Hi,
I am trying to use Rsolnp in a portfolio optimization context. While I
know that my current example is solvable as QP, my risk function, which is
variance right now, could become more complicated later. Basically, I want
to start with an equal weighted portfolio, and then minimize risk wit
Hi,
First off, this is a fantastic package, so very well done Alexios. I
am going to use the external regressor option in the rugarch package. In
order to do double check that everything is working properly, I decided to
fit a standard garch(0,1) where i use the lag of the residual as my
ext
There is extensive documentation available on R Cran including:
http://cran.r-project.org/web/packages/rugarch/rugarch.pdf
and also a very nice vignette here that explains what the rolling
estimation does (section 5):
http://cran.r-project.org/web/packages/rugarch/vignettes/Introduction_to_the_rug
?ugarchroll
Change the n.ahead parameter in the function. You can specify whether
there is a mean model or not with ugarchspec. Follow the the example
provided in the ugarchroll example, and then explore the resultant object
(called sp500.bktest).
On Tue, Jan 31, 2012 at 7:36 AM, Papa Senyo w
, VaR.alpha = c(0.01, 0.025, 0.05))
On Tue, Jan 31, 2012 at 8:49 AM, Papa Senyo wrote:
> Please, can you just cite an example via mail to me may a reproducible
> one. Kindly hope to hearing from you soon.
> Kind regards
> Papa
>
> --
> *Da
Hi,
I am running into a memory issue when I try to run a mean-CVaR
optimization on a large number of scenarios. I am running R 32-bit on
windows (see sessionInfo() output below). I have attached code to display
the problem below (using a sample from a random multi-variate normal
distribution a
%*% rep(1/3,3), alpha)
> # we are estimating [VaR AND the weights]
> sol = nloptr(
> x0 = c(q1,rep(1/3,3)),
> eval_f = func.cvar,
> eval_grad_f = grad.cvar,
> lb = c(-1, rep( wmin, 3) ),
> ub = c( 0, rep( wmax, 3) ),
> opts = ctrl,
> eval_g_ineq = func.ineq,
> eval_jac_g_ineq
, samps, hurd){
z <- samps %*%w
tail.v <- z[z<=hurd]
if(length(tail.v) == 0) tail.v <- 0
-(length(tail.v)/length(z))*mean(tail.v)
}
cvarStar(rep(1/3,3), samps, -.05)
On Wed, Feb 15, 2012 at 6:01 PM, Robert Harlow wrote:
> Wow, this is fantastic Alexios. Thank you ve
tensively covered (for the problem of the partial
> moment to the power 1 there is an LP representation, for the power 2 a QP
> representation, and for the rest or general LPM problem you can definitely
> represent it as a proper NLP).
>
> Regards,
> Alexios
>
> On 16/0
This seems more like an R-Help question than an R-Finance Question, but
here goes:
For the year example, you could use the month.day.year from the chron
package (here I chose to add a column to the data frame that adds the day):
library(xts)
library(chron)
n <- 100
df <- data.frame(Date = as.POSIX
Robert,
First off, this is a pretty interesting phenomenon, and let me start
off by saying that I have not figured out what is going on. However, I do
have some more information to contribute:
I am unable to recreate the bug in MATLAB, which uses a different
subroutine from LAPACK: DGESVD wher
hich
> > R's 'svd' throws an error because of non-convergence
> > in the algorithm. Changing the order of the rows
> > makes it go away.
> >
> > Pat
> >
> >
> > On 28/06/2012 16:52, Robert Harlow wrote:
> >
> >>
depmixS4 works pretty well. "Markov Regime Switching" can be implemented in
a variety of ways...read the vignette to learn how to specify the model you
want.
On Fri, Nov 2, 2012 at 6:56 AM, jaimie villanueva <
jaimie.villanu...@gmail.com> wrote:
> Hi R users,
>
> I'm wondering if is there a good
Hi,
I apologize if this is trivially simple or exists elsewhere, but I
couldn't find exactly what I am looking for through searching the web.
Basically, I would like to display what the feasible set of portfolios are
for a given linearly constrained problem. What the code below does, is,
given
Hi,
I am ran the following example with RQuantLib's AsianOption function:
AsianOption("geometric", "put", underlying=80, strike=85, div=-0.03,
riskFree=0.05, maturity=0.25, vol=0.2)
and it worked fine.
If I change the averageType parameter to "arithmetic", it crashes R.
AsianOption("arithmetic
I should add that I realize the arithmetic version is not fully written in
RQuantLib, I am mainly wondering why it crashes instead of throwing an
error or warning.
Thanks,
-Bob
On Fri, Dec 21, 2012 at 11:19 AM, Robert Harlow wrote:
> Hi,
> I am ran the following example with RQuan
Hi,
I was wondering where the rmgarch.tests folder is. In the vignette, it
says that examples are in the "rmgarch.tests folder of the source," but I
do not know exactly what that means and cannot find it in the directory
where the library is installed (on windows).
Also, I wanted to ask if any
it from CRAN and unpack/extract.
>
> I've installed it on Fedora 19 without problem...you do need to follow the
> instructions of Rmpfr/Bessel/fftw on installing the required dependencies
> first (see their readme files in the src installation).
>
> Regards,
> Alexios
&g
It looks like Rscript loads your .Rprofile file on startup, which is
located in ~/.Rprofile. You could just add library(methods) to your
.Rprofile file and it should fix the issue.
http://stackoverflow.com/questions/19680462/rscript-does-not-load-methods-package-r-does-why-and-what-are-the-conseq
Check out ?diff and ?lag. Also, it's kind of a personal preference, but
xts is a pretty attractive alternative to timeSeries that many of us use.
library(xts)
set.seed(10)
x <- xts(rnorm(10), Sys.Date()+1:10)
diff(x)
x - lag(x)
On Fri, Jun 20, 2014 at 8:14 AM, nacho wrote:
> Hi guys,
>
> Fir
Walmir,
In order for anyone to help you, please follow the posting guide (see
link below). You need to put a reproducable example in your question and
attaching .csv files is generally not necessary, so try to avoid it if
possible. What did you try? Which specific function "didn't work"? Be
spe
The issue has nothing to do with fPortfolio. You are not parsing your
input file correctly, run:
indices <- read.table("indices.csv", sep = ";", dec = ",", header = TRUE)
head(indices)
sapply(indices, class)
Notice that your variable types are not numeric, such that it is impossible
to take an a
Lasse,
Your problem is that you have NAs in your external regressor matrix. Since
you lagged it twice, you have 2 NAs. Try the code below.
Thanks for the reproducible example, it made it really easy.
Bob
require(rugarch)
require(quantmod)
x = runif(n = 1000, min = -5, max = 100)
ext.reg_lagged
Charles, my suggestion (which I doubt you will like) is to wrap your calls
to implied volatility root finders (quant lib or any others) in a try block
that returns NA on error. Your example (very little or no time value) can
be common if you are using mid prices from bid-ask quotes. Note that whe
Hi Michael,
Try not to post twice - this is really more of a general R question. To
answer the question, however, turn each element of your resultlist into an
xts (or zoo) object so that you have a list of xts objects (called xtsList
for example.) Then call do.call("merge", xtsList). Also, yo
ave been more desirable)
> By changing the engine I managed a workaround but I thought RQL team (or
> someone else having the same problem) could be interested with that
> non-desirable behavior.
>
> Regards
>
>
> From: Robert Harlow [mailto:rharlo...@gmail.com]
>
All,
I put out a package on CRAN last weekend that extends quadprog so that
constraints involving absolute values (e.g. book size and turnover in a
financial context) are handled in addition to allowing absolute value (L1)
considerations in the objective function, which can handle linear
transact
Have you tried rugarch or just "garch" in the tseries package? Reproducible
examples are always helpful as well.
> On Nov 2, 2017, at 1:56 PM, Curtis Miller wrote:
>
> Hello all,
>
> I have encountered bad behavior in fGarch's garchFit() function used for
> estimating the parameters of a GAR
Hi Antoine,
Yes - it handles turnover constraints. Check out the second to last
example from ?solveQPXT.
Bob
On Tue, Mar 6, 2018 at 9:03 AM, Antoine wrote:
> Hello Bob,
>
> First of all, thanks for your R Package quadprogXT, I hope that I will use
> it for my turnover constraint (in financial
29 matches
Mail list logo
