VARIMA: identification = "scalar_components" not working even if in the code is present

[matteopelagatti]

The parameter identification in the VARIMA function takes only the values "none" or "kronecker_indices". If I set identification == "scalar_components", I get an error suggesting I should pick the above options.
However, if I read the code in VARIMA.R (function train_varima, line 31), I read

} else if (identification == "scalar_components") {
with(
MTS::SCMid2(yd, max(p), max(q)),
MTS::SCMfit(
yd,
SCMorder, Tmatrix,
include.mean = "(Intercept)" %in% colnames(specials$xreg[[1]])
)
)
}

which is the code to estimate the scalar components VARIMA.

Can you fix this and allow for the scalar-components estimation?

[mitchelloharawild]

MRE

library(fable)
#> Loading required package: fabletools
#> Registered S3 method overwritten by 'tsibble':
#>   method               from 
#>   as_tibble.grouped_df dplyr
phi=matrix(c(0.2,-0.6,0.3,1.1),2,2); sigma=diag(2)
m1=MTS::VARMAsim(300,arlags=c(1),phi=phi,sigma=sigma)$series
colnames(m1) <- c("X","Y")
tsbl <- as_tsibble(as.ts(m1), pivot_longer = FALSE)

tsbl |> 
  model(
    VARIMA(vars(X, Y), identification = "scalar_components")
  )
#> Warning: 1 error encountered for VARIMA(vars(X, Y), identification = "scalar_components")
#> [1] non-conformable arrays
#> # A mable: 1 x 1
#>   `VARIMA(vars(X, Y), identification = "scalar_components")`
#>                                                      <model>
#> 1                                               <NULL model>

^{Created on 2026-01-15 with reprex v2.1.1}/