作者memphis (让你喜欢这世界~)
看板R_Language
标题Re: [问题] matrix of P-values (correlation test)
时间Wed Jul 31 00:17:04 2013
直接引用上一文章作者的例子
data.fr = data.frame(matrix(rnorm(1000),100)) #假资料 100 row 10 column
data.fr.cor <- cor(data.fr)
data.fr.cor.tval <- data.fr.cor*sqrt((100-2)/(1-(data.fr.cor*data.fr.cor)))
data.fr.cor.pval <- 2*pt(abs(data.fr.cor.tval), 98, lower.tail=F)
### 参考资料
(wiki) Pearson product-moment correlation coefficient
※ 引述《celestialgod (攸蓝)》之铭言：
: # examples of cor.test
: x <- c(44.4, 45.9, 41.9, 53.3, 44.7, 44.1, 50.7, 45.2, 60.1)
: y <- c( 2.6, 3.1, 2.5, 5.0, 3.6, 4.0, 5.2, 2.8, 3.8)
: cor.test(x, y, method = "kendall", alternative = "greater")
: # You need to input x and y, in addition, you can edit some parameters.
: # For your question, I would write the program like this way
: # Suppose your dataset is a data.frame which is named as data.fr
: data.fr = data.frame(matrix(rnorm(1000),100)) # assumed data
: n = dim(data.fr)[2]
: all.comb = t(combn(n,2))
: # To produce all combinations of all variables
: cor_test_res = lapply(1:dim(all.comb)[1], function(v) {
: cor.test(data.fr[,all.comb[v,1]],data.fr[,all.comb[v,2]])})
: # Get the results of tests
: cor.m = sapply(c(4, 1:3), function(i) {
: sapply(1:dim(all.comb)[1], function(v) cor_test_res[[v]][[i]])})
: # transfer the result to a matrix format
: colnames(cor.m) = c("estimate", "statistic (t)", "df", "p.value")
: rownames(cor.m) = sapply(1:dim(all.comb)[1], function(k) {
: paste0(names(data.fr)[all.comb[k,1]], "_",
: names(data.fr)[all.comb[k,2]])})
: # Another
: data.fr = data.frame(matrix(rnorm(1000),100)) # assumed data
: n = dim(data.fr)[2]
: all.comb = t(combn(n,2))
: # To produce all combinations of all variables
: cor.m = Reduce(rbind, lapply(1:dim(all.comb)[1], function(v) {
: Reduce(cbind, cor.test(data.fr[,all.comb[v,1]],
: data.fr[,all.comb[v,2]])[c(4,1:3)])}))
: colnames(cor.m) = c("estimate", "statistic (t)", "df", "p.value")
: rownames(cor.m) = sapply(1:dim(all.comb)[1], function(k) {
: paste0(names(data.fr)[all.comb[k,1]], "_",
: names(data.fr)[all.comb[k,2]])})
: ※ 引述《foam (Working Class Hero)》之铭言：
: : [问题类型]:
: : 程式谘询(我想用R 做某件事情，但是我不知道要怎麽用R 写出来)
: : [软体熟悉度]:
: : 入门(写过其他程式，只是对语法不熟悉)
: : [问题叙述]:
: : 我用 corMat <- cor(df)
: : 可以得到一个correlation matrix
: : (我的dataframe(df) 每一个column数值间的correlation coefficient)
: : 请问如果我想要得到matrix of p-values (correlation test的 p-value)
: : 应该怎麽写呢?
: : 谢谢
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 220.136.18.41
1F：→ celestialgod:我太依赖内建的函数了qq 07/31 00:31
2F：推 foam:谢谢 08/03 00:57