3

I am using R and have a question on correlations. 

A<-data.frame(A1=c(1,2,3,4,5),B1=c(6,7,8,9,10),C1=c(11,12,13,14,15 ))
B<-data.frame(A2=c(6,7,7,10,11),B2=c(2,1,3,8,11),C2=c(1,5,16,7,8))
cor(A,B)
#           A2        B2       C2
# A1 0.9481224 0.9190183 0.459588
# B1 0.9481224 0.9190183 0.459588
# C1 0.9481224 0.9190183 0.459588

I wanted to obtain the p-value for each of the correlation coefficients in the matrix. Is this possible?

I tried using rcorr function from Hmisc package but obtain only a single p-value and not for each correlation.

A <- as.vector(t(A))
B <- as.vector(t(B))
rcorr(A, B)
     x    y
x 1.00 0.13
y 0.13 1.00

n= 15 


P
  x      y     
x        0.6425
y 0.6425

Similarly, I also tried using "psych" package in R to do this but unable to.

Jilber Urbina
  • 58,147
  • 10
  • 114
  • 138
Paul
  • 1,077
  • 3
  • 14
  • 27

1 Answers1

3

You can apply rcorr directly on A and B if you convert them to matrices first :

library(Hmisc)
rcorr(as.matrix(A),as.matrix(B))

Which gives :

     A1   B1   C1   A2   B2   C2
A1 1.00 1.00 1.00 0.95 0.92 0.46
B1 1.00 1.00 1.00 0.95 0.92 0.46
C1 1.00 1.00 1.00 0.95 0.92 0.46
A2 0.95 0.95 0.95 1.00 0.97 0.16
B2 0.92 0.92 0.92 0.97 1.00 0.15
C2 0.46 0.46 0.46 0.16 0.15 1.00

n= 5 


P
   A1     B1     C1     A2     B2     C2    
A1        0.0000 0.0000 0.0141 0.0273 0.4361
B1 0.0000        0.0000 0.0141 0.0273 0.4361
C1 0.0000 0.0000        0.0141 0.0273 0.4361
A2 0.0141 0.0141 0.0141        0.0078 0.7981
B2 0.0273 0.0273 0.0273 0.0078        0.8125
C2 0.4361 0.4361 0.4361 0.7981 0.8125
juba
  • 47,631
  • 14
  • 113
  • 118
  • @ juba: I have one more clarification on this data but its regarding permutations. Should I post it as a new question? – Paul Sep 24 '13 at 09:44
  • 1
    @user2810362 Is it a data manipulation question or a statistical one ? In the second case the question could better go to Cross Validated. – juba Sep 24 '13 at 09:46