Put core MixAR coefficients into a canonical form, internal function
xxcanonic_coef.Rd
Put core MixAR coefficients into a canonical form, internal function.
Details
This is an internal function called by mixAR
to put MixAR
parameters into a canonical form. "shift"
, "scale"
and
"prob"
are normally vectors with one element for each
component but they may also be given as a single number, in which case
it is taken to be the common value for all parameters and
.canonic_coef
extends them correspondingly. Also, the AR
coefficients may be specified in a number of ways and this function
converts them to the format used by the MixAR classes.
Value
a list with the folowing components:
- order
MixAR order, a vector of integers
- prob
MixAR probabilities, vector of length
length(order)
- shift
MixAR shift parameters, vector of length
length(order)
- scale
MixAR scle parameters, vector of length
length(order)
- arcoef
AR coefficients as a
"raggedCoef"
object
Examples
## missing components are filled with 'filler', extended accordingly
mixAR:::.canonic_coef(list(order = c(2,3)), filler = NA)
#> $order
#> [1] 2 3
#>
#> $prob
#> [1] NA NA
#>
#> $shift
#> [1] NA NA
#>
#> $scale
#> [1] NA NA
#>
#> $arcoef
#> An object of class "raggedCoef"
#> Number of rows: 2
#> Components' lengths: 2 3
#>
#> co_1 co_2 co_3
#> Component_1 NA NA
#> Component_2 NA NA NA
#>
#>
# here 'scale' is replicated, the missing 'shift' is inserted
mo <- list(order = c(2,3), prob = c(0.4, 0.6), scale = 1,
arcoef = list(c(0.5, -0.5), c(1.1, 0.0, -0.5)) )
mixAR:::.canonic_coef(mo, filler = NA)
#> $order
#> [1] 2 3
#>
#> $prob
#> [1] 0.4 0.6
#>
#> $scale
#> [1] 1 1
#>
#> $arcoef
#> An object of class "raggedCoef"
#> Number of rows: 2
#> Components' lengths: 2 3
#>
#> co_1 co_2 co_3
#> Component_1 0.5 -0.5
#> Component_2 1.1 0.0 -0.5
#>
#>
#> $shift
#> [1] NA NA
#>