Remove NA from a "ts" object to make acf() and pacf() work [closed]

Question

How to remove NA from this time series?

x <- structure(c(NA, 15.3471263711111, 23.7495828967391, 17.6522039782609, 
    13.950342245, 18.6679701043956, 21.7398276684783, 22.4975495326087, 
    17.1164915533333, 17.7142983241758, 20.5018974402174, 20.841587, 
    18.9183467888889, 15.4600196181319, 17.8721988695652, 20.0170066630435, 
    18.5086658241758, 15.5306393346154, 16.0669595836957, 20.2165807554348, 
    16.25535685, 14.8567082505495, 15.2826896483696, 12.1608449365543, 
    8.88960227944444, 16.1804220681319, 19.7431830652174, 15.0564858630435, 
    15.1309696333333, 15.9665181153846, 17.7906991630435, 18.3519159722826, 
    15.7932861428571, 11.9033367478022, 16.378045125, 16.4570617070652, 
    11.8673526106667, 17.0768826978022, 17.6064028967391, 18.8143230434783, 
    15.1196432378882, NA, NA, NA), .Tsp = c(2004, 2014.75, 4), class = "ts")

#          Qtr1      Qtr2      Qtr3      Qtr4
#2004        NA 15.347126 23.749583 17.652204
#2005 13.950342 18.667970 21.739828 22.497550
#2006 17.116492 17.714298 20.501897 20.841587
#2007 18.918347 15.460020 17.872199 20.017007
#2008 18.508666 15.530639 16.066960 20.216581
#2009 16.255357 14.856708 15.282690 12.160845
#2010  8.889602 16.180422 19.743183 15.056486
#2011 15.130970 15.966518 17.790699 18.351916
#2012 15.793286 11.903337 16.378045 16.457062
#2013 11.867353 17.076883 17.606403 18.814323
#2014 15.119643        NA        NA        NA

This is causing me trouble:

acf(x)
#Error in na.fail.default(as.ts(x)) : missing values in object

pacf(x)
#Error in na.fail.default(as.ts(x)) : missing values in object

Orvilleorwell

Answer 1

Use window to select non-NA segments:

xx <- window(x, start = c(2004, 2), end = c(2014, 1))

acf(xx)
pacf(xx)

Or keep these NAs but use:

acf(x, na.action = na.pass)
pacf(x, na.action = na.pass)

Specter

Answer 2

x is a ts object, not a data.frame.

class(x)
## [1] "ts"

To get rid of leading and trailing NA's, assuming there are no other NA's, use na.omit like this:

na.omit(x)

Note

If there were internal NA's, i.e. NA's which were not leading or trailing, then use one of these all of which remove leading and trailing NA's plus the described action

library(zoo)

xx <- replace(x, 10, NA) # add an internal NA

na.trim(xx)    # preserve internal NA's
na.approx(xx)  # fill in internal NA's using linear interpolation

na.spline.ts <- function(x, ...) as.ts(na.spline(as.zoo(x), ...))
na.spline(na.trim(xx))  # spline interpolation

Elastic

na.omit does not omit internal NA's, i.e. NA's not at the ends, of ts objects. It gives an error in that case. – Elastic

But there are no internal NA's so it isn't needed here. – Elastic

na.spline fills all NA, and loses "ts" class. Is this expected? I am using zoo 1.8.10 (the latest). – Specter

I think there was suppo9sed to have been a ts method for that but it looks like there is no na.spline.ts. Have fixed example. – Elastic