3)
results$recall[n+6] <- round(rec,3)
results$f1[n+6] <- round(f1,3)
}
n = n + 7
}
}
# Summarise results
res <- results %>%
group_by(method, set) %>%
summarise(across(c(size, precision, recall, f1), .f = list(mean = mean), na.rm = TRUE))
res <- as.data.frame(res)
res_long <- melt(res, id.vars = c("method", "set"))
res_long$method[res_long$method == "WalktrapGM"] <- "WGM_RWR"
res_long$method[res_long$method == "WalktrapGMl"] <- "WGM_FC"
cols <- c("recall_mean", "precision_mean", "f1_mean")
file=paste0('all_meas.tiff')
tiff(file=file,width=8, height=5, units="in", res=300)
setEPS()
postscript(width=8, height = 6, "fig4.eps")
res_long %>%
dplyr::filter(variable %in% cols) %>%
ggplot(aes(x=method, y=value, group=as.factor(variable), fill=as.factor(variable)))+
geom_bar(position='dodge', stat='identity') +
ggtitle("") +
theme_minimal() +
ylab("") +
xlab("") +
facet_wrap(~ set, ncol = 2, scales = "free_y") +
theme(text = element_text(size = 13)) +
theme(axis.text.x = element_text(angle = 90)) +
theme(legend.title = element_text(size=9)) +
theme(legend.title=element_blank())+
scale_fill_discrete(labels=c('Precision', 'Recall', 'F-measure')) +
theme(legend.position="bottom") +
theme(legend.key.size = unit(0.3, "cm"))
dev.off()
cols <- c("size_mean")
file=paste0('size_meas.tiff')
tiff(file=file,width=8, height=5, units="in", res=300)
# setEPS()
# postscript(width=8, height = 5, "size.eps")
res_long %>%
dplyr::filter(variable %in% cols) %>%
ggplot(aes(x=method, y=value, group=as.factor(set), fill=as.factor(set)))+
geom_bar(position='dodge', stat='identity') +
ggtitle("") +
theme_minimal() +
ylab("") +
xlab("") +
theme(text = element_text(size = 16)) +
theme(axis.text.x = element_text(angle = 90)) +
scale_colour_discrete("Set") +
theme(legend.title=element_blank())+
theme(legend.position="bottom") +
theme(legend.key.size = unit(0.3, "cm"))
dev.off()
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.