array dimension changes with assignment

[Knut M. Wittkowski]

Why does the assignment of a 3178x93 object to
another 3178x93 object remove the dimension attribute?

 > GT <- array(dim = c(6,nrow(InData),ncol(InSNPs)))
 > dim(GT)
[1]    6 3178   93

 > SNP1 <- InSNPs[InData[,"C1"],]
 > dim(SNP1)
[1] 3178   93

 > SNP2 <- InSNPs[InData[,"C2"],]
 > dim(SNP2)
[1] 3178   93

 > dim(pmin(SNP1,SNP2))
[1] 3178   93

 > GT[1,,] <- pmin(SNP1,SNP2)
 > dim(GT)
NULL    # why??????????????????????????????????????

 > GT[2,,] <- pmax(SNP1,SNP2)
Error in GT[2, , ] <- pmax(SNP1, SNP2) : incorrect number of subscripts


Knut M. Wittkowski, PhD,DSc
----------------------------------------------------
Sr Research Associate,    The Rockefeller University
Biometrician,    The Rockefeller University Hospital
Director, Biostatistics/Epidemiology/Research Design
Center for Clinical and Translational Science (CCTS)
apl. Professor,  Eberhard-Karls-University, Tübingen
1230 York Ave #121B, Box 322, New York, NY 10021, US
Tel: [phone],       Fax: [phone]
kmw@...,   www.rockefeller.edu/ccts/rdbs

        [[alternative HTML version deleted]]


______________________________________________
R-help@... 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.

[Jeremiah Rounds]

>Why does the assignment of a 3178x93 object to >another 3178x93 object remove the dimension attribute?>> GT <- array(dim = c(6,nrow(InData),ncol(InSNPs)))>> dim(GT)>[1]   6 3178  93>> SNP1 <- InSNPs[InData[,"C1"],]>> dim(SNP1)>[1] 3178  93>> SNP2 <- InSNPs[InData[,"C2"],]>> dim(SNP2)>[1] 3178  93>> dim(pmin(SNP1,SNP2))>[1] 3178  93>> GT[1,,] <- pmin(SNP1,SNP2)>> dim(GT)>NULL   # why??????????????????????????????????????>> GT[2,,] <- pmax(SNP1,SNP2)>Error in GT[2, , ] <- pmax(SNP1, SNP2) : incorrect number of .subscripts
---------------------------------------------------

My understanding is that an array is just a list with a dimension attribute, so first note that loosing the dim attribute is not a great loss.  It does not represent an inefficiency.

But consider this code:
> GT <- array(dim = c(6,3178,  93) )> dim(GT)[1]    6 3178   93> > > SNP1 <-as.array(matrix(0,nrow=3178,ncol=93))> dim(SNP1)[1] 3178   93> > > > GT[1,,] <- SNP1> dim(GT)[1]    6 3178   93


Here what you wanted to happen happened just fine.  So the question you might ask yourself is: what is different?  And that leads to asking what class is the SNP1 object?  If you can coerce into an array you probably can avoid the issue.

Jeremiah Rounds
Graduate Student
Utah State University
_________________________________________________________________
Get Free (PRODUCT) RED™  Emoticons, Winks and Display Pics.

        [[alternative HTML version deleted]]


______________________________________________
R-help@... 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.

[Duncan Murdoch-2]

On 5/13/2008 12:47 AM, Knut M. Wittkowski wrote:
> Why does the assignment of a 3178x93 object to
> another 3178x93 object remove the dimension attribute?

Your example is not reproducible.  When I make one that is reproducible,
I don't see the error:

 > GT <- array(dim = c(6,3178,93))
 > dim(GT)
[1]    6 3178   93
 > GT[1,,] <- array(dim=c(3178,93))
 > dim(GT)
[1]    6 3178   93

(This was done in R 2.7.0.)

So I'd suggest that you try to make the example reproducible, and you'll
likely discover that something else you did caused the loss of dimension.

Duncan Murdoch

>
>  > GT <- array(dim = c(6,nrow(InData),ncol(InSNPs)))
>  > dim(GT)
> [1]    6 3178   93
>
>  > SNP1 <- InSNPs[InData[,"C1"],]
>  > dim(SNP1)
> [1] 3178   93
>
>  > SNP2 <- InSNPs[InData[,"C2"],]
>  > dim(SNP2)
> [1] 3178   93
>
>  > dim(pmin(SNP1,SNP2))
> [1] 3178   93
>
>  > GT[1,,] <- pmin(SNP1,SNP2)
>  > dim(GT)
> NULL    # why??????????????????????????????????????
>
>  > GT[2,,] <- pmax(SNP1,SNP2)
> Error in GT[2, , ] <- pmax(SNP1, SNP2) : incorrect number of subscripts
>
>
> Knut M. Wittkowski, PhD,DSc
> ----------------------------------------------------
> Sr Research Associate,    The Rockefeller University
> Biometrician,    The Rockefeller University Hospital
> Director, Biostatistics/Epidemiology/Research Design
> Center for Clinical and Translational Science (CCTS)
> apl. Professor,  Eberhard-Karls-University, Tübingen
> 1230 York Ave #121B, Box 322, New York, NY 10021, US
> Tel: [phone],       Fax: [phone]
> kmw@...,   www.rockefeller.edu/ccts/rdbs
>
> [[alternative HTML version deleted]]
>
>
>
> ------------------------------------------------------------------------
>
> ______________________________________________
> R-help@... 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.

______________________________________________
R-help@... 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.