Re: [R-SIG-Finance] moving averages on specific interval and merge

[Jim Green]

Thanks Garret, it worked  correctly.

I found a similar question here at
http://r.789695.n4.nabble.com/Speed-optimization-on-minutes-distribution-calculation-td929537.html
and Jeff gave a very fast solution. Just for the reference.

Now I am attempting to write a custom indicator doing what you showed
to be used in quantstrat.

Jim.


On 12 May 2012 12:54, G See  wrote:
> Hi Jim,
>
> We prefer if you share with us what you tried and what you've done to
> attempt to solve the problem yourself.  Anyway here's one way to
> start.
>
> library(quantmod) # for the Vo function as well as xts and zoo
> load("~/Downloads/a.rda")
> CalcMinuteVolume <- function(x) {
>    # This function will calculate the average of the volume for a given minute
>    #of every day
>    mvolm <- aggregate(Vo(x), format(index(x), "%H:%M"), mean)
>    xts(coredata(mvolm), as.POSIXct(paste(as.Date(end(x)), index(mvolm)),
>        format="%Y-%m-%d %H:%M"))
> }
> dtemp <- to.daily(temp)
> index(dtemp) <- as.Date(index(dtemp)) # in case you have an old version of xts
> # make strings to use to subset the data by 10 day periods
> subsets <- rollapplyr(index(dtemp), 10, function(x)
>                        paste0(first(x), "/", last(x)))
> out <- lapply(subsets, function(ss) {
>    CalcMinuteVolume(temp[ss])
> })
> names(out) <- index(dtemp)[-c(1:9)] # days 1:9 were burned to calculate 1st 
> mean
> # Now you have a list of days. e.g.
> out[["2012-02-15"]]
> # Now merge
> temp$Average.Volume.at.current.interval <- do.call(rbind, out)
> temp
> 
>
> Things to consider:
> - You almost certainly want to throw a lag in there.
> - You have a different number of rows for different days.
>
> Also, see ?agrregate.zoo or ?tapply, ?rollapply, ?period.apply, and ?do.call
>
> HTH,
> Garrett
>
>
> On Thu, May 10, 2012 at 8:11 PM, Jim Green
>  wrote:
>> Greetings!
>>
>> I am using quantstrat and xts to do some intraday work and come up
>> with this problem. the xts object temp in the following example is
>> attached as and rda file.
>>
>>> head(temp)
>>                    A.Open A.High  A.Low A.Close A.Volume
>> 2012-02-01 08:29:00  42.47  43.76 41.410   43.76     2071
>> 2012-02-01 09:30:00  43.38  43.38 42.970   43.15    40300
>> 2012-02-01 09:31:00  43.14  43.28 43.130   43.28    14990
>> 2012-02-01 09:32:00  43.27  43.37 43.270   43.37     3300
>> 2012-02-01 09:33:00  43.37  43.50 43.370   43.48     3056
>> 2012-02-01 09:34:00  43.49  43.50 43.396   43.44    10968
>>
>>> tail(temp)
>>                     A.Open  A.High   A.Low A.Close A.Volume
>> 2012-03-27 16:07:00 45.6650 45.6650 45.6650 45.6650      170
>> 2012-03-27 16:08:00 45.6710 45.6710 45.6710 45.6710      474
>> 2012-03-27 16:10:00 45.9131 45.9131 45.9131 45.9131     1800
>> 2012-03-27 16:13:00 45.6952 45.6952 45.6952 45.6952      300
>> 2012-03-27 16:15:00 45.9368 45.9368 45.9368 45.9368      791
>> 2012-03-27 16:21:00 45.7000 45.7000 45.7000 45.7000    22000
>>
>>
>> I would like to calculate moving averages of minute volume for
>> specific interval and merge with the original minute ohlc data.
>>
>> take 09:40:00 for example, calculate the average previous 10 days
>> volume between 09:39:00 to 09:40:00 and merge with exiting data.
>>
>> ultimately I want to get an xts object with columns
>>
>> Open   High   Low   Close   Volume   Average.Volume.at.current.interval
>> 2012-03-27 16:07:00 45.6650 45.6650 45.6650 45.6650      170   177
>> 2012-03-27 16:08:00 45.6710 45.6710 45.6710 45.6710      474  500
>> ...
>> ...
>> ..
>> 2012-03-27 16:21:00 45.7000 45.7000 45.7000 45.7000    22000 1000
>>
>> any pointers are appreciated!
>>
>> Jim.
>>
>> ___
>> R-SIG-Finance@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.

