【R】数据相关性的可视化

一千零一技|相关性分析及其可视化：copy&paste，搞定

.libPaths(c("/bioinfo/home/software/miniconda3/envs/R4.0/lib/R/library"))
#data("mtcars")
library("PerformanceAnalytics")
# pdf("test.pdf")
# my_data <- mtcars[, c(1,3,4,5,6,7)]
# print (head(my_data))
# chart.Correlation(my_data, histogram=TRUE, pch=19)
# dev.off()

args    <- commandArgs(trailingOnly = TRUE)
infile  <- args[1]
outdir <- args[2]
names<-basename(infile)
df <- read.delim(infile, header = T, stringsAsFactors = F,row.names = NULL)
#df <- df[,-ncol(df)]
print (df)
index <- row.names(df)
print(index)
# q()
# data_T=as.data.frame((data))
# df1 = anno_col[,"types",drop=FALSE]
pdf(paste0(outdir,"/",names,".correlation_chart.pdf"))
chart.Correlation(df, histogram=TRUE, pch=19,method='spearman')
dev.off()

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

输入：

输出：

热图

输入文件

library("pheatmap")


#args <- commandArgs(trailingOnly = TRUE)
#clusternum <- args[1]
#outdir <- args[2]

pdf(file.path('heatmap.pdf'))
data_frame <- read.table("heatmapFile.xls",sep="	",header=T,row.names = 1)
data_frame_sample <- data_frame[, 1:(ncol(data_frame)-1)]
groups1 <- data_frame[,c('clinic_label'),drop=FALSE]
p<-pheatmap(data_frame_sample,
            cluster_row = T,
            cluster_col=F,
            main="sample Heatmap",
            show_rownames = F,
            show_colnames=F,
            vmax = 0.6,
            #display_numbers = TRUE,
            annotation_row = groups1) #, cutree_cols = clusternum) fontsize_col=4
#col_cluster <- cutree(p$tree_col, k=clusternum)
#col_cluster = as.data.frame(col_cluster)
#print(col_cluster)
#write.csv(col_cluster, sep="\t", quote = FALSE, file.path(outdir, "cluster.csv"))
dev.off()

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