[R-SIG-Finance] Error check on "pspd" function from SPD package

[alexios galanos]

Once again, it's just a matter of reading the documentation.

1. rugarch "prefers" xts
2. spd "requires" numeric:

>From the documentation of pspd:
###########################
x,q    [pspd,dspd] -
a numeric vector of quantiles.
###########################

So I don't see any problem as the following works just fine:

library(rugarch)
library(spd)
library(xts)
data(sp500ret)   # in the "rugarch" package
sp = xts(sp500ret[,1], as.Date(rownames(sp500ret)))
# standard GARCH model:
spec =
ugarchspec(mean.model=list(armaOrder=c(0,0)),distribution.model="std")
tempgarch = ugarchfit(spec=spec,data=sp,solver="hybrid")
# fitting the SPD:
std.resid.sp = as.numeric(residuals(tempgarch,standardize=T))
fit.sp = spdfit(std.resid.sp, upper=0.9, lower=0.1, tailfit="GPD",
type="mle", kernelfit="information="observed")
sp.unif.variates = pspd(std.resid.sp, fit.sp, linear=T)


-A

On 14/10/2015 02:20, Gareth McEwan wrote:
> Hi all
>
> I used the "pspd" function from the SPD package last year and I'm running
> the code again, but it seems to throw an error now.
>
> I checked with data from the "rugarch" package and the error is still
> thrown.
>
> data(sp500ret)   # in the "rugarch" package
> sp = sp500ret
> class(sp)   # is a "data.frame" class
> sp=as.ts(sp)  #  I tried with and without this step
>
> # standard GARCH model:
> spec = ugarchspec(variance.model=list(model="sGARCH",garchOrder=c(1,1),
>             submodel=NULL,external.regressors=NULL,variance.targeting=F),
>
> mean.model=list(armaOrder=c(0,0),include.mean=T,external.regressors=NULL),
>             distribution.model="std")
> tempgarch = ugarchfit(spec=spec,data=sp,solver="hybrid")
>
> # fitting the SPD:
> (1)
> std.resid.sp = residuals(tempgarch,standardize=T)
> class(std.resid.sp) #  "xts"  "zoo"
>
> (2)
> stdResid.sp = as.ts(std.resid.sp)  #  NOTE: I tried (1) with and (2)
> without this step
> class(stdResid.sp)  #  "ts"
>
> # the problem:
> (1)
> fit.sp = spdfit(std.resid.sp, upper=0.9, lower=0.1, tailfit="GPD",
> type="mle", kernelfit="normal",
>             information="observed", title=NULL, description=NULL)
> sp.unif.variates = pspd(std.resid.sp, fit.sp, linear=T)
>
> Error in UseMethod("pspd") :
>   no applicable method for 'pspd' applied to an object of class "c('xts',
> 'zoo')"
>
> (2)
> fit.sp = spdfit(stdResid.sp, upper=0.9, lower=0.1, tailfit="GPD",
> type="mle", kernelfit="normal",
>             information="observed", title=NULL, description=NULL)
> sp.unif.variates = pspd(stdResid.sp, fit.sp, linear=T)
>
> Error in UseMethod("pspd") :
>   no applicable method for 'pspd' applied to an object of class "ts"
>
> >From sessionInfo(), I have SPD package version as: spd_2.0-1 (i.e., it
> seems to be up to date).
>
> I don't think Can anyone help out with a solution?
>
> Many thanks
> Gareth
>
>
> Additional:
> packageDescription("spd")
> Package: spd
> Type: Package
> Title: Semi Parametric Distribution
> Version: 2.0-1
> Date: 2015-07-02
> ...
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Finance at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> -- Subscriber-posting only. If you want to post, subscribe first.
> -- Also note that this is not the r-help list where general R questions should go.
>