vws: Eviews style regression specification

In bjorn81/vws: Multivariate time series regression in R

Description

vws Allows the specification of a regression with lag(x, k), diff(x, p) as well as ARMA errors. structure

Usage

vws(formula, ar = 0, ma = 0, data, method = "ML", subsett = NULL,
  datecolumn = NULL, lmclass = FALSE, ...)

Arguments

formula	Regression formula to be evaluated
ar	number of autoregressive (AR) terms to be included in regression
ma	Number of moving average (MA) terms to be included in regression
data	The dataframe which contains the variables used
method	Estimation method (CSS-ML, ML, CSS)
datecolumn	Optional column of class Date
lmclass	if true and no ARMA errors specifed, object returned is of class lm

Details

Returns an object of class "vws", which inherits from class "arima". If no ARMA structure of errors is specified and "lm" = true , the returned object is of class "lm". Generic functions plot and summary can be used.

Value

a list of x elements:

coef	a vector of AR, MA and regression coefficients, which can be extracted by the coef method.
sigma2	the MLE of the innovations variance.
var.coef	the estimated variance matrix of the coefficients coef, 'which can be extracted by the vcov method.
loglik	the maximized log-likelihood (of the differenced data), or the approximation to it used.
aic
arma	A compact form of the specification, as a vector giving the number of AR, MA, 'seasonal AR and seasonal MA coefficients, plus the period and the number of non-seasonal and seasonal differences.
residuals	the fitted innovations.
fitted	Fitted values of the formula
dates	Optional dates for the fitted series

Author(s)

Bjorn Backgard

Examples

#Generate some data: y0 follows an ARMA process conditional on x1, x2
error.model=function(n){rnorm(n, sd=.2)}
y0 <- arima.sim(model=list(ar=0.6, ma = -0.9), n=100,
               n.start=200, start.innov=rnorm(200, sd=.2),
               rand.gen=error.model )
x1 <- arima.sim(model=list(ar=0.95), n=100,
               n.start=200, start.innov=rnorm(200, sd=.2),
               rand.gen=error.model)
x2 <- arima.sim(model=list(ar=0.95), n=100,
               n.start=200, start.innov=rnorm(200, sd=.2),
               rand.gen=error.model)
y <- y0 + 0.5*x2 + 0.2 * x1
par(mfrow = c(2,2)) #Plot the variables
plot(y0)
plot(x1)
plot(x2)
plot(y)
dt <- data.frame(y,x1,x2)
par(mfrow = c(1,1))
library(vws)
vws.mod <- vws(y~x1+x2, ar = 0, ma = 0, data = dt) #Fit model
summary(vws.mod) #Get regression output
plot(vws.mod, 2) #Examine autocorrelation structure, ARMA terms needed
vws.mod <- vws(y~x1+x2, ar = 1, ma = 1, data = dt) #Fit new model
summary(vws.mod) #Get regression output
plot(vws.mod, 2) #Examine autocorrelation structure
plot(vws.mod, 3) #Check AR and MA roots
tsdiag(vws.mod) #No evidence of remaining autocorrelation in errors
plot(vws.mod, 1) #Check fit of final model

bjorn81/vws documentation built on May 16, 2019, 4:54 p.m.