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] IBrokers - reqOpenOrders and placeOrder not working anymore

[G See]

Checkout the code with svn and build it yourself.  The current version is 0.9-7

On linux or mac, you'd do something like this

svn checkout http://ibrokers.googlecode.com/svn/trunk/ ibrokers

That will checkout the code into a directory called ibrokers.  Then you can

R --vanilla CMD INSTALL --build ibrokers

HTH,
Garrett

On Thu, May 17, 2012 at 3:15 PM, omerle  wrote:
> Hi,
>
> I just dowload IBrokers from 
> http://code.google.com/p/ibrokers/downloads/detail?
> name=IBrokers_0.9-0.tar.gz&can=2&q=
> and install IBrokers with
> install.packages("C:/IBrokers_0.9-0).tar.gz", repos = NULL, type="source")
> but reqOpenOrders and placeOrders is not working anyway !
>
> Any other advice ?
>
> Thanks
>
>
> Try updating to the newest googlecode version. A change in IB response.
>
> Jeff
>
> Jeffrey Ryan | Founder | jeffrey.r...@lemnica.com
>
> www.lemnica.com
>
> On May 11, 2012, at 5:26 PM, G See  wrote:
>
>> FWIW, I have the opposite problem. When connected with a live account, I see:
>>
>>> tws <- twsConnect()
>>> tws$nextValidId
>> [1] "1"
>>
>> But, with the demo account, the account number is being stored in 
>> nextValidId:
>>
>>> tws <- twsConnect()
>>> tws$nextValidId
>> [1] "DU123456"
>>
>> # But, I changed the actual account number
>>
>> Best,
>> Garrett
>>
>> On Fri, May 11, 2012 at 8:03 AM, omerle  wrote:
>>> Dear All,
>>>
>>> First thanks to Jeff Ryan for the very interesting IBrokers package.
>>>
>>> After looking for answers on Rseek without success, your are my last hope !
>>>
>>> Here is my problem :
>>>
>>> placeOrder & reqOpenOrders don't work on my trading accout but its working 
>>> on my demo
>>> account !
>>>
>>> For instance when I do
>>>
>>> reqOpenOrders(twsconn) or IBrokers:::.reqOpenOrders(twsconn)
>>> => I get only the error message 53 (OPEN_ORDER_END) even if I have open 
>>> order in my
>>> console.
>>>
>>> OR
>>>
>>> placeOrder
>>> => it doesnt do anything
>>>
>>> => But I can get TWS real time quotes or accountUpdate
>>>
>>> (I think it was ok before at least for placeOrder)
>>>
>>> I set the same parameter in configurationAPIs for both my demo account and 
>>> trading
> account
>>> but trading account is still not working.
>>>
>>> Do you have any idea what I did wrong and how I can use both placeOrder and
> reqOpenOrders
>>> ?
>>>
>>>
>>>
>>> Many Thanks,
>>>
>>> Olivier Merle
>>>
>>> Une messagerie gratuite, garantie à vie et des services en plus, ça vous 
>>> tente ?
>>> Je crée ma boîte mail www.laposte.net
>>>
>>> ___
>>> R-SIG-Finance@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.
>> ___
>> R-SIG-Finance@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.
>
>
> Une messagerie gratuite, garantie à vie et des services en plus, ça vous 
> tente ?
> Je crée ma boîte mail www.laposte.net
>
> ___
> R-SIG-Finance@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.

