Skip to content

can't merge zoo ojects and convert to ts (been trying for 2 days)

2 messages · Darren Norris, Gabor Grothendieck

Darren Norris
# 
I'm stuck, but am sure it can be done I just don't understand how.
I have data in an irregular timeseries. I want to be able to use stl to
visualise the data (see seasonal parts etc), so I need to change to regular
series of class ts (I think).

I am using 2 zoo objects one is regular and the other is my irregular data.
I am then merging to create the object I want but when I try to change to a
ts class using as.ts I get the following error:
Error in zoo(coredata(x), tt) : 
  ?x? : attempt to define illegal zoo object.

Can anyone show the error of my ways? I have looked throuh help and the zoo
pdf's but can't see the answer.
Many thanks for any help.
All code and data below (short example).
Darren

R version 2.6.1 (2007-11-26) 
i386-pc-mingw32 

locale:
LC_COLLATE=English_United Kingdom.1252;LC_CTYPE=English_United
Kingdom.1252;LC_MONETARY=English_United
Kingdom.1252;LC_NUMERIC=C;LC_TIME=English_United Kingdom.1252

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] chron_2.3-21 zoo_1.4-2   

loaded via a namespace (and not attached):
[1] grid_2.6.1     lattice_0.17-2 tools_2.6.1   

###Import data:
 ttest<-read.csv("la_time_test2.csv", header = T, as.is=TRUE);
 tlist<-read.csv("la_time_list2.csv", header = T, as.is=TRUE)

####Convert to zoo:
 zooreg_time<-zoo(, as.chron(strptime(tlist$a_datetime, format="%d/%m/%Y
%H"), format = c(dates = "Y-m-d ", times = "H:M:S")),frequency=8760 )
 class(zooreg_time)
[1] "zooreg" "zoo"   

###Create my zoo objects
###First irregular series. Steps: 1) create vector to order by 2) create zoo
object.

#####Vector to order by:
%H"), format = c(dates = "Y-m-d ", times = "H:M:S"))
zoo_test
[1] (2004-Apr-28 08:00:00) (2004-Apr-28 13:00:00)

#####Zoo object:
zoo_test2<-zoo(ttest$Minutos.de.viaje,zoo_test)
 zoo_test2
(2004-Apr-28 08:00:00) (2004-Apr-28 13:00:00) 
                   125                    295

#####Merge (need to approximate NAs so can change to ts?)
x<-na.approx(merge(zooreg_time,zoo_test2))
x
(2004-Apr-28 08:00:00) (2004-Apr-28 09:00:00) (2004-Apr-28 10:00:00)
(2004-Apr-28 11:00:00) (2004-Apr-28 12:00:00) (2004-Apr-28 13:00:00) 
                   125                    159                    193                   
227                    261                    295

######Now change to ts
as.ts(x)
Error in zoo(coredata(x), tt) : 
  ?x? : attempt to define illegal zoo object

THE DATA
la_time_test2.csv
TRIP_ID	Ruta_datetime_start	Minutos de viaje
28/04/2004_CM2-CM1_downriver_08:15	28/04/2004 08:15	125
28/04/2004_CM1-CM2_upriver_13:05	28/04/2004 13:05	295

la_time_list2.csv
a_datetime
28/04/2004 01:00
28/04/2004 02:00
28/04/2004 03:00
28/04/2004 04:00
28/04/2004 05:00
28/04/2004 06:00
28/04/2004 07:00
28/04/2004 08:00
28/04/2004 09:00
28/04/2004 10:00
28/04/2004 11:00
28/04/2004 12:00
28/04/2004 13:00
28/04/2004 14:00
28/04/2004 15:00
28/04/2004 16:00
28/04/2004 17:00

ttest
                             TRIP_ID Ruta_datetime_start Minutos.de.viaje
1 28/04/2004_CM2-CM1_downriver_08:15    28/04/2004 08:15              125
2   28/04/2004_CM1-CM2_upriver_13:05    28/04/2004 13:05              295

tlist
         a_datetime
1  28/04/2004 01:00
2  28/04/2004 02:00
3  28/04/2004 03:00
4  28/04/2004 04:00
5  28/04/2004 05:00
6  28/04/2004 06:00
7  28/04/2004 07:00
8  28/04/2004 08:00
9  28/04/2004 09:00
10 28/04/2004 10:00
11 28/04/2004 11:00
12 28/04/2004 12:00
13 28/04/2004 13:00
14 28/04/2004 14:00
15 28/04/2004 15:00
16 28/04/2004 16:00
17 28/04/2004 17:00

zooreg_time
Data:
numeric(0)
Index:
 [1] (2004-Apr-28 01:00:00) (2004-Apr-28 02:00:00) (2004-Apr-28 03:00:00)
(2004-Apr-28 04:00:00) (2004-Apr-28 05:00:00) (2004-Apr-28 06:00:00)
 [7] (2004-Apr-28 07:00:00) (2004-Apr-28 08:00:00) (2004-Apr-28 09:00:00)
(2004-Apr-28 10:00:00) (2004-Apr-28 11:00:00) (2004-Apr-28 12:00:00)
[13] (2004-Apr-28 13:00:00) (2004-Apr-28 14:00:00) (2004-Apr-28 15:00:00)
(2004-Apr-28 16:00:00) (2004-Apr-28 17:00:00)


x
(2004-Apr-28 08:00:00) (2004-Apr-28 09:00:00) (2004-Apr-28 10:00:00)
(2004-Apr-28 11:00:00) (2004-Apr-28 12:00:00) (2004-Apr-28 13:00:00) 
                   125                    159                    193                   
227                    261                    295
Gabor Grothendieck
# 
Its a bug in na.approx.  I just fixed it and (1) until a
new version of zoo comes out add this to your code:

source("http://r-forge.r-project.org/plugins/scmsvn/viewcvs.php/*checkout*/pkg/R/na.approx.R?rev=448&root=zoo")

or (2) a workaround not requiring that source statement
is to add the na.rm = FALSE argument to
na.approx since the bug only gets triggered with the
default na.rm = TRUE setting.

Actually in your case you don't need na.approx anyways
since as.ts can do it all itself and if you want to force
a specific frequency (rather than the one it picks) you
can convert to zooreg first specifying that frequency.

Also please try to provide code in such a manner that
one can simply copy it from the post and paste it into
one's R session to reproduce the problem using the
style shown below.  I have also illustrated some
simplifications:

library(zoo)
library(chron)

la_time_test2 <- "TRIP_ID,Ruta_datetime_start,Minutos de viaje
28/04/2004_CM2-CM1_downriver_08:15,28/04/2004 08:15,125
28/04/2004_CM1-CM2_upriver_13:05,28/04/2004 13:05,295
"

t.df <- read.csv(textConnection(la_time_test2))
t.zoo <- zoo(test2.df[,3], as.chron(strptime(test2.df[,2], "%d/%m/%Y %H:%M")))
t.ts <- as.ts(as.zooreg(test2.zoo, freq = 24 * 365))
On Thu, Mar 6, 2008 at 5:00 PM, Darren Norris <doon75 at hotmail.com> wrote: