R help - Nov 2009 - multiple tests: t-statistic for vectors in 4-dimensional array

Hi everyone,

I created a four dimensional vector (dim (128,128,1,8)).  This third 
dimension is necessary for another function somewhere.  Now I'd like to 
perform a t-test on every vector of length 8 in my array on the fourth 
dimension.
I'd like to obtain a new array of three dimensions with dimensions 
128x128x1 with all these test statistics.

I tried this with a double loop:

A <- array(rep(seq(-1:2),128*128*8),dim=c(128,128,1,8)) #in simulations 
I fill the array with random variables
tA <- array(rep(NA,128*128),dim=c(128,128,1))

for (i in 1:128){
for (j in 1:128){
print(c(i,j))
tA[i,j,1] <-
t.test(A[i,j,1,],alternative="two.sided",mu=0)$statistic
}}

This works, but it takes way to much time, since I'd like to run this 
about 500 times.

I read something about the apply function, but there I don't know how to 
specify the outcome values ($statistic), plus I have some difficulties 
with the specification of my vectors...
This doesn't work:
tA <- apply(smoothA,4,t.test$statistic)

"Error in t.test$statistic : object of type 'closure' is not
subsettable"

I sought in the package multtest, but can't find any helpful function.

Thanks,

J.

On Mon, 9 Nov 2009, J. wrote:
> Hi everyone,
>
> I created a four dimensional vector (dim (128,128,1,8)).  This third 
> dimension is necessary for another function somewhere.  Now I'd like to
> perform a t-test on every vector of length 8 in my array on the fourth 
> dimension.

Vectorize the whole thing:


adf <- as.data.frame( aperm( A, c(4,1,2,3)))
tA2 <- array( mean(adf) / sd(adf) * sqrt(8), c(128,128,1))


HTH,

Chuck
> I'd like to obtain a new array of three dimensions with dimensions
128x128x1
> with all these test statistics.
>
> I tried this with a double loop:
>
> A <- array(rep(seq(-1:2),128*128*8),dim=c(128,128,1,8)) #in simulations
I
> fill the array with random variables
> tA <- array(rep(NA,128*128),dim=c(128,128,1))
>
> for (i in 1:128){
> for (j in 1:128){
> print(c(i,j))
> tA[i,j,1] <-
t.test(A[i,j,1,],alternative="two.sided",mu=0)$statistic
> }}
>
> This works, but it takes way to much time, since I'd like to run this
about
> 500 times.
>
> I read something about the apply function, but there I don't know how
to
> specify the outcome values ($statistic), plus I have some difficulties with
> the specification of my vectors...
> This doesn't work:
> tA <- apply(smoothA,4,t.test$statistic)
>
> "Error in t.test$statistic : object of type 'closure' is not
subsettable"
>
> I sought in the package multtest, but can't find any helpful function.
>
> Thanks,
>
> J.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
Charles C. Berry                            (858) 534-2098
                                             Dept of Family/Preventive Medicine
E mailto:cberry at tajo.ucsd.edu	            UC San Diego
http://famprevmed.ucsd.edu/faculty/cberry/  La Jolla, San Diego 92093-0901