5

I've seen that multiplicative Holt-Winters requires strictly positive data points. I was wondering why it does not allow zero values?

andrey
  • 61
  • Does it require strictly positive data? Neither this nor this suggest it in their discussion of multiplicative HW. Who says so? – Glen_b Mar 15 '14 at 04:59
  • 1
    here: http://stat.ethz.ch/R-manual/R-patched/library/stats/html/HoltWinters.html The data in x are required to be non-zero for a multiplicative model, but it makes most sense if they are all positive. http://www.okstate.edu/sas/v8/saspdf/ets/chap12.pdf The WINTERS method assumes that the series values are positive. If negative or zero values are found in the series, a warning is printed and the values are treated as missing. – andrey Mar 15 '14 at 14:14

1 Answers1

4

Because in multiplicative model for Holt Winter trend is calculated as division of two data points rather than subtraction (incase of additive model). Multiplicative model is generally more stable than additive model.

Shreyak
  • 141