Question

我想通过矩阵中的值的wilcoxon符号秩检验（配对数据）来计算p值

t.matrix.12x<-read.csv("test.csv", header= TRUE, sep = ",", 
row.names = 1,check.names = FALSE)
group1<-t.matrix.12x[,c(1:24)]# defined first group
group2<-t.matrix.12x[,c(25:48)] # defined second group
p_value<-sapply(1:nrow(t.matrix.12x), function(i){
  wilcox.test(x=as.numeric(group2[i,]),y= 
as.numeric(group1[i,]),paired = TRUE, exact = FALSE, correct = 
FALSE)$p.value
})
p_value_data<-cbind(t.matrix.12x,p_value)
#write.csv(p_value_data,'z.csv')

现在，我想显示与上述相同的两个不同组中的数据分布。我从test.csv中取出一些值，并做一个较小的矩阵。然后我在ggplot中将其处理为：

library(ggplot2)
comparison_1<-read.csv("input_WRS1.csv", header= TRUE, sep = ",", check.names = FALSE)
merged_comparison_1<-melt(comparison_1,id.vars = c("Metabolite"))
merged_comparison_1$variable <- rep(c("visit1","visit2"), times = c(432,432))
comparison_1_dataframe_log2<-data.frame(log2(merged_comparison_1[,3]))
comparison_1_final_dataframe_log2<-data.frame(cbind(merged_comparison_1[,1:2],comparison_1_dataframe_log2))
#write.csv(comparison_1_final_dataframe_log2,'ss.csv')

box_plot <- ggboxplot(comparison_1_final_dataframe_log2, x = "variable", y = "log2.merged_comparison_1...3..",
                  color = "variable", palette = "jco",
                  add = "jitter",
                  short.panel.labs = TRUE,facet.by = "Metabolite", ylab = "Normalized value\n(log2)", xlab = "Visit #")+labs(title = "Comparison_1_(Positive_ACR@_Visit_1 vs Positive_ACR@_Visit_2) ")+ facet_wrap( ~ Metabolite, ncol = 5)
box_plot + stat_compare_means(paired = TRUE,label = "p.format",method = "wilcox.test",method.args = list(paired = "TRUE", exact = "FALSE", correct = "FALSE"), label.x.npc = "middle")

尽管我仔细检查了较大和较小的数据矩阵之间没有差异，但是ggplots中的p值与第一个矩阵不同。

请帮助。我不知道代码有什么问题。

您可以在此处查看两个输入文件：

input files

ggplot和总矩阵中的不同p值

0 个答案: