On Nov 17, 2012, at 11:53 PM, Eric Thungstom wrote:
> Getting this message when I try to install xtsExtra.
>
>> install.packages('xtsExtra', source = "http://r-forge.r-project.org";)
> Installing package(s) into ‘C:/PROGRA~1/R/R-212~1.0/library’
> (as ‘lib’ is unspecified)
> Warning in install
On Dec 28, 2012, at 5:17 AM, "Robert A'gata" wrote:
> Hi,
>
> I would like to confirm my intuition about backtesting. My strategy is a
> simple moving average crossing over (something similar). I open a long
> position of 1 unit if fast line crosses above the slow one and vice versa.
> After s
Intraday data usually requires a paid data subscription. Do you have any
particular source in mind?
MW
On Jan 11, 2013, at 8:34 PM, "Simone Gogna" wrote:
> Dear R users,
> I hope this is not an already discussed topic. I tried with RSiteSearch(“high
> frequency italian market data”) but I was
Moved to r-sig-fin.
On Feb 15, 2013, at 6:43 AM, Aleksi Mattila wrote:
> Dear R experts,
>
> After updating PerformanceAnalytics from 1.0.4.4. to 1.0.5.2 I have
> encountered a following issue (reproducible example below). I am not quite
> sure whether this is due to my sloppy coding (any hi
What exactly do you mean by end of day intraday?
Regardless, I assume the answer is 'yes, many by subscription: some cheap: few
free.'
MW
On Mar 11, 2013, at 22:26, Bill Blount wrote:
> is there a source for end of day intraday data (i am assuming that yahoo is
> daily only)?
>
> Thanks. Bi
On Apr 25, 2013, at 20:00, Mark Knecht wrote:
> Hi,
> Is there a good home page or Wiki for how folks are approaching
> trading within R using quantStrat, TradeAnalytics, or something else?
>
> I'm working in RStudio/R-3.0-64-bit and am pretty much a putz when
> it comes to R programming.
To echo what Deo is saying, if the question is 'how to do X with ts objects'
the answer is almost always going to be 'Use xts instead'. It's not quite a
drop in for one or two things, but its basically always going to be better for
whatever you may wish.
M
On May 30, 2013, at 15:34, Deo Jaisw
Take a look at runquantile from caTools.
http://svitsrv25.epfl.ch/R-doc/library/caTools/html/runquantile.html
Though I agree this would make sense in TTR as well...
Michael
On Jul 27, 2013, at 5:03, Andreas Voellenklee wrote:
> Hi,
>
> first I would like to say thanks to the developer(s) of
On Sep 30, 2013, at 17:52, Preston Li wrote:
> if (!is.loaded('etfdata')) data(etfdata)
Totally off topic: that doesn't do what you think it does. You're looking for
exists()
M
___
R-SIG-Finance@r-project.org mailing list
https://stat.ethz.ch/mail
Something like:
cor(tail(merge(ret.SPY, lag(ret.MDY)), 10))
would work.
Merge your series into a single object, take the last ten obs (tail(x, 10)) and
get the correlation matrix thereof.
MW
On Oct 29, 2013, at 7:49, "Martin Bauer" wrote:
> hi,
>
> How can I calculate the correlation on
If you are using my xtsExtra::plot.xts or Jeff's xts::plot.xts, the plot engine
uses POSIXct (unix epoch seconds) to align the x-axis; that sometimes throws
folks off. Take a look at xtsExtra::lines.xts for an example.
Michael
On Nov 24, 2013, at 18:05, fernando wrote:
> Please,
> I am ploti
On Jan 27, 2014, at 8:01, arnaud gaboury wrote:
> On Mon, Jan 27, 2014 at 1:52 PM, Anil Bishnoie wrote:
>> Hi All,
>> As Blopen API,emul and rblapi etc. are very useful ,It will be gr8
>> if a VM be formed and stacked in pub domain for download for future
>> testing/use.
>>
> Sorr
Not about finance. Please don't cross post.
Michael
On Apr 10, 2014, at 13:56, ì¥ì¹ì± wrote:
>
> My data as follows.
> group Y1 Y2...Y100 X1 X2
> A
> A
>
> ...
> A
> B
> B
>
> ...
>
> B
> C
>
> ...
>
> Z
>
>
> I want to run regression analysis. The model as follow
> Y1A=a+b1*X1+b2*X
> On Jun 29, 2014, at 4:07 PM, Christofer Bogaso
> wrote:
>
> Hi again,
>
> I would like ask a small question however not really related to R.
>
> We all know that non-arbitrage Forward price of any underlying (except
> perhaps Interest Rate) is just the spot price plus the cost of carry.
>
ort themselves out.
> On Jun 29, 2014, at 5:09 PM, Christofer Bogaso
> wrote:
>
> Ofcourse Electricity can not be stored or storage cost would be
> extraordinarily high. Therefore it would have zero CY. I feel Power
> prices should always be in contango
>
> On Mon, Jun 30, 2014 a
> On Jun 29, 2014, at 18:32, Christofer Bogaso
> wrote:
>
> I defined backwardination as forward curve is downward sloping as
> compared to spot. So you may term it as you said 'forward prices
> increasing/decreasing'. I felt this is bit different than what you
> said as 'difference between expe
> On Jul 14, 2014, at 6:38, u0...@wolke7.net wrote:
>
> Dear R-SIGs,
>
> Is there anybody out there working with parma ?
> It's a great package.
> I can use parts of it,
> but don't understand everything.
>
> Does anybody know the function of parameter "targetType"
> in method "parmaspec()" ?
>
> On Feb 15, 2015, at 7:32 PM, Robert Sherry wrote:
>
> Thank you for your response. The symbol GLD is not really going to work
> for me because it is my understanding how much gold GLF represents
> varies over time. Therefore,
> I thought I would use IKTrading. However, when I try installing
If you have data from each trading day, something like
quantmod::apply.monthly(x, NROW)
might work.
If your data is spotty, you need to tell us more about your data provider.
As always, http://sscce.org is your friend.
Michael
> On Mar 27, 2015, at 1:39 PM, Pankaj K Agarwal via R-SIG-Financ
I think you're overthinking. Look at the auto.assign arg to getSymbols.
for(x in tickers){
data <- getSymbols(x, auto.assign=FALSE)
# Process `data` ...
}
Michael
> On Apr 6, 2015, at 12:59 PM, Samuel Wilson wrote:
>
> As an experienced R Programmer, I should be able to figure this out
> On Oct 26, 2015, at 2:47, Amelia Marsh via R-SIG-Finance
> wrote:
>
> Dear Forum,
>
> I have series of say 100 (say equity) instrument prices. From these prices,
> for each of these 100 instruments, I generate returns using ln(current price
> / previous price).
>
> Assuming originally I
> On Oct 26, 2015, at 9:21, Am Gut wrote:
>
> Dear R Users,
>
> I am trying to calculate trailing month and annual returns with daily data.
> The data is already in returns, and not in price. I have been looking for a
> handy function to calculate trailing returns where I can specify the look
>
On Mon, Oct 26, 2015 at 9:53 AM, Am Gut wrote:
> Good Morning Michael,
>
> I have simple return data, in a daily periodicity. I am essentially trying
> to calculate the trailing returns for say 252 periods, assuming I am trying
> to look at trailing 12 month returns. So I have been trying to use
On Tue, Nov 24, 2015 at 6:31 PM, Nick White wrote:
> You might want to check out the derivation of the Thorp /
> Black-Scholes-Merton formula as it deals with essentially the same
> concepts...
>
> On Wed, Nov 25, 2015 at 11:27 AM, Ernest Stokely
> wrote:
>
>> Maybe a naive question but given the
?rq=1
(though note there's a mistake in the latter)
Hope this helps,
Michael
On Tue, Nov 24, 2015 at 7:23 PM, Michael Weylandt
wrote:
> On Tue, Nov 24, 2015 at 6:31 PM, Nick White wrote:
>> You might want to check out the derivation of the Thorp /
>> Black-Scholes-Merton
The Hessian is the matrix of second derivatives
(https://en.wikipedia.org/wiki/Hessian_matrix) -- in scalar terms,
you're finding a point where the second derivative is zero and then
trying to divide by the second derivative to calculate the step size.
I haven't gone through your code in any detai
the meaning of "comfortably," but I can create a
> Cholesky decomposition without it blowing up, which is usually how I can tell
> if I have done something stupid. Well, stupider than normal.
>
> -Original Message-
> From: Michael Weylandt [mailto:michael.weyla...
14, 2015 at 10:10 PM, Michael Ashton
wrote:
> Well, not sure whether this makes any sense but kappa(cov.mat) gives me
> 11148245007.
>
> That seems large. But I am not sure what it is supposed to be.
>
> -----Original Message-
> From: Michael Weylandt [mailto:michae
You can use a sink(NULL) (R equivalent of >/dev/null) to suppress all
output temporarily.
As far as I know, there's no simple way to just catch this one cat().
On Tue, Dec 29, 2015 at 12:34 AM, George Kumar wrote:
> Hi all,
>
> I am using getSymbols to with the following options:
>
> setDefaults
On Fri, Jan 1, 2016 at 2:41 AM, Evelyn Nyamadi via R-SIG-Finance
wrote:
>
> Dear All,
> Please, I would like to use the nnetar in the forecast package to do both in
> sample and out of sample forecasting.
>
> But what is package does is not very clear to me. It gives you rather an
> extrapolate
Take a look at Section 2.2.3 of the package vignette [1, 2].
Michael
[1]
https://cran.r-project.org/web/packages/rugarch/vignettes/Introduction_to_the_rugarch_package.pdf
[2]
https://bitbucket.org/alexiosg/rugarch/src/1bf0f673286b22124fe3a55dfd79d94b3169fb6b/vignettes/rugarch.tex?at=master&file
Hi Marco,
Can you put together a minimal reproducible example [1,2] so that it's
easier for others to answer your question?
For this problem, I'd recommend using the edhec data distributed with
PerformanceAnalytics.
Thanks,
Michael
[1]
http://stackoverflow.com/questions/5963269/how-to-make-a-g
>>
>> #The value of the objective function is:
>> *constrained_objective(w=rep(1/4,4), R=indexes, portfolio=myCVaRObjSpec)
>> #* ES *0.1217594*
>>
>> #This should be the CVaR of the equal-weight portfolio as computed by the
>> function ES in the P
I'm not sure what you're trying to do. What should the `newdata`
argument be used for here? (The shocks/innovations, external
regressors, etc.?)
On Tue, Nov 1, 2016 at 4:57 AM, Be Water wrote:
> ## R version 3.3.1 (2016-06-21)
> ## Platform: x86_64-w64-mingw32/x64 (64-bit)
> ## Running under: Win
Hi Pierre,
I'm willing to take a look, but could you format the data in a way
that would make it a bit easier to use?
The dput() function can produce a representation which is safe for
emailing and can be read directly back into R.
Cheers,
Michael
On Tue, Nov 1, 2016 at 4:19 AM, wrote:
>
> I
The fGARCH predict() method will do that with the n.ahead argument; no
need to supply newdata. Try running
> example('predict-methods', package="fGARCH")
to see an example (with plots!).
The standard ARMA and GARCH models don't need external data to make
predictions: their predictions are made b
Hi FMH,
Perhaps you could set up a minimal reproducible example [1,2] to help
us answer your question.
Note that the biwavelet [sic] package has lots of examples that you
can access. The most useful ones are probably found by running
example(wt)
and
example(biwavelet)
Cheers,
Michael
[1]
ht
ach(data)
>
> library(biwavelet)
> dat <- cbind(1:length(data$Temp),data$Temp)
> wt.dat <- wt(dat)
>
> par(mfrow=c(2,1))
> plot.ts(dat[,2])
> plot(wt.dat, plot.cb=TRUE, plot.phase=FALSE)
>
> Thank you very much for your guideline.
>
> Fir
>
>
ng too many inferences. Since this is a somewhat thorny issue, my
> recommendation is that you handle the plot call yourself so you know exactly
> what you are getting.
>
> Hope this helps and good luck getting started with R,
>
> Michael Weylandt
>
>
>
>
>
>
> On
na.locf() will handle the second part of your question. I've not been super
successful with the maxlag optional argument, however, so do be wary if you
use that one.
Michael
On Sun, Aug 28, 2011 at 9:42 PM, Worik Stanton wrote:
> Forgive me if this has been asked/answered.
>
> I have some data t
You might try resending this message while actually including the error
message if you want a more substantial reply...
Michael Weylandt
On Mon, Sep 12, 2011 at 4:18 AM, krisan haria wrote:
> Hi
>
> Basic Question
>
> I havev R .2.13.1. Im trying to use RBLOOMBERG for
of 2 variables:
> $ V1: Factor w/ 1 level "spy,20110815": 1 1 1 1 1 1 1 1 1 1 ...
> $ V2: Factor w/ 23399 levels
> "01:00:00,119.92,119.92,119.92,119.92,0,0,0.0 , ...
>
> I need to access the individual elements of each row of V2.
>
> Please let me know,
> It
There's no automatic function for this sort of request that I know of, but
it shouldn't be hard to write one:
require(quantmod)
getSymbols('^GSPC',from='1990-01-01')
CollapseOpCl <- function(P){
stopifnot(all(has.Op(P),has.Cl(P)))
O <- Op(P); time(O) <- as.POSIXlt(time(O)); time(O)$hour <
No idea why, but just chiming in to say same problem on my (mac) machine:
> sessionInfo()
R version 2.13.1 (2011-07-08)
Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
locale:
[1] C/en_US.UTF-8/C/C/C/C
attached base packages:
[1] stats graphics grDevices utils datasets methods bas
at 11:27 PM, chandra bajpai wrote:
> HI Michael...thanks for the quick reply, I'm to to R, how does one
> report such a bug?
> I wonder if it's the bug exists in 2.12 at all.
>
> Thanks,
> Chandra
>
> On Tue, Sep 27, 2011 at 11:21 PM, R. Michael Weylandt
>
I don't know if Josh/Brian/Jeff got you an answer on this, but on my
version of PerformanceAnalytics (1.0.3.2), these different VaR methods
seem to have been collapsed into VaR with different methods available.
Perhaps your reference is for an older version of the package in which
these have been d
I don't have the OP's data, but, just for the record, the same
phenomenon can be seen with SPY returns.
library(quantmod)
getSymbols("SPY")
R = na.omit(ROC(Ad(SPY)))
identical((ES(R, p = 0.99)-VaR(R, p = 0.99))[1,1],0)
TRUE
And thanks for the insightful explanation. Do you happen to have the
Jor
library(quantmod)
getSymbols("CBA.AX")
Hope that helps,
Michael
On Tue, Oct 25, 2011 at 6:13 AM, Fan wrote:
> Hi there,
> I am trying to load data using getSymbols {quantmod}, however, it looks to
> me this function can only load data from US exchange, since I trade stocks
> in ASX, so I tried
This isn't specifically financial so any follow up is perhaps better
done on the general R-help list, but try demo(persp) to see some
example code that's built in. If I remember correctly, demo(lattice)
also has at least one 3d example.
Michael
On Wed, Oct 26, 2011 at 1:43 PM, financial engineer
Perhaps you should load the lattice package first with
library(lattice)I just confirmed it has 2 3d examples in it.
As to specific formatting advice, I'd usually prefer delta/moneyness
than straight strike if you were preparing the graph for me. Otherwise
there's one very large source of varia
Another thought, you might want to look at the rgl package: it allows
for interactive 3d graphics (though with a little bit of work) which
can add a potentially useful level of exploratory power. I haven't
used it much but would also be interested in learning more: if you'd
be interested, I'd love
I think you mean to put get(ticker) in the rbind() to get the xts object, not
the name.
Michael
On Nov 6, 2011, at 12:37 PM, Hans Grapenthin wrote:
> Hello
>
> this does it for one stock. I am struggling when I insert this expression in
> a
> loop that worked before:
>
> tickers <- c("MS
If no one here knows, it may be worth it to contact the package
maintainer directly. See maintainer("vars")
Michael
On Tue, Nov 8, 2011 at 9:44 AM, Richard Saba wrote:
> Does anyone know if the bootstrap CI intervals generated by the irf()
> function in the " vars" package are bias corrected?
You can probably specify the problem with model.matrix and use lm.fit
directly, but what's probably even better is to remember that the
slope can be calculated as correlation * std_y / std_x for this simple
case of one independent variable and implement directly . E.g.,
something like
apply(Data,
plot.xts()
chartSeries() if you are using financial data?
It's going to depend on your (unstated) form of time series, but
generally R isn't set up for interactive graphics. (chartSeries being
the only time-series specific exception I know; you can also do Rgl
and Rggobi stuff with a little bit o
What version of forecast do you have loaded (use sessionInfo() after
library(forecast) to check)? If it was only uploaded very very
recently, it might not have propagated to your mirror. It should be
3.14.
If your mirrors don't have it built, you might have to download a
source package and do a co
Have you looked at the quantstrat demos? (If you don't know what I
mean, google quanstrat and download the required packages from
R-forge)
Michael
On Sat, Dec 24, 2011 at 4:09 PM, Michael wrote:
> Hi all,
>
> Happy holidays!
>
> I would like to ask a question about using R to backtest strategies
Slightly beyond the scope of your question, but this might be of
interest: http://quanttrader.info/public/betterHedgeRatios.pdf
Michael
On Fri, Jan 6, 2012 at 2:08 PM, Mark Breman wrote:
> Hello List,
>
> I'm testing multivariate timeseries for cointegration using the
> Phillips-Ouliaris test fr
If you have tick/high(-ish) frequency data, you'll have to transform it to
OHLC; there's no way to keep the frequency you have and to make an OHLC object
out of it, since those metrics are by definition aggregates. The easiest way to
do so is to read it in as an xts object then use one of the to
uantmod easily handles tick data.
Michael
On Jan 11, 2012, at 10:10 AM, "Brian G. Peterson" wrote:
> On Wed, 2012-01-11 at 09:49 -0500, R. Michael Weylandt wrote:
>> If you have tick/high(-ish) frequency data, you'll have to
>> transform it to OHLC; there's
Perhaps runMax() and runMin() in TTR would let you code this up easily.
Michael
On Fri, Jan 13, 2012 at 2:30 PM, Michael wrote:
> Hi all,
>
> I am looking for the command/functions for defining price channel:
>
> A price channel is defined by the highest high and lowest low over some
> past numb
Use something like:
period.apply(X, endpoints(X, "minutes", 5), median)
Michael Weylandt
On Thu, Jan 19, 2012 at 1:54 PM, Michael wrote:
> I have also searched "crazily"... and read into both "xts" and "zoo"... but
> couldn't find a solutio
storage.mode(as.POSIXct(123757, origin = "1970-01-01")) # POSIXct
sapply(strptime("1970-01-01 03:43:32.345", format = "%Y-%m-%d
%H:%M:%OS"), storage.mode) #POSIXlt
They are numeric, not character.
Michael Weylandt
On Fri, Jan 20, 2012 at 10:49 AM, Michael wrote:
n?
Or (if you actually read the xts documentation!) there's a (much) better trick:
? `[.xts`
tmp1["::20110920 12:01:33.386168"]
Michael Weylandt
On Fri, Jan 20, 2012 at 11:32 AM, Michael wrote:
> These are char strings:
> "2011-09-20 12:01:33.386168"
> In fa
It's remarkable how often people's R problems are actually Excel problems.
:-)
Michael Weylandt
On Fri, Jan 20, 2012 at 11:35 AM, Michael wrote:
> I followed the "period.apply" advice given previously and it worked...
>
> And the write.zoo problem is just an
gt; (in the previously discussed formatting)... I found the tt$DateTimeStamps is
> actually "factor" ...
>
> So my question is: ordering them like the one I did above,
>
> is it correct?
>
> Thanks a lot!
>
>
>
>
> On Fri, Jan 20, 2012 at 10:44 AM, R.
Short answer: I'm pretty sure it's impossible to remove elements "in place".
Medium answer: If you are worried about memory usage, I might suggest
using the data.table() package. It's pretty efficient for most things,
though you'll loose xts specific time-functionality. You might be able
to wedge
The OP's confusion might also stem from the fact that
?merge
doesn't actually lead to
?merge.xts
which would have been much more helpful here.
I think it's a suboptimal consequence of method dispatch that there
are no pointers to documented methods...I'll try to work up a small
patch to `?` a
I'm not sure time() is very good for what you want to do. It's tied to
R's builtin ts class, which, and it pains me to say this about R,
really isn't very good (at least for finance-y things). I think all
your problems come from that...
Perhaps construct your new index sequence as:
seq(start(alph
ULL
> Error in coredata.xts(x) : currently unsupported data type
>>
>
> As you can see, all of the data from the last day for which there was data
> in the file was processed (in this case by str()), but then it looks like
> apply.data() tries to apply myfun on data for the day afte
R-Forge doesn't build for outdated versions of R. You'll need to
upgrade your R or simply build it from source. I believe it has no
compiled code so it shouldn't be hard.
Michael
On Tue, Jan 31, 2012 at 11:09 AM, Michael wrote:
> Hi all, good morning and good evening!
