Plots of Sample Autocorrelation Function       

  rain.nyc1: A dataset included in Splus. It is the yearly total 
of New York City precipitation, in inches, from 1869-1957.

> par(mfrow=c(2,2))       # Sets up a 2 x 2 plotting region
> tsplot(rain.nyc1)
> mtext("Annual Rainfall: NY City")

 "acf" is the Splus command for calculating, and plotting, the sample
autocorrelation function. It has a default setting for deciding at how
many lags the sample autocorrelation is calculated.

> acf(rain.nyc1)
Call: acf(x = rain.nyc1)

Autocorrelation matrix:
   lag rain.nyc1
 1   0    1.0000
 2   1    0.1091
 3   2   -0.0655
 4   3   -0.2355
 5   4    0.0507
 6   5    0.0519
 7   6    0.0496
 8   7   -0.0167
 9   8   -0.0630
 .
 .
 .
20  19    0.0019

 ##############################################

 The data say.wavelet is the signal of a voice saying the word 
"wavelet" , sampled at 11025 Hz.  There are 8192 samples, ranging in 
time from 0 to approximately 0.7429 seconds. We will look at 50 of the
samples for illustration. 

> plot(say.wavelet[2501:2550,])
 
 In acf, we can tell Splus the maximum number of lags at which to
calculate the acf. We do this using the "lag.max" option.

> acf(say.wavelet[2501:2550,], lag.max = 7) 
Call: acf(x = say.wavelet[2501:2550,  ], lag.max = 7)
 
Autocorrelation matrix:
  lag say.wavelet
1   0      1.0000
2   1      0.6997
3   2      0.3161
4   3      0.1302
5   4      0.2183
6   5      0.2288
7   6      0.0426
8   7     -0.1740