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] IBrokers - reqOpenOrders and placeOrder not working anymore

[omerle]

Hi,

I just dowload IBrokers from http://code.google.com/p/ibrokers/downloads/detail?
name=IBrokers_0.9-0.tar.gz&can=2&q=
and install IBrokers with
install.packages("C:/IBrokers_0.9-0).tar.gz", repos = NULL, type="source")
but reqOpenOrders and placeOrders is not working anyway !

Any other advice ?

Thanks


Try updating to the newest googlecode version. A change in IB response. 

Jeff

Jeffrey Ryan | Founder | jeffrey.r...@lemnica.com

www.lemnica.com

On May 11, 2012, at 5:26 PM, G See  wrote:

> FWIW, I have the opposite problem. When connected with a live account, I see:
>
>> tws <- twsConnect()
>> tws$nextValidId
> [1] "1"
>
> But, with the demo account, the account number is being stored in nextValidId:
>
>> tws <- twsConnect()
>> tws$nextValidId
> [1] "DU123456"
>
> # But, I changed the actual account number
>
> Best,
> Garrett
>
> On Fri, May 11, 2012 at 8:03 AM, omerle  wrote:
>> Dear All,
>>
>> First thanks to Jeff Ryan for the very interesting IBrokers package.
>>
>> After looking for answers on Rseek without success, your are my last hope !
>>
>> Here is my problem :
>>
>> placeOrder & reqOpenOrders don't work on my trading accout but its working 
>> on my demo
>> account !
>>
>> For instance when I do
>>
>> reqOpenOrders(twsconn) or IBrokers:::.reqOpenOrders(twsconn)
>> => I get only the error message 53 (OPEN_ORDER_END) even if I have open 
>> order in my
>> console.
>>
>> OR
>>
>> placeOrder
>> => it doesnt do anything
>>
>> => But I can get TWS real time quotes or accountUpdate
>>
>> (I think it was ok before at least for placeOrder)
>>
>> I set the same parameter in configurationAPIs for both my demo account and 
>> trading 
account
>> but trading account is still not working.
>>
>> Do you have any idea what I did wrong and how I can use both placeOrder and 
reqOpenOrders
>> ?
>>
>>
>>
>> Many Thanks,
>>
>> Olivier Merle
>>
>> Une messagerie gratuite, garantie à vie et des services en plus, ça vous 
>> tente ?
>> Je crée ma boîte mail www.laposte.net
>>
>> ___
>> R-SIG-Finance@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.
> ___
> R-SIG-Finance@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.


Une messagerie gratuite, garantie à vie et des services en plus, ça vous tente ?
Je crée ma boîte mail www.laposte.net

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] look at the underlying source code

[Brian G. Peterson]

On Thu, 2012-05-17 at 19:40 +0200, jaimie villanueva wrote:
> hi
> 
> someone can show me how can i get the source code of a function. Exactly,
> Function "ugarchsim" from library (rugarch).
> I need to know (in detailed ) how the variance and  mean ecuation of a
> arma/garch process are calculated.
> With other packages like "fGarch" i used to invoked the function debug ()
> and allows me to step into the functions but with ugarchsim i can't use it.
> 
> any suggestions are well received

1> this is not a finance question.  Please direct general R questions to
r-help, and/or follow the advice that Alexios already gave you to look
up the answer.

2> all the code is on R-Forge, download it and look.

3> much of the rugarch and rmgarch code is compiled, and other parts of
it are S4, these are not as easy to debug into as vanilla R code.  See
the answer to 1> above for details on how to do this, and pursue further
non-finance questions about it on r-help, not here.

Regards,

   - Brian

-- 
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] look at the underlying source code

[julien cuisinier]

simply type the name of the function in the console without parenthesis or 
anything - much easier than debug

HTH,
Julien


> Date: Thu, 17 May 2012 19:40:00 +0200
> From: jaimie.villanu...@gmail.com
> To: R-SIG-Finance@r-project.org
> Subject: [R-SIG-Finance] look at the underlying source code
> 
> hi
> 
> someone can show me how can i get the source code of a function. Exactly,
> Function "ugarchsim" from library (rugarch).
> I need to know (in detailed ) how the variance and  mean ecuation of a
> arma/garch process are calculated.
> With other packages like "fGarch" i used to invoked the function debug ()
> and allows me to step into the functions but with ugarchsim i can't use it.
> 
> any suggestions are well received
> 
> thanks
> 
> Jamie
> 
>   [[alternative HTML version deleted]]
> 
> ___
> R-SIG-Finance@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.
  
