Skip to content

Two similar zoo objects with different structures, how to get same structure?

6 messages · Sergey Goriatchev, Gabor Grothendieck

Original
Sergey Goriatchev
# 
Dear all,

I have a zoo object that has following structure:
zoo [1:5219, 1:12] 91.9 91.8 91.7 91.8 91.7 ...
 - attr(*, "index")=Classes 'dates', 'times'  atomic [1:5219] 7305
7306 7307 7308 7309 ...
  .. ..- attr(*, "format")= chr "m/d/y"
  .. ..- attr(*, "origin")= Named num [1:3] 1 1 1970
  .. .. ..- attr(*, "names")= chr [1:3] "month" "day" "year"
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:12] "ED4 COMDTY" "ED12 COMDTY" "ER4 COMDTY" "ER12 COMDTY" ...

I also have a vector of dates of same length as zoo object (5219):
chr [1:5219] "01/01/90" "01/02/90" "01/03/90" "01/04/90" "01/05/90"
"01/08/90" "01/09/90" "01/10/90" "01/11/90" "01/12/90" ...

I do the following:
and get the following:
'zoo' series from 1990-01-01 to 2009-12-31
  Data: num [1:5219, 1:12] 91.9 91.8 91.7 91.8 91.7 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:12] "ED4 COMDTY" "ED12 COMDTY" "ER4 COMDTY" "ER12 COMDTY" ...
  Index: Class 'Date'  num [1:5219] 7305 7306 7307 7308 7309 ...

But what I really want to get is the following:
'zoo' series from 01/01/90 to 12/31/09
  Data: num [1:5219, 1:12] 91.9 91.8 91.7 91.8 91.7 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:12] "ED4 COMDTY" "ED12 COMDTY" "ER4 COMDTY" "ER12 COMDTY" ...
  Index: Classes 'dates', 'times'  atomic [1:5219] 7305 7306 7307 7308 7309 ...
  ..- attr(*, "format")= chr "m/d/y"
  ..- attr(*, "origin")= Named num [1:3] 1 1 1970
  .. ..- attr(*, "names")= chr [1:3] "month" "day" "year"

The array bldata is the same, only the last one is on a machine
running RBloomberg, and then I tried to save bldata (and rownames of
bldata in bldates) and
transfer bldata to the other machine which is not a Bloomberg
terminal, and the structure of bldata changes.

What should I do in this case to adjust the structure? I guess I have
to somehow change the class of Index and set attributes of "Index" and
attribute "names" of attribute "origin" of attribute "Index"?

Thank you for your help in advance!

Regards,
Sergey
Gabor Grothendieck
# 
Please reduce your examples down to small amounts of data and use dput so
that they are reproducible.
On Wed, Jan 21, 2009 at 9:32 AM, Sergey Goriatchev <sergeyg at gmail.com> wrote:
Sergey Goriatchev
# 
Dear Dr. Grothendieck,

First of all, I realized I did not load zoo package before I tried the
first str(bldata). If I load zoo and then do str(bldata) I get the
following:

'zoo' series from 7305 to 14609
  Data: num [1:5219, 1:12] 91.9 91.8 91.7 91.8 91.7 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:12] "ED4 COMDTY" "ED12 COMDTY" "ER4 COMDTY" "ER12 COMDTY" ...
  Index: Classes 'dates', 'times'  atomic [1:5219] 7305 7306 7307 7308 7309 ...
  ..- attr(*, "format")= chr "m/d/y"
  ..- attr(*, "origin")= Named num [1:3] 1 1 1970
  .. ..- attr(*, "names")= chr [1:3] "month" "day" "year"

