Semi-Regular Time Series with Missing Values

Hi,

I'm trying to make a ts object that has both NA values and a frequency other
than 1 (so I can use stl). I've tried all permutations I can think of, but
cannot get the desired (expected?) results.

The values live in x and the corresponding semi-regular time stamps are in
t:

library('zoo')
z = zoo(x, order.by=t, frequency=24)
zzr = as.zooreg(z, start=0)
zr = zooreg(x, order.by=t, start=0, frequency=24)
zrz = as.zoo(zr)

The objects z, zr, and zzr all have frequency 24, as desired, and plot
correctly. Object zrz plots correctly but has frequency 1. Now I try to make
it a ts object:

zrt = ts(zzr)

... omits the missing values.

zrt = ts(zzr, frequency=24)

... omits the missing values and changes the timestamps.

plot(ts(zr))

... omits the missing values.

zrt = ts(zr, frequency=24)

... omits the missing values and changes the timestamps.

zrt = as.ts(zr)

... inserts 23 NA values between each legitimate value.

zrt = as.ts(zr, frequency=24)

... inserts 23 NA values between each legitimate value.

zrzt = ts(zrz)

... omits the missing values.

zrzt = ts(zrz, frequency=24)

... omits the missing values and changes the timestamps.

I don't seem to have that method:

zts = as.ts.zoo(z)

Error: could not find function "as.ts.zoo"

I'm finding forum posts from people trying to use stl directly with zoo[reg]
functions (despite the documentation specifying ts) but discovering that stl
gives errors with NA entries. I encountered similar errors.

Can anyone on this list confirm whether stl works with NA values? If so,
could someone please point me to a working example?

This still does not give me the desired behavior:

z = zoo(x, order.by=t, frequency=24)
t = as.ts(z)

... inserts 23 NA values between every actual value. This is not correct;
the original data has a frequency of 24 and doesn't need one forced upon it
during coercion.

z = zoo(x, order.by=t)
t = as.ts(z)

... obviously doesn't give me a ts with frequency 24.

Keep in mind that I'm trying to get a ts with frequency 24 so I can feed it
to stl. I haven't yet found a sequence of operations that lets me do this.
Any ideas?

library(zoo)
z <- zooreg(1001:1012, frequency = 4); z

frequency(z)

tt <- as.ts(z); tt

frequency(tt)

Gabor,

If you follow those steps on a series with missing values, the resulting tt
will contain tons of new NAs. (See my original email and the call to "zrt =

My series was regularly spaced, it simply contained missing values. I don't
see why the new values were added.

Besides, my previous email showed exactly how to convert a zoo object with

The objective is to get the stl calls to work without making up values for
the NAs:

library("zoo")
data = c(50, 32, 22, 13, 14, 6, 12, 3, 6, 21, NA, NA, 15, 13, 17, 9, 11, 34,
63)
dataz = zoo(data, 1:length(data))
t = as.ts(dataz)
datat = as.ts(dataz)
datatt = ts(t, frequency=2)
frequency(datatt)
length(datatt)
# these all give errors
stl(datatt, s.window="per")
# this next error is especially bizarre to me, because datatt is periodic
with more than two periods
stl(datatt, s.window="per", na.action=F)
stl(datatt, s.window="per", na.action=na.omit)
stl(datatt, s.window="per", na.action=na.pass)
stl(datatt, s.window="per", na.action=na.exclude)
stl(dataz, s.window="per", na.action=F)
stl(datat, s.window="per", na.action=F)

I was expecting at least one of the stl calls to work, because stl claims to
require a ts object with a frequency greater than one and at least two
periods (which I gave it) and it claims to handle NA values.