Skip to content

Filling missing days in xts time series

2 messages · thierrydb, R. Michael Weylandt

thierrydb
# 
Hi, 


I have a bunch of irregularly spaced xts time series (with a POSIX index),
and I'm trying to write a function that fillls the missing days. Using a
solution suggested by Gabor Grothendieck for zoo, I wrote the following:

# FD: Fill missing days
FD<-function(ser) {rng<-range(time(ser))
However,using S&P end-of-day time series, I obtain doubled entries such as:

2011-09-29 00:00:00 1160.40
2011-09-29 02:00:00 1131.42
2011-09-30 00:00:00 1131.42
2011-09-30 02:00:00 1099.23
2011-10-01 02:00:00 1099.23
2011-10-02 02:00:00 1099.23

It looks like it is due to the way xts handles time zones. What is the
correct way to do this?


Thanks

TDB

--
View this message in context: http://r.789695.n4.nabble.com/Filling-missing-days-in-xts-time-series-tp3884830p3884830.html
Sent from the R help mailing list archive at Nabble.com.
R. Michael Weylandt
# 
Can you provide an example of where using the zoo methods directly on
the xts object fails to work? You'll note that

R> X = xts(1:5, Sys.Date()+(1:5)/2)

R> class(X)
[1] "xts" "zoo"

so the zoo methods should work automatically. This gives you access to
all sorts of options for handling NAs.

Michael
On Sat, Oct 8, 2011 at 6:53 AM, thierrydb <thierrydb at gmail.com> wrote: