R help - Mar 2009 - How to avoid switching on input type?

Hi,

I need some help improving this ugly code I wrote.  I would like to shift
forward a zoo object, matrix, ts, or list by "shift" items (default 1)
and
fill the holes with 0's.  

The code below works, but it looks ugly.  I could write a function
lag.zerofill() which calls the two functions below depending on the class()
of the item passed in, but that feels very unlike R.

What is the proper way to write this code so that it works for all input
types?  

Is a switch depending on the input type really necessary?  I am hoping the
answer is "no".

Thanks in advance.

- Ken

#
-----------------------------------------------------------------------------------

require( zoo );

inp <- c( 5, 9, 4, 2, 1 ); inp2 <- c( 6, 6, 0, 4, 2 );
inp.zoo <- zoo( cbind( inp, inp2 ), as.Date("2003-02-01") +
(0:(length(inp)-1)));

lag.zerofill.list <- function( m, shift=1, ...) {
  c( rep(0,shift), m[-((length(m)-shift+1):length(m))] );
}
  
lag.zerofill.zoo <- function( m, shift=1, ...) {
  k <- rbind( m[1:shift,], lag(m,-(shift))); k[1:shift,] <- 0; k; 
}

lag.zerofill.list( inp ) # works ok
lag.zerofill.zoo( inp.zoo ) # works ok

-- 
View this message in context:
http://www.nabble.com/How-to-avoid-switching-on-input-type--tp22637249p22637249.html
Sent from the R help mailing list archive at Nabble.com.

Create an S3 generic lag.zerofill and then define methods for each class:

lag.zerofill <- function(x, k = 1, ...) UseMethod("lag.zerofill")
lag.zerofill.zoo <- function(x, k, ...) {
		m <- merge(x, lag(x, k), fill = 0)
		structure(m[, -(1:ncol(x))], dimnames = list(NULL, colnames(x)))
}
# lag.zerofill.list <- ...

# test
lag.zerofill(inp.zoo, -1)

Running the above gives (last line shown only):
> lag.zerofill(inp.zoo, -1)
           inp inp2
2003-02-01   0    0
2003-02-02   5    6
2003-02-03   9    6
2003-02-04   4    0
2003-02-05   2    4


On Sat, Mar 21, 2009 at 11:47 AM, Ken-JP <kfmfe04 at gmail.com>
wrote:
>
> Hi,
>
> I need some help improving this ugly code I wrote. ?I would like to shift
> forward a zoo object, matrix, ts, or list by "shift" items
(default 1) and
> fill the holes with 0's.
>
> The code below works, but it looks ugly. ?I could write a function
> lag.zerofill() which calls the two functions below depending on the class()
> of the item passed in, but that feels very unlike R.
>
> What is the proper way to write this code so that it works for all input
> types?
>
> Is a switch depending on the input type really necessary? ?I am hoping the
> answer is "no".
>
> Thanks in advance.
>
> - Ken
>
> #
>
-----------------------------------------------------------------------------------
>
> require( zoo );
>
> inp <- c( 5, 9, 4, 2, 1 ); inp2 <- c( 6, 6, 0, 4, 2 );
> inp.zoo <- zoo( cbind( inp, inp2 ), as.Date("2003-02-01") +
> (0:(length(inp)-1)));
>
> lag.zerofill.list <- function( m, shift=1, ...) {
> ?c( rep(0,shift), m[-((length(m)-shift+1):length(m))] );
> }
>
> lag.zerofill.zoo <- function( m, shift=1, ...) {
> ?k <- rbind( m[1:shift,], lag(m,-(shift))); k[1:shift,] <- 0; k;
> }
>
> lag.zerofill.list( inp ) # works ok
> lag.zerofill.zoo( inp.zoo ) # works ok
>
> --
> View this message in context:
http://www.nabble.com/How-to-avoid-switching-on-input-type--tp22637249p22637249.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> 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.
>