I would lik to use {highcharter} to plot stl objects. To to so, I would like to obtain seasonal, trend, remainder etc. from stl objects:
str(fit_stl)
> List of 8 $ time.series: Time-Series [1:164, 1:3] from 2010 to 2024:
> 284.6 48.7 156.2 -78.8 -85 ... ..- attr(*, "dimnames")=List of 2 .. ..$ : NULL .. ..$ : chr [1:3] "seasonal" "trend" "remainder" $
> weights : num [1:164] 1 1 1 1 1 1 1 1 1 1 ... $ call :
> language stl(x = TotalGen_ts, s.window = 10) $ win : Named num
> [1:3] 10 23 13 ..- attr(*, "names")= chr [1:3] "s" "t" "l" $ deg
> : Named int [1:3] 0 1 1 ..- attr(*, "names")= chr [1:3] "s" "t" "l"
> $ jump : Named num [1:3] 1 3 2 ..- attr(*, "names")= chr [1:3]
> "s" "t" "l" $ inner : int 2 $ outer : int 0
> - attr(*, "class")= chr "stl"
plot can handle stl objects:
fit_stl = stl(TotalGen_ts, s.window = 10)
plot(fit_stl)
However, so far, stl seems to be not supported by highcharter::hchart(). I'm wondering is there another way I can go about this rather than digging into creating individual charts and using highchart multiples.
To extract relevant data I have tried
do.call(rbind.data.frame, fit_stl)
as suggested in R Language Collective, which issues
error: invalid list argument: all variables should have the same length
and
df <- data.frame(matrix(unlist(fit_stl), nrow=length(fit_stl), byrow=TRUE))
which issues
Warning message: In matrix(unlist(fit_stl), nrow = length(fit_stl), byrow = TRUE) : data length [668] is not a sub-multiple or multiple of the number of rows [8]
I do not see any problem.
As you do not provide any data, I use nottem(Average Monthly Temperatures at Nottingham, 1920–1939) from in-built {datasets}
X = stl(nottem, "per")
class(X)
#> [1] "stl"
str(X)
#> List of 8
#> $ time.series: Time-Series [1:240, 1:3] from 1920 to 1940: -9.35 -9.86 -6.85 -2.76 3.5 ...
#> ..- attr(*, "dimnames")=List of 2
#> .. ..$ : NULL
#> .. ..$ : chr [1:3] "seasonal" "trend" "remainder"
#> $ weights : num [1:240] 1 1 1 1 1 1 1 1 1 1 ...
#> $ call : language stl(x = nottem, s.window = "per")
#> $ win : Named num [1:3] 2401 19 13
#> ..- attr(*, "names")= chr [1:3] "s" "t" "l"
#> $ deg : Named int [1:3] 0 1 1
#> ..- attr(*, "names")= chr [1:3] "s" "t" "l"
#> $ jump : Named num [1:3] 241 2 2
#> ..- attr(*, "names")= chr [1:3] "s" "t" "l"
#> $ inner : int 2
#> $ outer : int 0
#> - attr(*, "class")= chr "stl"
class(X$time.series)
#> [1] "mts" "ts" "matrix" "array"
This is all you need to build plots:
X$time.series
seasonal trend remainder
Jan 1920 -9.3471980 49.68067 0.266525379
Feb 1920 -9.8552496 49.54552 1.109728805
Mar 1920 -6.8533008 49.41037 1.842931803
Apr 1920 -2.7634710 49.32862 0.134848770
May 1920 3.5013569 49.24688 1.351767558
...
If you like, transform X$time.series to a data.frame.
Also, highcharter::hchart(stl(nottem, "per")) works without any issues.