作者LSCAT (秘密客)
看板Statistics
标题[问题] R软体的矩阵问题!
时间Tue Sep 18 07:05:45 2012
如果是跟统计软体有关请重发文章
如果跟论文有关也烦请您重发文章
文章类别是为了帮助大家搜寻资料与解答,造成不便之处请见谅
Simulate 2,000 independent sets of i.i.d. standard Normal random numbers,
with 50 observations in each set. Find the maximum absolute correlation coefficient among all
pairs of sets. Comment on the results. What do you expect to see if there are 20,000 such
independent sets?
有人可以帮忙解答一下吗?
--
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 204.52.215.1
1F:→ bmka:作业? 09/18 10:06
2F:→ gsuper:DATA<-matrix(NA,50,2000); 09/18 13:36
3F:→ gsuper:for(g in 1:2000){DATA[,g]<-rnorm(50)} 09/18 13:36
4F:→ gsuper:x<-cor(DATA);x[abs(x)==1]<-0; 09/18 13:36
5F:→ gsuper:y<-abs(x)==max(x) ; 09/18 13:36
6F:→ gsuper:which(colSums(y)>0) ; which(rowSums(y)>0); 09/18 13:37
7F:→ gsuper:Assume 只有一个最大值 & 最大值不可能为 1 09/18 13:38
8F:→ gsuper:其它还有些手工的部分就自己学吧 09/18 13:38