[R] writing autocorrelation and partial auto correlation functions to a file

Glen Barnett glnbrntt at gmail.com
Thu May 20 11:51:14 CEST 2010


The problem is that the acf function (like many R functions) returns a list
containing many different things. For example, I have a short series in
the vector z:

> acz <- acf(z)
> str(acz)
List of 6
 $ acf   : num [1:11, 1, 1] 1 -0.0668 -0.7401 0.0627 0.5954 ...
 $ type  : chr "correlation"
 $ n.used: int 11
 $ lag   : num [1:11, 1, 1] 0 1 2 3 4 5 6 7 8 9 ...
 $ series: chr "z"
 $ snames: NULL
 - attr(*, "class")= chr "acf"

Let's say you are interested in extracting the acf and lag values from that

> acd <- data.frame(cbind(lag=acz$lag,acf=acz$acf))
> acd
   lag         acf
1    0  1.00000000
2    1 -0.06676557
3    2 -0.74006273
4    3  0.06266564
5    4  0.59542482
6    5 -0.06786556
7    6 -0.41370293
8    7  0.03464621
9    8  0.21887326
10   9 -0.07485743
11  10 -0.04835570



More information about the R-help mailing list