Now, I've done what you said and here is the ASCII representation of
the data (I don't know if one can attach files here, and anyway I
cannot attach files where I am, but the following would reproduce
exactly):

a <- (structure(c(98.585, 98.355, 98.48, 98.585, 98.67, 98.695, 98.81,
98.865, 98.865, 98.865, 98.735, 98.805, 98.805, 97.435, 97.18,
97.165, 97.265, 97.34, 97.415, 97.445, 97.505, 97.525, 97.635,
97.625, 97.53, 97.53, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 0.25,
0.25, 0.25, 0.25, 0.25, 0.25, 0.25, 2, 2, 2, 2, 2, 1.5, 1.5,
1.5, 1.5, 1.5, 1.5, 1.5, 1.5), .Dim = c(13L, 4L), index = structure(c(14245,
14246, 14249, 14250, 14251, 14252, 14253, 14256, 14257, 14258,
14259, 14260, 14263), format = "m/d/y", origin = structure(c(1,
1, 1970), .Names = c("month", "day", "year")), class = c("dates",
"times")), class = "zoo", .Dimnames = list(NULL, c("ED4 COMDTY",
"ED12 COMDTY", "FDTR INDEX", "UKBRBASE INDEX"))))

Copy-paste to R prompt shows following:

       ED4 COMDTY ED12 COMDTY FDTR INDEX UKBRBASE INDEX
14245     98.585      97.435       0.25            2.0
14246     98.355      97.180       0.25            2.0
14249     98.480      97.165       0.25            2.0
14250     98.585      97.265       0.25            2.0
14251     98.670      97.340       0.25            2.0
14252     98.695      97.415       0.25            1.5
14253     98.810      97.445       0.25            1.5
14256     98.865      97.505       0.25            1.5
14257     98.865      97.525       0.25            1.5
14258     98.865      97.635       0.25            1.5
14259     98.735      97.625       0.25            1.5
14260     98.805      97.530       0.25            1.5
14263     98.805      97.530       0.25            1.5

On my Bloomberg machine, in R console, the rownames look like:
           ED4 COMDTY ED12 COMDTY FDTR INDEX UKBRBASE INDEX
01/01/09     98.585      97.435       0.25            2.0
01/02/09     98.355      97.180       0.25            2.0
01/05/09     98.480      97.165       0.25            2.0
01/06/09     98.585      97.265       0.25            2.0
01/07/09     98.670      97.340       0.25            2.0

and that is what I try to get as well.

Doing:

attributes(a)[[2]] <- format(as.Date(attributes(a)[[2]]), "%m/%d/%y")

changes structure of a to:

'zoo' series from 01/01/09 to 01/19/09
  Data: num [1:13, 1:4] 98.6 98.4 98.5 98.6 98.7 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:4] "ED4 COMDTY" "ED12 COMDTY" "FDTR INDEX" "UKBRBASE INDEX"
  Index:  chr [1:13] "01/01/09" "01/02/09" "01/05/09" "01/06/09"
"01/07/09" "01/08/09" "01/09/09" "01/12/09" "01/13/09" "01/14/09"
"01/15/09" "01/16/09" "01/19/09"

That is not the same structure, attributes disappeared.

What am I doing wrong?
Thank you in advance for your help.

Regards,
Sergey




On Wed, Jan 21, 2009 at 15:56, Gabor Grothendieck
<ggrothendieck at gmail.com> wrote:

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Gabor Grothendieck
    

    

      
      #
    

  


  

      
Is your purpose to change the times in some way?
If z is a zoo series you can change the times like this:

library(zoo)
library(chron)

time(z) <- ...whatever...

e.g.

            

      
      
      
    

        
11 12 13
 1  2  3

            

      
      
      
    

        
1970-01-12 1970-01-13 1970-01-14
         1          2          3

        
On Wed, Jan 21, 2009 at 10:51 AM, Sergey Goriatchev <sergeyg at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Sergey Goriatchev
    

    

      
      #
    

  


  

      
Dear Dr. Grothendieck,

My purpose in this case it to have the structure of the object on
non-Bloomberg machine the same as that of the object on the Bloomberg
machine.
That way I am sure that whatever code I write away from Bloomberg
machine will work on it when I copy the code to it.
Also, I am very curious how I could change the structure of an object,
in this case a zoo object. Why did the attributes disappear? How can I
restore them?

Kind Regards,
Sergey

On Wed, Jan 21, 2009 at 17:01, Gabor Grothendieck

        
<ggrothendieck at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Gabor Grothendieck
    

    

      
      #
    

  


  

      
So far we have dput(a).  Does that represent what you have?
what you want?  Can you provide both and a description in words.
See the last line to every message to r-help.

        
On Wed, Jan 21, 2009 at 11:13 AM, Sergey Goriatchev <sergeyg at gmail.com> wrote: