Skip to contents

Classes ArmaModel, ArModel and MaModel in package sarima.

Objects from the Class

Classes "ArModel" and "MaModel" are subclasses of "ArmaModel" with the corresponding order always zero.

The recommended way to create objects from these classes is with the functions ArmaModel, ArModel and MaModel. Objects can also be created by calls of the form new("ArmaModel", ..., ar, ma, mean, check). See also coerce-methods for further ways to create objects from these classes.

Slots

center:

Object of class "numeric" ~~

intercept:

Object of class "numeric" ~~

sigma2:

Object of class "numeric" ~~

ar:

Object of class "BJFilter" ~~

ma:

Object of class "SPFilter" ~~

Extends

Class "ArmaSpec", directly. Class "VirtualArmaModel", directly. Class "ArmaFilter", by class "ArmaSpec", distance 2. Class "VirtualFilterModel", by class "VirtualArmaModel", distance 2. Class "VirtualStationaryModel", by class "VirtualArmaModel", distance 2. Class "VirtualArmaFilter", by class "ArmaSpec", distance 3. Class "VirtualAutocovarianceModel", by class "VirtualArmaModel", distance 3. Class "VirtualMeanModel", by class "VirtualArmaModel", distance 3. Class "VirtualMonicFilter", by class "ArmaSpec", distance 4.

Methods

modelOrder

signature(object = "ArmaModel", convention = "ArFilter"): ...

modelOrder

signature(object = "ArmaModel", convention = "MaFilter"): ...

modelOrder

signature(object = "ArmaModel", convention = "missing"): ...

modelOrder

signature(object = "SarimaModel", convention = "ArmaModel"): ...

sigmaSq

signature(object = "ArmaModel"): ...

Author

Georgi N. Boshnakov

Examples

arma1p1 <- new("ArmaModel", ar = 0.5, ma = 0.9, sigma2 = 1)
autocovariances(arma1p1, maxlag = 10)
#> An object of class "Autocovariances"
#>       Lag_0       Lag_1       Lag_2       Lag_3       Lag_4       Lag_5 
#> 3.613333333 2.706666667 1.353333333 0.676666667 0.338333333 0.169166667 
#>       Lag_6       Lag_7       Lag_8       Lag_9      Lag_10 
#> 0.084583333 0.042291667 0.021145833 0.010572917 0.005286458 
autocorrelations(arma1p1, maxlag = 10)
#> An object of class "Autocorrelations"
#>           0           1           2           3           4           5 
#> 1.000000000 0.749077491 0.374538745 0.187269373 0.093634686 0.046817343 
#>           6           7           8           9          10 
#> 0.023408672 0.011704336 0.005852168 0.002926084 0.001463042 
partialAutocorrelations(arma1p1, maxlag = 10)
#> An object of class "PartialAutocorrelations"
#>       Lag_0       Lag_1       Lag_2       Lag_3       Lag_4       Lag_5 
#>  1.00000000  0.74907749 -0.42512100  0.29448975 -0.22337052  0.17832394 
#>       Lag_6       Lag_7       Lag_8       Lag_9      Lag_10 
#> -0.14703750  0.12391932 -0.10606466  0.09181545 -0.08015626 
partialAutocovariances(arma1p1, maxlag = 10)
#> An object of class "PartialAutocovariances"
#>           0           1           2           3           4           5 
#>  3.61333333  1.18790975 -0.55232849  0.34942744 -0.25181673  0.19464072 
#>           6           7           8           9          10 
#> -0.15702172  0.13030164 -0.11027274  0.09465347 -0.08210298 

new("ArmaModel", ar = 0.5, ma = 0.9, intercept = 4)
#> An object of class "ArmaModel"
#> intercept:  4 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.5
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.9
new("ArmaModel", ar = 0.5, ma = 0.9, center = 1.23)
#> An object of class "ArmaModel"
#> mean:  1.23 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.5
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.9

new("ArModel", ar = 0.5,  center = 1.23)
#> An object of class "ArModel"
#> mean:  1.23 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.5
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  0 
#> Coefficients: 
#> numeric(0)
new("MaModel", ma = 0.9,  center = 1.23)
#> An object of class "MaModel"
#> mean:  1.23 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  0 
#> Coefficients: 
#> numeric(0)
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.9

# argument 'mean' is an alias for 'center':
new("ArmaModel", ar = 0.5, ma = 0.9, mean = 1.23)
#> An object of class "ArmaModel"
#> mean:  1.23 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.5
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.9

##  both center and intercept may be given
## (the mean is not equal to the intercept in this case)
new("ArmaModel", ar = 0.5, ma = 0.9, center = 1.23, intercept = 2)
#> An object of class "ArmaModel"
#> mean:  2.563333 
#> intercept:  2 (full intercept:  3.845 ) 
#> sigmaSq:  NA 
#> 
#> slot "ar":
#> An object of class "BJFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.5
#> 
#> slot "ma":
#> An object of class "SPFilter"
#> order:  1 
#> Coefficients: 
#> [1] 0.9

## Don't use 'mean' together with 'center' and/or 'intercept'.
##    new("ArmaModel", ar = 0.5, ma = 0.9, center = 1.23, mean  = 4)
##    new("ArmaModel", ar = 0.5, ma = 0.9, intercept = 2, mean  = 4)
## Both give error message:
##   Use argument 'mean' only when 'center' and 'intercept' are missing or zero