R语言 - t.test：不应该＆＃39; x＆＃39;意见

Question

嘿，

我在R中有一个表格，格式如下

氨基酸|标签|比率1 |比率2

对于此表，我想应用以下函数为每行添加t.test值

table<- apply(table , 1,function(x) {

i <- as.numeric(x[3])
j <- as.numeric(x[4])

#Perform t.test
if( !is.na(i) && !is.na(j)){
test <- t.test( i, j, paired = TRUE)

out <- data.frame("Amino acids" = colnames(r)[r[1]],
                  "Tag" = colnames(r)[r[2]],
                  "LRatio 1" = colnames(r)[r[3]],
                  "LRatio 2" = colnames(r)[r[4]],
                  "Tvalue" = sprintf("%.3f", test$statistic),
                  "df" = test$parameter,
                  "Pvalue" = sprintf("%.3f", test$p.value))
return (out)
}
out <- data.frame("Amino acids" = colnames(r)[r[1]],
                "Tag" = colnames(r)[r[2]],
                "LRatio 1" = colnames(r)[r[3]],
                "LRatio 2" = colnames(r)[r[4]],
                "Tvalue" = "NA",
                "df" = "NA",
                "Pvalue" = "NA")  
return (out)
})

我的t.test方法，抱怨我没有足够的＆＃39; x＆＃39;观测＆＃34;

调试我拥有的值的示例时：

i: -0.2480872

j: -0.1382415

x: merged.ac
A0A075B5K6, A0A0B4J1H6, P01629, P01630, P03976

merged.r2
127

merged.lratio.x
-0.2480872152

merged.lratio.y
-1.382415e-01

Answer 1

这个问题的答案是t.test方法将两个非空数字向量作为参数。在上面的代码中，i和j是数值，因此x值不足的误差。