Question

我有两个源自同一数据集的矩阵，但每个矩阵都有不同的数据量。我想创建一个数据集，它在列名和行名方面是x的复制，但包含y中的数据值。如果数据不可用，则NA将用作该坐标的值。

并非x中的所有行名都出现在y中，反之亦然。列名也是如此。

对于我在下面给出的示例输入数据，x中与y中的rowname相对应的rownames是rowname的开始和结束|（我希望保留后的everthing）其他映射的|。

最有效的方法是什么？

期望的输出

z = structure(c(NA, 1, NA, NA, NA, NA, NA, NA, 0, NA, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 0, NA, NA, NA, NA, NA, 
NA, 0, NA, NA, NA, 0, NA, NA, NA, NA, NA, NA, 0, NA, NA, NA, 
NA, NA, NA, NA, NA, NA, NA, NA, NA, NA), .Dim = c(11L, 5L), .Dimnames = list(
c("AACSL|729522", "AACS|65985", "AADACL2|344752", "AADACL3|126767", 
"AADACL4|343066", "AADAC|13", "AADAT|51166", "AAGAB|79719", 
"AAK1|22848", "AAK12|14", "AANAT|15"), c("S18", "S20", "S45", 
"S95", "S100")))

示例输入

x = structure(c(0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 
1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 
0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0), .Dim = c(11L, 
5L), .Dimnames = list(c("AACSL|729522", "AACS|65985", "AADACL2|344752", 
"AADACL3|126767", "AADACL4|343066", "AADAC|13", "AADAT|51166", 
"AAGAB|79719", "AAK1|22848", "AAK12|14", "AANAT|15"), c("S18", 
"S20", "S45", "S95", "S100")))

y = structure(c(0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 
0, 0, 0, 0, 0, 0), .Dim = c(11L, 4L), .Dimnames = list(c("A1BG", 
"A1CF", "A2ML1", "A4GALT", "AACS", "AAK1", "AARD", "AARS2", "AASDHPPT", 
"AASS", "BAACS"), c("S18", "S10", "S45", "S95")))

Answer 1

我认为您提供的示例可能存在轻微问题，我无法看到z如何来自上面的x和y ..请参阅此代码：

intersect(sapply(rownames(x), #I am just extracting the letter codes here
             function(i){
                     return(
                             strsplit(x=i,split="|",fixed=TRUE)[[1]][[1]])
             }),rownames(y))

#[1] "AACS" "AAK1"

很奇怪，对吗？我的意思是，与x相比，y中只有2个代码。但是，我认为下面的代码符合您的计划（除了这种不一致）：

library(data.table)
library(reshape2)
library(dplyr)
x %>% as.data.frame %>% mutate(rownames=rownames(x)) %>%
        mutate(nms=sapply(rownames(x),
                          function(i){
                                  return(
                                          strsplit(x=i,split="|",fixed=TRUE)[[1]][[1]])
                          })) %>%
        melt(id.vars=c("nms","rownames")) %>%
        merge(., y %>% as.data.frame %>% mutate(nms=rownames(y))%>% melt(id.vars="nms"), by=c("variable","nms"), all.x=TRUE) %>%
        select(-nms, -value.x) %>% dcast(formula = rownames~variable, value.var="value.y") -> xy
#now put back the column names where they belong
rownames(xy)<-xy$rownames
#now the only thing left is to arrange the columns
xy[rownames(x),colnames(x)] -> xy

或者我在理解你的一些观点时错了吗？

将一个矩阵的内容映射到另一个矩阵的结构上

1 个答案: