Skip to content

library(forecast): Error in SD.test(x, m) : Insufficient data

4 messages · Jean Jacques Dureau, Rainer Hurling

Original
Jean Jacques Dureau
# 
I have these R code:

###########################################################################
time.test.data<-c(2.88645418326693, 2.91546949823027, 2.94130799329234,
  2.93313338038109, 2.89478957915832, 2.86029757243540, 2.78648486664669,
  2.80183167535133, 2.75435512226307, 2.78992352676563, 2.76028433151845,
  2.68741721854305, 2.70691974293828, 2.683833847881, 2.65041551246537,
  2.65169020111254, 2.58837541686517, 2.66549241844080, 2.58451314648945,
  2.60250871080139, 2.61253722876188, 2.59921041087878, 2.71727961060032,
  2.65440192667915, 2.74799149338374, 2.70649994101687, 2.80636027009366,
  2.81801086502298, 2.82555635319454, 2.87133347201997, 2.75746714456392,
  2.7660659236424, 2.71688375522241, 2.72655367231638, 2.72461997828447,
  2.78455790784558, 2.71160702495708, 2.65754456439869, 2.85673280918507,
  2.7053919591233, 2.7532637075718, 2.74272237196766, 2.75893306492199,
  2.62584686181772, 2.75230602278893, 2.82781018027572, 2.80220656652931,
  2.80242587601078, 2.84061953534849, 2.87123514783089, 2.76991605276683,
  2.77796934865900)

library(forecast)
auto.arima(ts(time.test.data, start=c(2011,1), frequency=52))
###########################################################################

I obtain the error message:
Error in SD.test(x, m) : Insufficient data

What is wrong?

jj
Rainer Hurling
#
On 08.02.2012 15:35 (UTC+1), Jean Jacques Dureau wrote:
This looks good to me, I got no error:

auto.arima(ts(time.test.data, start=c(2011,1), frequency=52))
Series: ts(time.test.data, start = c(2011, 1), frequency = 52)
ARIMA(2,0,2) with non-zero mean

Coefficients:
          ar1      ar2      ma1     ma2  intercept
       1.4540  -0.5762  -1.0235  0.5484     2.7569
s.e.  0.3351   0.3194   0.3240  0.1902     0.0328

sigma^2 estimated as 0.003059:  log likelihood=76.05
AIC=-140.09   AICc=-138.23   BIC=-128.38


So perhaps something is wrong with your data object or with your 
dependend packages?
Jean Jacques Dureau
# 
Hi rainer,
how can I control  dependend packages?

thanks

jj

Il 08 febbraio 2012 17:10, Rainer Hurling <rhurlin at gwdg.de> ha scritto:
Rainer Hurling
#
On 08.02.2012 17:19 (UTC+1), Jean Jacques Dureau wrote:
You did not tell us very much about your installation and versions of 
packages you are using.

On my system sessionInfo() gives me the following after loading your 
example:

sessionInfo()
R Under development (unstable) (2012-02-07 r58290)
Platform: amd64-portbld-freebsd10.0 (64-bit)

locale:
[1] 
de_DE.ISO8859-15/de_DE.ISO8859-15/de_DE.ISO8859-15/C/de_DE.ISO8859-15/de_DE.ISO8859-15

attached base packages:
[1] parallel  stats     graphics  grDevices utils     datasets  methods 
   base

other attached packages:
[1] forecast_3.17        RcppArmadillo_0.2.34 Rcpp_0.9.9 fracdiff_1.4-0
[5] tseries_0.10-27      zoo_1.7-6            quadprog_1.5-4

loaded via a namespace (and not attached):
[1] grid_2.15.0    lattice_0.20-0


Here you can see, that package forecast uses other packages like 
RcppArmadillo, Rcpp, fracdiff and others. If your dataset 
'time.test.data' is ok, there is a (little) chance, that there are some 
not updated packages used by forecast.

But I am for sure not an expert on this. Perhaps someone else has an idea?

Rainer