>
> Could you please help
It looks like you need to wrap coredata() because some funny
arithmetic is happening when the xts-ness is preserved.
e.g.
findPeaks(coredata(Ad(SPY)), 5)
But this is perhaps a less-than-desirable feature. I'll patch it and
send it to Josh.
Michael
On Thu, Feb 9, 2012 at 9:56 PM, Michael wrote
Happy to follow procedure.
Thanks for jumping on it so quickly.
Michael Weylandt
On Thu, Feb 9, 2012 at 10:40 PM, Jeffrey Ryan wrote:
> On Thu, Feb 9, 2012 at 9:00 PM, R. Michael Weylandt
> wrote:
>> It looks like you need to wrap coredata() because some funny
>> arithmetic
Peaks(cc, 5)
> points(p, cc[p])
> points(p, cc[p])
>
>>
>
> p
>
> [1] 3 22 41
>
>
>
> It only picks up 3 peaks, which is obviously wrong, am I correct?
>
> Thanks a lot!
>
> On Thu, Feb 9, 2012 at 9:55 PM, R. Michael Weylandt
> wrote:
>>
r 5?
Michael
On Thu, Feb 16, 2012 at 10:18 AM, Luna wrote:
> But using my eye from the plot it should have more local peaks even with
> tolerance 5... no?
>
> I cannot change the number 5
>
> On Thu, Feb 16, 2012 at 9:17 AM, R. Michael Weylandt
> wrote:
>&g
I think you're over-thinking this: if you have adjusted prices, they
already incorporate splits+dividends --- so the return in adjusted
price *is* the total return. (Up to some fuzziness in how that
adjustment should be done)
Michael
On Sat, Feb 18, 2012 at 5:20 PM, SW wrote:
> Hello All,
>
>
>
This isn't really a finance question...
Your problem is that you use names() instead of just getting the row
numbers from outlierTest but then, when you convert the names to an
integer, your attempts to remove the row by that number and so doesn't
actually get the right row, then an outlier remain
t there quite a few?
iii) What does this mean: "the 3 big blocks/regimes need to be
contiguous within each block/regime itself..."
Perhaps you should look into reformulating your idea in terms of
regime switching models.
Michael Weylandt
On Tue, Feb 28, 2012 at 12:55 PM, Michael
You can write your own getQuote "method" and use the Defaults package to change
from src="yahoo" to "yahoopaid" or whatever you decide to call it. I don't have
a paid subscription so I can't really help with writing it.
Michael
On Mar 2, 2012, at 3:01 PM, Andre Zege wrote:
> Sorry, my bad. I
I think you are interpreting your frequency argument incorrectly and
thus R is giving you an error message because you are asking for
seasonal trends but giving only one observation per "season" -- it's
rather subtle: as the help page says,
frequency -- the number of observations per unit of time.
>> start.on.monday=TRUE)
> Error in as.numeric(test) :
> cannot coerce type 'S4' to vector of type 'double'
>
> I havent yet been able to identify the cause of this issue.
>
> It will also be good to know if the same extraction can be performed on the
I'm not sure I understand. PerformanceAnalytics is part of the
"xts-family" (as opposed to the Rmetrics family) of finance packages.
Your problem may be that it requires returns, but consider this
(somewhat opaque) one liner.
charts.PerformanceSummary(ROC(Cl(to.weekly(Ad(getSymbols("IBM",auto.assi
e data.frame conversion in
> between.
>
> Thank you in advance.
>
> Golam
>
> Sent from my iPhone
>
> On 24 Mar 2012, at 13:13, "R. Michael Weylandt"
> wrote:
>
>> I'm not sure I understand. PerformanceAnalytics is part of the
>> "
You might google around for anything involving the term quantstrat.
Michael
On Thu, Mar 29, 2012 at 6:03 AM, John Hardy wrote:
> Hi, Everyone
>
> I am getting going with R - mostly only for programming trading systems with
> RBloomberg data. I have figured out how to download daily and intraday
olute pass/fail (like most other statistical tests) so you
have to pick your desired significance threshold, but that's not hard.
As Brian said though, it makes most sense to do this inside your loop
construction.
Hope this helps,
Michael Weylandt
On Fri, Mar 30, 2012 at 12:00 PM, Brian G. P
procedure is, it seems that one needs to be
careful of unpenalized model selection here, but that might be already
taken care of in Johansen's method: I simply don't know. If you get
Pfaff's book I'm sure this is explained: I don't have a copy handy and
it's far too cold
Just a guess, but you are missing a comma in the GSPC part of the "y <- " line.
