Time Series Analysis
Section 2 Correlation
3.5 Summary of important functions
Section 4 Basic Stochastic Models
Section 5 Regression
Section 6 Stationarity
Datasets (massey.ac.nz)
wine.dat #Australian wines
global.dat
CBE.dat #chocolate,Beer and electricity
Herald.dat #Carbon dioixde emissions at Herald Square, Manhattan.
Section 1
1.1 Purpose of Time Series Data
1.2 Time series
1.3 The R language
R can be insalled free of charge from www.r-project.org
An online guide "An Introduction to R" can be access by typing
help.start() at the command prompt to access this.
R is case sensitive.
A convention is to use define a variable name with a capital letter.
This reduces the chance of overwriting inbuild R functions, which are
usually written in lowercase letters.
Functions in R can be treated as "objects" that can be manipulated or
used recursively.
R shares many aspects with both Object orietnated and Functional
programming langguages.
all data in R is stored an objects, which have a range of "methods" available.
The "class" of an object can be found using the class() function.
1.4 Plots Trends and Seasonal Variation
dataset: Air Passengers
#########################
data(AirPassengers)
AP<-Air.Passengers
AP
######################################
plot(decompose(Elec.ts))
Elec.decom <- decompose(Elec.ts, type = "mult")
plot(Elec.decom)
Trend<-Elec.decom$trend
Seasonal<-Elec.decom$seasonal
ts.plot(cbind(Trend,Trend*Seasonal),lty =1:2)
Section 2 Correlation
Correllelograms
2.2.2 The ensemble and stationarity
The mean function of a time series model is a function of
(t) =E(xt)
2.2.3 ergodic series
2.2.5 Summary of useful function
cor
cov
Section 3 Forecasting strategies
3.3 The Bass Model
f(t) density
F(t)cumulative distribution function
h(t) hazard function
3.4 Exponential Smoothing and Holt-Winters
3.4.1. Exponential smoothing
3.5 Summary of important functions
nls()
predict()
coef()
ts.union() union of two time series analysis
The CCf
################################
ccf
Section 4 Basic Stochastic Models
4.3 Random Walks
xt=xt-1+wt
The backward shift operator
B.xt=xt-1
The differenceoperator
Section 5 Regression
Section 6 Stationarity
Section 7