[[alternative HTML version deleted]]

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] look at the underlying source code

[Joshua Ulrich]

On Thu, May 17, 2012 at 12:40 PM, jaimie villanueva
 wrote:
> hi
>
> someone can show me how can i get the source code of a function. Exactly,
> Function "ugarchsim" from library (rugarch).
> I need to know (in detailed ) how the variance and  mean ecuation of a
> arma/garch process are calculated.
> With other packages like "fGarch" i used to invoked the function debug ()
> and allows me to step into the functions but with ugarchsim i can't use it.
>
> any suggestions are well received
>
I reiterate the suggestion Alexios already gave you, namely that you
use some of the, "many resources online to learn about R, S3, S4
etc.".  Also, "how do I look at source code?" is not a finance-related
question, even though you want to look at the source code of a
finance-related function.

> thanks
>
> Jamie
>

Best,
--
Joshua Ulrich  |  FOSS Trading: www.fosstrading.com

___
R-SIG-Finance@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.

[R-SIG-Finance] look at the underlying source code

[jaimie villanueva]

hi

someone can show me how can i get the source code of a function. Exactly,
Function "ugarchsim" from library (rugarch).
I need to know (in detailed ) how the variance and  mean ecuation of a
arma/garch process are calculated.
With other packages like "fGarch" i used to invoked the function debug ()
and allows me to step into the functions but with ugarchsim i can't use it.

any suggestions are well received

thanks

Jamie

[[alternative HTML version deleted]]

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] IBrokers: issues with combo contracts

[Jeffrey Ryan]

I don't think I've tried this, so I may be wrong, but twsBAG isn't a real
contract, it is for the order itself.  Therefore you can't get mktData on
it.

Additionally, snapshot= isn't in the documentation for reqMktData.

Read the docs,and then see if you can get further along.  It (API and
IBrokers) is intentionally tricky to prevent wide-spread self-inflicted
harm ;-)  And 'paper account' is a good step...

Best,
Jeff

On 5/17/12 12:03 PM, "Alain Burt"  wrote:

>Hi there,
>
>I am having some troubles with the use of the IBrokers package on my IB
>paper account. I am trying to get market data and an order execution for a
>combo contract, with no success in both cases. I wonder if anyone could
>help. Here is the code I tried
>
>
>
>  ibg <- ibgConnect()
>
>
>
>leg1 <- twsComboLeg( conId = "79265697",
>
>   ratio = 1,
>
>action = "BUY",
>
>   exchange = "SMART")
>
>
>
>leg2 <- twsComboLeg( conId = "79265687",
>
>   ratio = 1,
>
>action = "SELL",
>
>   exchange = "SMART")
>
>
>
> reqMktData(conn = ibg, Contract = twsBAG(leg1,leg2), tickGenerics = "",
>tickerId = "1", snapshot = T)
>
>
>
> placeOrder(ibg, twsBAG(leg1,leg2), twsOrder(reqIds(ibg),"BUY",1,"MKT"))
>
>
>
>About the market data, I am not getting anything, even by setting snapshot
>= F. As for the order, it is placed but never executed. Could it be a
>problem with contract objects created by twsBAG ? Does anyone know how to
>fix these issues ?
>
>
>
>Best regards,
>
>Alain
>
>   [[alternative HTML version deleted]]
>
>___
>R-SIG-Finance@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.

