search

robjhyndman / forecast

Forecasting Functions for Time Series and Linear Models
http://pkg.robjhyndman.com/forecast
1.1k stars 340 forks source link

Error in ets(myTS, alpha = 0.1, beta = 0.45, gamma = 0.1) : No model able to be fitted #884

Closed Mayalaroz closed 3 years ago

Mayalaroz commented 3 years ago

Hello,

I try to forecast a time serie with the ets function. Here is an reproducible example:

library(forecast)

myTS <- c(600, 500, 225, 500, 50, 1025, 1000, 950, 6975, 2400, 2400, 1250, 3500, 3125, 900, 2900, 600, 2150, 900, 1800,  750, 1650,  925,  900, 1075, 1800, 4575, 3725, 1100, 450
          , 175, 1275, 1025, 1025, 1575, 1500, 4525, 2000, 1350, 300,  425,  475,   25,  575,  225, 1000,  325,  225,  275,  325,  200,  450,  350, 1175,  250,  400,  650, 2150, 1075,  425)

myTS_calc <- ets(myTS, alpha = 0.1, beta = 0.45, gamma = 0.1)

I get the following error:

[1] "Model: ETS(A,N,N)"
[1] "Model: ETS(A,Ad,N)"
[1] "Model: ETS(A,A,N)"
[1] "Model: ETS(M,N,N)"
[1] "Model: ETS(M,Ad,N)"
[1] "Model: ETS(M,A,N)"
Error in ets(myTS, alpha = 0.1, beta = 0.45, gamma = 0.1) : 
  No model able to be fitted

Where does this error come from? I thought that the values of alpha, beta and gamma are between 0 and 1. Is there another restriction?

robjhyndman commented 3 years ago

beta must be less than alpha. See https://otexts.com/fpp2/estimation-and-model-selection.html