Error in is.constant(y)

[Steviey]

R 4.0.5, smooth 3.1.5

I try the following code and get the error-msg. below:

liveData<-0
if(liveData<1){
    idx=120
    y<-rnorm(idx, mean=15, sd=5)
    x<-cbind(
        x1 =  rnorm(idx, mean=15, sd=5)
        ,x2 =  rnorm(idx, mean=15, sd=5)
        ,x3 =  rnorm(idx, mean=15, sd=5)
        ,x4 =  rnorm(idx, mean=15, sd=5)
        ,x5 =  rnorm(idx, mean=15, sd=5)
    )   
    data    <- data.frame(x=x,value=y,stringsAsFactors=F)
    data    <- data %>% dplyr::relocate(value)
}

myModel <- adam(data,"ANN",silent=TRUE,h=1,holdout=FALSE)

myPlot<-plot(forecast::forecast(myModel,h=10,interval="complete",nsim=100),main="Complete prediction interval")
print(myPlot)
View(myModel)
stop()

Error in is.constant(y) : 
  'list' object cannot be coerced to type 'double'
    3. +-forecast:::forecast.default(...)
 4. | \-forecast:::forecast.ts(object, ...)
 5. |   +-fabletools::forecast(...)
 6. |   \-forecast::ets(...)
 7. |     \-forecast::is.constant(y)
 8. \-(function () ...
 9.   \-lobstr::cst()
 10.      

[config-i1]

Hi, This is because of the conflicting methods in the packages. I've submitted on CRAN the new version of greybox that is supposed to resolve this. Please, give it a try with the development versions of the package:

• greybox v1.0.4
• smooth v3.1.6.41004

[Steviey]

Windows 7, greybox 1.0.5.41001, smooth 3.1.6.41004, R version 4.0.5

I get the same result with the versions above...

liveData<-0
if(liveData<1){
    idx=120
    y<-rnorm(idx, mean=15, sd=5)
    x<-cbind(
        x1 =  rnorm(idx, mean=15, sd=5)
        ,x2 =  rnorm(idx, mean=15, sd=5)
        ,x3 =  rnorm(idx, mean=15, sd=5)
        ,x4 =  rnorm(idx, mean=15, sd=5)
        ,x5 =  rnorm(idx, mean=15, sd=5)
    )   
    data    <- data.frame(x=x,value=y,stringsAsFactors=F)
    data    <- data %>% dplyr::relocate(value)
}

myModel <- adam(data,"ANN",silent=TRUE,h=1,holdout=FALSE)

myPlot<-plot(forecast::forecast(myModel,h=10,interval="complete",nsim=100),main="Complete prediction interval")
print(myPlot)
View(myModel)
stop()

Error in is.constant(y) : 
  'list' object cannot be coerced to type 'double'

[config-i1]

Please, use:

myPlot<-plot(forecast(myModel,h=10,interval="complete",nsim=100),main="Complete prediction interval")

instead of forecast::forecast(myModel,...). If this doesn't work (which could mean that forecast package hasn't updated its imports) then use generics::forecast(myModel,...).

[Steviey]

Wow, generics works...

One question: I like to customize the graph. ggplot+plotly allows zooming and panning. I can do this with smooth::es()...but I have no entries: 'lower' and 'upper' for the intervals in the adam-model to do so...Can you provide these?

[config-i1]

In adam(), you need to produce forecasts, it does not generate intervals inside the function itself. So, it is closer to the style of forecast functions. Here is an example:

forecast(adamModel, h=h, interval="prediction", level=c(0.9,0.95))