insert element in IBrokers/XTS time series?

still not working...
Sorry, try this

?rbind(myhist, xts(data.frame(NA,NA,NA,NA,NA,NA),
order.by=as.Date("2009-08-01")))

also, check class(time(myhist)) and if it is not "Date" you may need
to replace as.Date with the appropriate constructor for your time
class.

Sean.

On Sun, Jul 5, 2009 at 2:47 AM, Michael<comtech.usa at gmail.com> wrote:
I got an error:

Error in xts(data.frame(rep(NA, 6), order.by = as.Date("2009-08-01"))) :
?order.by requires an appropriate time-based object

On Sat, Jul 4, 2009 at 2:26 AM, Sean Carmody<seancarmody at gmail.com> wrote:
Eliminating a row is straightforward, e.g.

myhist[time(myhist) != as.Date("2009-06-29"),]

will return everything except the entry for 2009-06-29.

Adding a row is, I believe, a bit more fiddly. You can use

rbind(myhist, xts(data.frame(rep(NA,6), order.by=as.Date("2009-08-01")))

although this will give a warning message about column names
differing. If you have an additional xts with the same column names,
you can simply use merge()

Sean.

On Sat, Jul 4, 2009 at 10:01 AM, Michael<comtech.usa at gmail.com> wrote:
Hi all,

Let's say I have the following timeseres in IBrokers and XTS:

2009-06-25 ? ? ?18038 ? ? ?18350 ? ? 18011 ? ? ? 18327 ? ? ? ? 5396 ? ? 18183
2009-06-26 ? ? ?18420 ? ? ?18708 ? ? 18331 ? ? ? 18515 ? ? ? ? 7367 ? ? 18499
2009-06-29 ? ? ?18546 ? ? ?18688 ? ? 18423 ? ? ? 18577 ? ? ? ?28728 ? ? 18555
2009-07-03 ? ? ?17951 ? ? ?18198 ? ? 17866 ? ? ? 18149 ? ? ? ?38019 ? ? 18059

I want to add a few more rows, indexed by

"2009-06-30", "2009-07-01", "2009-07-02", and using NA's,

I did the following:

myhist["2009-07-01",]=c(NA, NA, NA, NA, NA, NA, NA, NA)

But it didn't really work.

What's the good way to insert missing points in time series?

Thanks!

_______________________________________________
R-SIG-Finance at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only.
-- If you want to post, subscribe first.

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

It works for me (albeit with the warning about column names I
mentioned)! What do you get from

class(time(myhist))?

Sean.
still not working...

On Sat, Jul 4, 2009 at 2:10 PM, Sean Carmody<seancarmody at gmail.com> wrote:
Sorry, try this

?rbind(myhist, xts(data.frame(NA,NA,NA,NA,NA,NA),
order.by=as.Date("2009-08-01")))

also, check class(time(myhist)) and if it is not "Date" you may need
to replace as.Date with the appropriate constructor for your time
class.

Sean.

On Sun, Jul 5, 2009 at 2:47 AM, Michael<comtech.usa at gmail.com> wrote:
I got an error:

Error in xts(data.frame(rep(NA, 6), order.by = as.Date("2009-08-01"))) :
?order.by requires an appropriate time-based object

On Sat, Jul 4, 2009 at 2:26 AM, Sean Carmody<seancarmody at gmail.com> wrote:
Eliminating a row is straightforward, e.g.

myhist[time(myhist) != as.Date("2009-06-29"),]

will return everything except the entry for 2009-06-29.

Adding a row is, I believe, a bit more fiddly. You can use

rbind(myhist, xts(data.frame(rep(NA,6), order.by=as.Date("2009-08-01")))

although this will give a warning message about column names
differing. If you have an additional xts with the same column names,
you can simply use merge()

Sean.

