R help - Aug 2008 - Iterative Averages

Dear all,

I have a data frame of 2160 rows and 4320 columns, which I hope to condense to a
smaller dataset by finding averages of 6 by 6 blocks of values (to produce a
data frame of 360 rows by 720 columns).

How would I go about finding the mean of a 6 x 6 block, then find the mean of
the next adjacent 6 x 6 block, and so on, until the whole data frame has been
covered?

One slight twist is that I have NA values, which I don't want to be included
in the calculations unless a particular 6 x 6 block is entirely composed of NA
values - in which case, NA should be the output value.

Thanks very much for any advice and solutions.

Steve

_________________________________________________________________
Get Hotmail on your mobile from Vodafone

Thanks very much for both your suggestions.

When you refer to doing a 'double for loop', do you mean finding the
average for rowgrp and colgrp within each 6x6 block? If so, how would this be
done so that the whole data frame is covered? It would seem to me that the
'mean' operation would need to combine rowgrp AND colgrp together? How
would I get the loop to cycle through each set of 6x6 blocks of the data frame?

Thanks again,

Steve

_________________________________________________________________
Win New York holidays with Kellogg?s & Live Search

Use kronecker products. For example,
> mat <- matrix(1001:1144, 12, 12)
> result <- k1 <- matrix(1:9, 3, 3)
> k2 <- matrix(1, 4, 4) # size of one block
> result[] <- tapply(mat, kronecker(k1, k2), mean)
> result
       [,1]   [,2]   [,3]
[1,] 1020.5 1068.5 1116.5
[2,] 1024.5 1072.5 1120.5
[3,] 1028.5 1076.5 1124.5
> mean(mat[1:4, 1:4]) # check first block
[1] 1020.5


On Sat, Aug 2, 2008 at 9:28 AM, Steve Murray <smurray444 at hotmail.com>
wrote:
>
> Dear all,
>
> I have a data frame of 2160 rows and 4320 columns, which I hope to condense
to a smaller dataset by finding averages of 6 by 6 blocks of values (to produce
a data frame of 360 rows by 720 columns).
>
> How would I go about finding the mean of a 6 x 6 block, then find the mean
of the next adjacent 6 x 6 block, and so on, until the whole data frame has been
covered?
>
> One slight twist is that I have NA values, which I don't want to be
included in the calculations unless a particular 6 x 6 block is entirely
composed of NA values - in which case, NA should be the output value.
>
> Thanks very much for any advice and solutions.
>
> Steve
>
> _________________________________________________________________
> Get Hotmail on your mobile from Vodafone
>
> ______________________________________________
> 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.
>