rstandard.KFS: Extract Standardized Residuals from KFS output

Description

Extract Standardized Residuals from KFS output

## S3 method for class 'KFS':
rstandard(model, type = c("recursive", "deviance", "pearson",
  "state"), ...)

model

KFS object

type

Type of residuals. See details.

...

Ignored.

For object of class KFS, several types of standardized residuals can be computed:

'recursive': One-step ahead prediction residuals defined as$$v_{t,i})/\sqrt{F_{i,t}},$$with residuals being undefined in diffuse phase. Only supported for fully Gaussian models.
'pearson': Standardized Pearson residuals$$(y_{t,i}-\theta_{t,i})/\sqrt{V(\mu)_{t,i}\phi_i\sqrt{1-h_{t,i}}}, \quad i=1,\ldots,p,t=1,\ldots,n,$$where$V(\mu_{t,i})$is the variance function of the model,$\phi_i$is the dispersion parameter and$h_{t,i}$is the hat value. For gaussian models, these coincide with the smoothed$\epsilon$disturbance residuals.
'state': Residuals based on the smoothed disturbance terms$\eta$are defined as$$L^{-1}_t \hat \eta_t, \quad t=1,\ldots,n,$$where$L_t$is the lower triangular matrix from Cholesky decomposition of$V_{\eta,t}$.
'deviance': Deviance residuals.