On Sat, Jul 4, 2009 at 10:01 AM, Michael<comtech.usa at gmail.com> wrote:
Hi all,

Let's say I have the following timeseres in IBrokers and XTS:

2009-06-25 ? ? ?18038 ? ? ?18350 ? ? 18011 ? ? ? 18327 ? ? ? ? 5396 ? ? 18183
2009-06-26 ? ? ?18420 ? ? ?18708 ? ? 18331 ? ? ? 18515 ? ? ? ? 7367 ? ? 18499
2009-06-29 ? ? ?18546 ? ? ?18688 ? ? 18423 ? ? ? 18577 ? ? ? ?28728 ? ? 18555
2009-07-03 ? ? ?17951 ? ? ?18198 ? ? 17866 ? ? ? 18149 ? ? ? ?38019 ? ? 18059

I want to add a few more rows, indexed by

"2009-06-30", "2009-07-01", "2009-07-02", and using NA's,

I did the following:

myhist["2009-07-01",]=c(NA, NA, NA, NA, NA, NA, NA, NA)

But it didn't really work.

What's the good way to insert missing points in time series?

Thanks!

_______________________________________________
R-SIG-Finance at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only.
-- If you want to post, subscribe first.

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody
Here's a more elegant approach that should not generate the error message:

xts(rbind(coredata(myhist), rep(NA,dim(myhist)[2])),
order.by=c(index(myhist), as.Date("2009-01-01")))

Note that if class(index(myhist)) is not type "Date" you will need to
replace as.Date with the appropriate constructor, e.g. as.POSIXct or
as.chron.

Regards,
Sean.
still not working...

On Sat, Jul 4, 2009 at 2:10 PM, Sean Carmody<seancarmody at gmail.com> wrote:
Sorry, try this

?rbind(myhist, xts(data.frame(NA,NA,NA,NA,NA,NA),
order.by=as.Date("2009-08-01")))

also, check class(time(myhist)) and if it is not "Date" you may need
to replace as.Date with the appropriate constructor for your time
class.

Sean.

On Sun, Jul 5, 2009 at 2:47 AM, Michael<comtech.usa at gmail.com> wrote:
I got an error:

Error in xts(data.frame(rep(NA, 6), order.by = as.Date("2009-08-01"))) :
?order.by requires an appropriate time-based object

On Sat, Jul 4, 2009 at 2:26 AM, Sean Carmody<seancarmody at gmail.com> wrote:
Eliminating a row is straightforward, e.g.

myhist[time(myhist) != as.Date("2009-06-29"),]

will return everything except the entry for 2009-06-29.

Adding a row is, I believe, a bit more fiddly. You can use

rbind(myhist, xts(data.frame(rep(NA,6), order.by=as.Date("2009-08-01")))

although this will give a warning message about column names
differing. If you have an additional xts with the same column names,
you can simply use merge()

Sean.

On Sat, Jul 4, 2009 at 10:01 AM, Michael<comtech.usa at gmail.com> wrote:
Hi all,

Let's say I have the following timeseres in IBrokers and XTS:

2009-06-25 ? ? ?18038 ? ? ?18350 ? ? 18011 ? ? ? 18327 ? ? ? ? 5396 ? ? 18183
2009-06-26 ? ? ?18420 ? ? ?18708 ? ? 18331 ? ? ? 18515 ? ? ? ? 7367 ? ? 18499
2009-06-29 ? ? ?18546 ? ? ?18688 ? ? 18423 ? ? ? 18577 ? ? ? ?28728 ? ? 18555
2009-07-03 ? ? ?17951 ? ? ?18198 ? ? 17866 ? ? ? 18149 ? ? ? ?38019 ? ? 18059

I want to add a few more rows, indexed by

"2009-06-30", "2009-07-01", "2009-07-02", and using NA's,

I did the following:

myhist["2009-07-01",]=c(NA, NA, NA, NA, NA, NA, NA, NA)

But it didn't really work.

What's the good way to insert missing points in time series?

Thanks!

_______________________________________________
R-SIG-Finance at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only.
-- If you want to post, subscribe first.

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

--
Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody

Sean Carmody

The Stubborn Mule
http://www.stubbornmule.net
http://twitter.com/seancarmody