Re: [R] extend.series not zero padding

From: Spencer Graves <>
Date: Sat 10 Dec 2005 - 13:12:38 EST

HELLO ERIC:           The documentation for "extend.series" indicates that "zero" is an option for "method". However, consider the followining:

 > extend.series(c(0, pi), method="zero") Error in extend.series(c(0, pi), method = "zero") :

        Invalid argument value for 'method'

          If you replace "zero" with "zeros" in the documentation, it will be consistent with the actual function behavior:

 > extend.series(c(0, pi), method="zeros") [1] 0.000000 3.141593 0.000000 0.000000

HELLO KEITH:           Permit me to outline for your general edification how I found this. First, I installed "wavelets" and replicated your error. I noticed that your sample series was named "c". Since that's also a function name, I replaced "c" with "c.". That did not fix the error. However, I generally avoid using the names of standard functions for other purposes. In most but not all cases, R can tell from the context whether the function or the non-function object 'c' should be used. To avoid being trapped by this in the occasionally cases where R can NOT tell the difference, I routine type a variabile name at a command prompt before using it: If I get 'object not found', I feel more comfortable using it. If R returns something, I tend to avoid that name.

          Next, I tried to simplify your example. I determined from the documnetation that we should be able to delete the last two arguments and still get the same thing. I therefore tried the following:


          When I got the same error message, I then listed the function by typing "extend.series" at a command prompt and copied the text into a script file so I could look at it.

          Then I invoked "debug(extend.series)" and tried the 'dew0<-extend.series(...)" command again. From this, I found it failed on the following:

     if (, c("periodic", "reflection", "zeros",
         "mean", "reflection.inverse"))))
	  I then tried it with "method='zeros'", and it worked.

	  Then, in preparing an email to Eric Aldrich, I simplified the example 
still further to make it easier for him to understand the issue.
	  hope this helps.
	  spencer graves

Keith Chamberlain wrote:

> Dear List,
> I was trying to verify that I could use extend.series in the wavelets
> package and kept getting an error when trying to use method="zero". I'm not
> seeing where my syntax has gone awry.
> According to the documentation, [see ?extend.series]
> " method: A character string indicating which extension method to use.
> Possible values are '"periodic"', '"reflection"', '"zero"',
> '"mean"', and '"reflection.inverse"'."
> c<-cbind(0:60, 60:0) # setup a series, length will be 61


> [1] 122

> j=log(length(c),2)%/%1)
>>Error in extend.series(c, method = "zero", length = "powerof2", j =

> log(length(c), :
> Invalid argument value for 'method'
> Other methods work great, such as method="mean".

> j=log(length(c),2)%/%1)

> [1] 128
> Has this come up in anyone else's experience? If so, what's the workaround
> so that I can use "zero" as a method?
> Rgds,
> KeithC.
> ______________________________________________
> mailing list
> PLEASE do read the posting guide!
Spencer Graves, PhD
Senior Development Engineer
PDF Solutions, Inc.
333 West San Carlos Street Suite 700
San Jose, CA 95110, USA <>
Tel:  408-938-4420
Fax: 408-280-7915

______________________________________________ mailing list
PLEASE do read the posting guide!
Received on Sat Dec 10 13:17:54 2005

This archive was generated by hypermail 2.1.8 : Sat 10 Dec 2005 - 17:24:59 EST