[R] time series has no or less than 2 periods

[Clint Bowman]

Perhaps looking at your data will suggest an appropriate number, viz.

plot(data,type="b",xlim=c(0,20),ylim=c(0,50))
par(new=T)
ind<-1:19  # in this case where the data length is 19
data.ind<-data.frame(ind,data)
data.lo<-loess(data~ind,data.ind)
data.pre<-predict(data.lo,data.frame(ind = seq(1,19,1))) 
plot(data.pre,pch=3,col=2,xlim=c(0,20),ylim=c(0,50))

If you now plot the difference between the data and the loess prediction,

data.ind<-cbind(data.ind,data.pre)
data.diff<-with(data.ind,data-data.pre)
data.ind<-cbind(data.ind,data.diff)
with(data.ind,plot(ind,data.diff,type="b"))
abline(h=0)

there is also a pretty strong two week signal--is that of any interest?

Now you should be able to decide how to proceed.

Clint

Clint Bowman			INTERNET:	clint at ecy.wa.gov
Air Quality Modeler		INTERNET:	clint at math.utah.edu
Department of Ecology		VOICE:		(360) 407-6815
PO Box 47600			FAX:		(360) 407-7534
Olympia, WA 98504-7600

         USPS:           PO Box 47600, Olympia, WA 98504-7600
         Parcels:        300 Desmond Drive, Lacey, WA 98503-1274

On Fri, 4 Oct 2013, Daniel Hickman wrote:

> Bill,
>
> Thanks for replying.
>
>
> The data is weekly time series data.  Assume there is 52 weeks in the year.  Of the 52 weeks, I typically only have data for weeks 8 through 40.
>
>
> 4-Apr-10, 8, 27.2
> 11-Apr-10, 9, 32.3
> 18-Apr-10, 10, 31.7
>
> DataXYZ, 40, 13.4
>
>
> data <- c(0,24.57,29.93,24.19,12.25,48.07,36.68,24.78,48.69,30.39,48.17,36.51,36.43,36.52,48.75,24.17,37.07,0,18.89)
> ts <- ts(data= data, start = 8, end = 40, frequency = ????)
>
>
> There is a weekly seasonality effect.  What should I set my frequency value to?
>
>
> Thanks,
>
> Dan Hickman
>
>
>
>
>
>
> From: William Dunlap
> Sent: ???Thursday???, ???October??? ???3???, ???2013 ???3???:???57??? ???PM
> To: David Winsemius, Daniel Hickman
> Cc: r-help at r-project.org
>
>
>>> ts <- ts(data$QtyPerWeek, frequency=52)
>>> HoltWinters(ts,0.46924,0.05,0.2)
>>>
>>> This results in the following error. "Error in decompose(ts(x[1L:wind], start = start(x),
>> frequency = f), seasonal) : time series has no or less than 2 periods"
>
> Since you have set the frequency of the time series to 52, you need
> to have 104 observations to get the initial estimate of the seasonal
> pattern.  How many observations are in 'ts'?  If you don't have enough
> you can omit the seaonal component (HoltWinters(gamma=FALSE,...)),
> change start.periods from the default 2 to 1, or supply a 52-long vector
> of the initial seasonal pattern as the s.start argument.
>
> If you do have more than 104 observations then you will have to tell
> us more about the data.
>
> Bill Dunlap
> Spotfire, TIBCO Software
> wdunlap tibco.com
>
>
>> -----Original Message-----
>> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf
>> Of David Winsemius
>> Sent: Thursday, October 03, 2013 12:39 PM
>> To: Daniel Hickman
>> Cc: r-help at r-project.org
>> Subject: Re: [R] time series has no or less than 2 periods
>>
>>
>> On Oct 3, 2013, at 8:32 AM, Daniel Hickman wrote:
>>
>>> Hello,
>>>
>>>
>>>
>>> I have been tasked with taking an excel file that my colleague had implemented Triple
>> Exponential Smoothing and recreate using R.
>>>
>>> The following image shows the before and after of smoothing out a fixed interval time
>> series data using Triple Exponential Smoothing inside of Excel.
>>>
>>> enter image description here
>>
>> The image file formats that I know are acceptable are .ps, .pdf or .png. Not sure about
>> jpeg.
>>
>>>
>>> I am trying to perform the same triple exponential smoothing in R.  I created a csv file
>> with the before smoothing data.  The csv file is attached and can also be found here.
>>
>> Need to send with .txt extension.
>>
>>>
>>> I found the HoltWinters method but I keep getting an error when I try to apply
>> HoltWinters against the csv.
>>> setwd("C:/temp")
>>> data <- read.table("TripleExpSmoothingXLS.csv", header=TRUE, sep=",")
>>> ts <- ts(data$QtyPerWeek, frequency=52)
>>> HoltWinters(ts,0.46924,0.05,0.2)
>>>
>>> This results in the following error. "Error in decompose(ts(x[1L:wind], start = start(x),
>> frequency = f), seasonal) : time series has no or less than 2 periods"
>>
>> Perhaps a data entry problem. We would need to see either the file or output of
>> str(data).
>>>
>>> In case it helps,  excel file with the triple exponential smoothing formulas and original
>> data can be found here.
>>
>> Again.... there is no here here.
>>
>>>
>>> Any advice?
>>>
>>> Thanks, Dan______________________________________________
>>> 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.
>>
>> David Winsemius
>> Alameda, CA, USA
>>
>> ______________________________________________
>> 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.
> 	[[alternative HTML version deleted]]