___
R-SIG-Finance@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.

[R-SIG-Finance] IBrokers: issues with combo contracts

[Alain Burt]

Hi there,

I am having some troubles with the use of the IBrokers package on my IB
paper account. I am trying to get market data and an order execution for a
combo contract, with no success in both cases. I wonder if anyone could
help. Here is the code I tried



  ibg <- ibgConnect()



leg1 <- twsComboLeg( conId = "79265697",

   ratio = 1,

action = "BUY",

   exchange = "SMART")



leg2 <- twsComboLeg( conId = "79265687",

   ratio = 1,

action = "SELL",

   exchange = "SMART")



 reqMktData(conn = ibg, Contract = twsBAG(leg1,leg2), tickGenerics = "",
tickerId = "1", snapshot = T)



 placeOrder(ibg, twsBAG(leg1,leg2), twsOrder(reqIds(ibg),"BUY",1,"MKT"))



About the market data, I am not getting anything, even by setting snapshot
= F. As for the order, it is placed but never executed. Could it be a
problem with contract objects created by twsBAG ? Does anyone know how to
fix these issues ?



Best regards,

Alain

[[alternative HTML version deleted]]

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] mark random rows of xts object

[Brian G. Peterson]

On Thu, 2012-05-17 at 14:01 +0200, Andreas Voellenklee wrote:
> To compare results of a trading system against a
> random-entry-benchmark, I'm trying to program a function that marks
> random 5% of bars of a xts object.
> 
> The result should be a xts object with the same indexes as the OHLCV
> object and one column "random_entry" that contains boolean values. 

getSymbols('MMM')
MMM$random<-0
MMM$random[sample(1:nrow(MMM),round(nrow(MMM)/20),replace=FALSE),]<-1

Regards,

   - Brian

-- 
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock

___
R-SIG-Finance@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.

[R-SIG-Finance] mark random rows of xts object

[Andreas Voellenklee]

Hi,

To compare results of a trading system against a
random-entry-benchmark, I'm trying to program a function that marks
random 5% of bars of a xts object.

The result should be a xts object with the same indexes as the OHLCV
object and one column "random_entry" that contains boolean values.

I tried this without success:


# returns true in 5% of calls
random <- function() {runif(1,0,1) >= .95}

getSymbols("MMM")

> as.xts(apply(MMM, 1, random))
Fehler in FUN(newX[, i], ...) : unbenutzte(s) Argument(e) (newX[, i])


Thanks for your help,
Andreas

___
R-SIG-Finance@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.

Re: [R-SIG-Finance] About rugarch (message to alexios)

[alexios ghalanos]

Jaimie,
1. This is a mailing list so do not start a message by addressing it to 
one individual.
2. Following from 1, your message has no context for anyone but myself 
which defeats the purpose of posting to a mailing list.


Your original question was:

> My question is in line with function "ugarchsim". With the package
> "fGarch" to be sure about what the simulation function is doing, i
> used to use a function call "debug" wich allows you to make a step by 
> step analysis. To me it's usefull to understand, for example, how de 
> variance of a garch process are calculated, or the innovations, how

> they have been calculated. The main purpose is to ensure that
> functions are doing what i really want. The debug function allows you 
> to see the formulas and you can type any variable and returns it's

> value.
> But when i try to do the same with the function "ugarchsim",  i
> can't. I can't get into the formula to see the process of
> calculating."

I mentioned that the package uses S4 methods and classes which add a 
layer of abstraction on top of the code. If  you do not know what these 
are, there are many resources online to learn about R, S3, S4 etc. The 
same goes about how to source the code or see under the hood (it is OPEN 
SOURCE).


Regards,
Alexios


On 17/05/2012 09:36, jaimie villanueva wrote:

First, I apologize because i`m not an expert using R. I'm not sure to
understand you when you say "The rugarch provide an added layer of
abstraction on top of the code".
you mean that i can access to that added layer?

Second is : How can i look at the source code?

regards

Jaimie

___
R-SIG-Finance@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.



___
R-SIG-Finance@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.

[R-SIG-Finance] About rugarch (message to alexios)

[jaimie villanueva]

First, I apologize because i`m not an expert using R. I'm not sure to
understand you when you say "The rugarch provide an added layer of
abstraction on top of the code".
you mean that i can access to that added layer?

Second is : How can i look at the source code?

regards

Jaimie

___
R-SIG-Finance@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.