R Dataset / Package DAAG / bomsoi2001
On this Rdata statistics page, you will find information about the bomsoi2001 data set which pertains to Southern Oscillation Index Data. The bomsoi2001 data set is found in the DAAG R package. You can load the bomsoi2001 data set in R by issuing the following command at the console data("bomsoi2001"). This will load the data into a variable called bomsoi2001. If R says the bomsoi2001 data set is not found, you can try installing the package by issuing this command install.packages("DAAG") and then attempt to reload the data with the library() command. If you need to download R, you can go to the R project website. You can download a CSV (comma separated values) version of the bomsoi2001 R data set. The size of this file is about 8,046 bytes.
Southern Oscillation Index Data
Description
The Southern Oscillation Index (SOI) is the difference in barometric pressure at sea level between Tahiti and Darwin. Annual SOI and Australian rainfall data, for the years 19002001, are given. Australia's annual mean rainfall is an areaweighted average of the total annual precipitation at approximately 370 rainfall stations around the country.
Usage
bomsoi2001
Format
This data frame contains the following columns:
 Year

a numeric vector
 Jan

average January SOI values for each year
 Feb

average February SOI values for each year
 Mar

average March SOI values for each year
 Apr

average April SOI values for each year
 May

average May SOI values for each year
 Jun

average June SOI values for each year
 Jul

average July SOI values for each year
 Aug

average August SOI values for each year
 Sep

average September SOI values for each year
 Oct

average October SOI values for each year
 Nov

average November SOI values for each year
 Dec

average December SOI values for each year
 SOI

a numeric vector consisting of average annual SOI values
 avrain

a numeric vector consisting of a weighted average annual rainfall at a large number of Australian sites
Source
Australian Bureau of Meteorology web pages:
http://www.bom.gov.au/climate/change/rain02.txt and http://www.bom.gov.au/climate/current/soihtm1.shtml
References
Nicholls, N., Lavery, B., Frederiksen, C.\ and Drosdowsky, W. 1996. Recent apparent changes in relationships between the El Nino – southern oscillation and Australian rainfall and temperature. Geophysical Research Letters 23: 33573360.
See Also
bomsoi
Examples
bomsoi < bomsoi2001 plot(ts(bomsoi[, 15:14], start=1900), panel=function(y,...)panel.smooth(1900:2001, y,...)) pause()# Check for skewness by comparing the normal probability plots for # different a, e.g. par(mfrow = c(2,3)) for (a in c(50, 100, 150, 200, 250, 300)) qqnorm(log(bomsoi[, "avrain"]  a)) # a = 250 leads to a nearly linear plotpause()par(mfrow = c(1,1)) plot(bomsoi$SOI, log(bomsoi$avrain  250), xlab = "SOI", ylab = "log(avrain = 250)") lines(lowess(bomsoi$SOI)$y, lowess(log(bomsoi$avrain  250))$y, lwd=2) # NB: separate lowess fits against time lines(lowess(bomsoi$SOI, log(bomsoi$avrain  250))) pause()xbomsoi < with(bomsoi, data.frame(SOI=SOI, cuberootRain=avrain^0.33)) xbomsoi$trendSOI < lowess(xbomsoi$SOI)$y xbomsoi$trendRain < lowess(xbomsoi$cuberootRain)$y rainpos < pretty(bomsoi$avrain, 5) with(xbomsoi, {plot(cuberootRain ~ SOI, xlab = "SOI", ylab = "Rainfall (cube root scale)", yaxt="n") axis(2, at = rainpos^0.33, labels=paste(rainpos)) ## Relative changes in the two trend curves lines(lowess(cuberootRain ~ SOI)) lines(lowess(trendRain ~ trendSOI), lwd=2) }) pause()xbomsoi$detrendRain < with(xbomsoi, cuberootRain  trendRain + mean(trendRain)) xbomsoi$detrendSOI < with(xbomsoi, SOI  trendSOI + mean(trendSOI)) oldpar < par(mfrow=c(1,2), pty="s") plot(cuberootRain ~ SOI, data = xbomsoi, ylab = "Rainfall (cube root scale)", yaxt="n") axis(2, at = rainpos^0.33, labels=paste(rainpos)) with(xbomsoi, lines(lowess(cuberootRain ~ SOI))) plot(detrendRain ~ detrendSOI, data = xbomsoi, xlab="Detrended SOI", ylab = "Detrended rainfall", yaxt="n") axis(2, at = rainpos^0.33, labels=paste(rainpos)) with(xbomsoi, lines(lowess(detrendRain ~ detrendSOI))) pause()par(oldpar) attach(xbomsoi) xbomsoi.ma0 < arima(detrendRain, xreg=detrendSOI, order=c(0,0,0)) # ordinary regression modelxbomsoi.ma12 < arima(detrendRain, xreg=detrendSOI, order=c(0,0,12)) # regression with MA(12) errors  all 12 MA parameters are estimated xbomsoi.ma12 pause()xbomsoi.ma12s < arima(detrendRain, xreg=detrendSOI, seasonal=list(order=c(0,0,1), period=12)) # regression with seasonal MA(1) (lag 12) errors  only 1 MA parameter # is estimated xbomsoi.ma12s pause()xbomsoi.maSel < arima(x = detrendRain, order = c(0, 0, 12), xreg = detrendSOI, fixed = c(0, 0, 0, NA, rep(0, 4), NA, 0, NA, NA, NA, NA), transform.pars=FALSE) # error term is MA(12) with fixed 0's at lags 1, 2, 3, 5, 6, 7, 8, 10 # NA's are used to designate coefficients that still need to be estimated # transform.pars is set to FALSE, so that MA coefficients are not # transformed (see help(arima))detach(xbomsoi) pause()Box.test(resid(lm(detrendRain ~ detrendSOI, data = xbomsoi)), type="LjungBox", lag=20)pause()attach(xbomsoi) xbomsoi2.maSel < arima(x = detrendRain, order = c(0, 0, 12), xreg = poly(detrendSOI,2), fixed = c(0, 0, 0, NA, rep(0, 4), NA, 0, rep(NA,5)), transform.pars=FALSE) xbomsoi2.maSel qqnorm(resid(xbomsoi.maSel, type="normalized")) detach(xbomsoi)
Dataset imported from https://www.rproject.org.