Your bigger problem is that you don't need sapply() in the first place
here (best I remember): I think this works:
y <- merge(Ad(FNMIX), Ad(DBLTX), Ad(AGG), Ad(GSPC))
ROC(y) # Note that you don't need sapply() because
As I said to you before, ca.jo is in the urca package.
It's defined in arguments: to wit,
K: The lag order of the series (levels) in the VAR.
It's also used in that manner throughout the details section.
Michael Weylandt
On Mon, Apr 2, 2012 at 5:33 PM, Michael wrote:
> Hi all,
I know IBrokers is reasonable and it has top-notch R integration
thanks to Jeff Ryan
(http://cran.r-project.org/web/packages/IBrokers/index.html) -- here's
some of their pricing data:
http://individuals.interactivebrokers.com/en/p.php?f=marketData
Michael
On Wed, Apr 4, 2012 at 3:52 AM, Sebastian
Following up on what Josh said, does it work with the built-in data
set MSFT provided by timeSeries?
library(xts)
library(timeSeries)
data(MSFT)
head(MSFT)
head(as.xts(MSFT))
head(as.zoo(MSFT))
all give the same dates for me.
Michael
On Tue, Apr 17, 2012 at 1:39 PM, Joshua Ulrich wrote:
> On
In addition to what's been said, you could much more easily write:
GSPC <- getSymbols("^GSPC", from = "2000-01-01", auto.assign = F)
which will give names that are (directly) compatible with other
quantmod functions and would seem to avoid your problem in the first
place.
Michael
On Mon, Apr 2
I'm no Jeff Ryan, but I'd be willing to put money on him wanting to
ask what you've tried and what error message(s) you got? (As well as
system config details you get from sessionInfo() )
I just installed it with no problems (well, none that I can check
since i don't have Berkeley DB) straight fro
This isn't really a finance question, but do you have the source code? If you
do, just use install.packages() with type = "source" and point to the
appropriate tarball.
The only somewhat difficult step might be if you need to add a namespace.
Michael
On Apr 25, 2012, at 10:00 PM, Wei-han Li
sion 2.14.1)
>
> I've installed Berkeley DB11gR2
>
> Thank you to tell me how to install RBerkeley if you have some information.
>
>
>
> -Message d'origine-
> De : R. Michael Weylandt [mailto:michael.weyla...@gmail.com]
> Envoyé : mercredi 25 avril
Your xts is a little out of date, I doubt that that's relevant but it
might not hurt to just go ahead and update.
But yes, sending the file (just a small sample thereof) would help --
the easiest way to do so would be to type
dput(head(x, 30)) # x is name of your object
and send us the textual o
Not runnable exactly as is where does the for loop have its close brace?
This works fine for me:
getSymbols("SPY", from = "1990-01-01")
y <- to.weekly(SPY)
chartSeries(y, type = "candlesticks", multi.col = TRUE, theme =
"white", subset = 'last 54 months', name = "SPY")
addMACD(fast = 12, slow
Use xts (always use xts!) and do something like this:
Suppose your data is in three columns called "dat"
xts(dat[,3], as.POSIXct(paste(dat[,1], dat[,2], format = "%m/%d/%Y %H%M"))
This will make an xts of your data taking the third column as the
data-bit and using the first two to make a POSIXct
;
> I am getting:
>
> Error in as.POSIXlt.character(x, tz, ...) :
> character string is not in a standard unambiguous format
>
> I am also using:
>
> Sys.setenv(TZ="GMT")
>
> just in case.
>
> Any thoughts?
>
> On 25 May 2012 19:49, R. Michael Weyl
Note also that you're probably using a somewhat outdated version -- as
of a few days ago the package is no longer called "RBloomberg" (see
John's posting here:
https://stat.ethz.ch/pipermail/r-sig-finance/2012q2/010275.html) --
you might try updating to Rbbg and seeing what happens.
[In addition t
H... looking at the source code, you see the return value is
wrapped in invisible() so it won't auto-print but it is returned
--it's not a bug, but I don't know exactly why Dirk would have done
that. There are other functions that return invisibly (e.g., most plot
functions) but those are usual
1 - 100 of 146 matches
Mail list logo