Hi,

        I am attempting to fit a ARMA/GARCH regression model without success.

                ### ARIMA-GARCH model with regressor ###

### Time series data: A multivariate data set.
cov.ts.dq = cov.ts[1:4,"dq1"][!is.na(cov.ts[,"dq1"])]
cov.ts.day = ts.intersect(dq = diff(q.ts), day = lag(q.ts, -1))

### The following R scripts work:
(summary(no.day.fitr <- garchFit(dq ~ arma(0,3) + garch(1,1), data = 
cov.ts.day)))
(summary(no.day.fitr2 <- garchFit(dq ~ arma(0,3) + garch(1,1), data = 
cov.ts.day,
        include.mean=FALSE)))

### ERROR: I add in the regressor "day".
(summary(no.day.fitr3 <- garchFit(dq ~ day + arma(0,3) + garch(1,1), data = 
cov.ts.day,
        include.mean=FALSE)))
### Error in .garchArgsParser(formula = formula, data = data, trace = FALSE) : 
###  object 'formula.mean' not found

### ERROR: 
day.fitr4 <- garchFit(formula.mean = dq ~ day + arma(0,3),formula.var = 
~garch(1,0), data = cov.ts.day,include.mean = FALSE)
### Error in garchFit(formula.mean = dq ~ day + arma(0, 3), formula.var = 
~garch(1,  : 
### Multivariate data inputs require lhs for the formula.
### Note: If I remove "day" I obtain the same error message.
        
        I would greatly appreciate knowing how to overcome this problem.

        Paul
        [[alternative HTML version deleted]]

______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to