R help - Nov 2006 - problem with indexing a zoo object

My problem is the following : I create 2 zoo objects and then I try to
subset one of them using logic. indicesthatpass is a vector of trues and
falses but when I send it into bckret, it returns an empty bckret.
Obviously it has something to do with bckret being a zoo object and if I
do the same subsctripting off of coredata(bckret), I'm confident it will
work. But, I need to keep the minute days of bckret.

So, I have two questions :

1) is there a way around this ? 

2) even if the answer to 1) is yes, there could be some other way better
than what I am doing ? It took me a while to even get to this point
because I was taking is.infinite of fwdret and abs of bckret rather than
their respective coredatas and that was really causing a mess because of
the different indexing that was getting returned. 

Thanks.


# SET SEED
set.seed(1)

# CREATE  2 ZOO OBJECTS, bckkret and fwdret OF MINUTE DATA OVER THE
WHOLE DAY

x.date<-as.Date(paste(2003,01,01,sep="-"))
allminutes<-as.POSIXct(format(x.date,"%Y-%m-%d")) + 60 * seq(1, 24
* 60)
bckret<-zoo(matrix(abs(rnorm(1440)),nc=1),allminutes)
fwdret<-zoo(matrix(abs(rnorm(1440)),nc=1),allminutes)

indicesthatpass<-!is.infinite(coredata(fwdret)) &
!(abs(coredata(bckret)) > 1.00)

#SUBSET bckret 

bckret<-bckret[indicesthatpass,drop=FALSE]
--------------------------------------------------------

This is not an offer (or solicitation of an offer) to buy/se...{{dropped}}

Hi guys:
           I'm testing some statistic discriminant analysis methods in R. 
But I couldn't find the package for these methods:
           1. Discriminant analysis implemented with nearest neighbor 
method,
           2. logistic discriminant analysis
           3. POLYCLASS algorithm( there is such a function in S-Plus, 
namely poly.fit from polyclass library. Is there a corresponding library in 
R? )
           Any kind of help is highly appreciated!

Best,

Feng

On Wed, 29 Nov 2006, Leeds, Mark (IED) wrote:
> My problem is the following : I create 2 zoo objects and then I try to
> subset one of them using logic. indicesthatpass is a vector of trues and
Nope, it's a matrix (because you want the coredata to be a matrix) but
should be a vector. (Maybe we could extend [.zoo here...)
> 2) even if the answer to 1) is yes, there could be some other way better
> than what I am doing ? It took me a while to even get to this point
> because I was taking is.infinite of fwdret and abs of bckret rather than
> their respective coredatas and that was really causing a mess because of
> the different indexing that was getting returned.
Logical comparisons on zoo objects return zoo objects as documented in
various places (which you should know about by now) including ?zoo and the
package vignette. The documentation suggests various strategies (including
using coredata).
Z