+ All Categories
Home > Documents > Statistics 248 D.R. Brillinger

Statistics 248 D.R. Brillinger

Date post: 06-Jan-2016
Category:
Upload: khanh
View: 40 times
Download: 2 times
Share this document with a friend
Description:
Statistics 248 D.R. Brillinger. Process data indexed values Y( ) ,  : time, space, set, function. curves, surfaces, shapes, measures, images, sequences, scatter, spirals, trajectories, mosaics - and they may be moving. Random process data. - PowerPoint PPT Presentation
17
Statistics 248 D.R. Brillinger Process data indexed values Y(), : time, space, set, function curves, surfaces, shapes, measures, images, sequences, scatter, spirals, trajectories, mosaics - and they may be moving Random process data Put process datums in an urn and pick one, Y(;), a random variable time series, point processes, marked point processes, spatial processes, spatial-temporal processes, categorical-valueds, hybrids, ...
Transcript
Page 1: Statistics 248       D.R. Brillinger

Statistics 248 D.R. Brillinger

Process data

indexed values Y(), : time, space, set, function

curves, surfaces, shapes, measures, images, sequences, scatter, spirals, trajectories, mosaics - and they may be moving

Random process data

Put process datums in an urn and pick one,

Y(;), a random variable

time series, point processes, marked point processes, spatial processes, spatial-temporal processes, categorical-valueds, hybrids, ...

Page 2: Statistics 248       D.R. Brillinger

Time series data.

t: time (but could be space)

0 t T (continuous)

t = 0,...,T-1 (discrete)

t = 1 ,..., N unequally spaced, missings

(unique) value Y(t)

Y in R

{0,1}

Rp

N, category, [0, 2)

Page 3: Statistics 248       D.R. Brillinger

Tufte (1983). Perhaps 10th or 11th century movement of planets and sun

Page 4: Statistics 248       D.R. Brillinger

Symbol graph. Good for long term behavior

cannot appreciate middle and high frequency behavior

irregular sampling can be unclear

plot(type="p",...) seals T = 940

Elephant seal dives

Page 5: Statistics 248       D.R. Brillinger

Connected. Good for smooth series

individual data points not unambiguously displayed

irregular sampling can be unclear

plot(type="l",

Page 6: Statistics 248       D.R. Brillinger

Both. Points and line segments

plot(type="b",

Page 7: Statistics 248       D.R. Brillinger

Both. Points on top of connected

plot(type="o",

Page 8: Statistics 248       D.R. Brillinger

plot(type="n"

points(pch="*"

Page 9: Statistics 248       D.R. Brillinger

plot(type="n"

lines(type=3

Page 10: Statistics 248       D.R. Brillinger

plot(type="h"

Good when need to see individual values

when series long

about centralvalue (next)

Not good when strong trend

Page 11: Statistics 248       D.R. Brillinger
Page 12: Statistics 248       D.R. Brillinger

plot(type="h",...) Amazon T = 32875

Good about central value

Amazon River at Manaus

Page 13: Statistics 248       D.R. Brillinger

Mississippi Locks

Page 14: Statistics 248       D.R. Brillinger

Vector ts case matplot(type="l"

use several line types

colors

use to provide comparison

Page 15: Statistics 248       D.R. Brillinger

Hipel K. W. McLeod A. I. (1994). Time Series Modelling of Water Resources and Environmental Systems Elsevier

Buys-Ballot

Useful when there is special period P

t=nP+p, p=0,...,P-1

n=0,1,2,...

Yn(p)=Y(nP+p)

Page 16: Statistics 248       D.R. Brillinger

Computer software

Ross Ihaka is one the originators of R, a software package for statistical computing that has had phenomenal uptake internationally. It can be downloaded free and easily customised for a very wide variety of applications. The package and the paper introducing it have been cited over 1700 times, by far the highest for publications in the mathematical sciences over the last ten years, worldwide. It is now disseminated from over 75 internet sites in 30 countries.

R

CRAN

Page 17: Statistics 248       D.R. Brillinger

The package is used both for teaching and research by hundreds of universities around the world, including Stanford, Oxford, Cambridge and Berkeley. There are over 40 books written about, or featuring, the use of R.

http://www.nytimes.com/2009/01/07/technology/business-computing/07program.html?scp=2&sq=ihaka&st=cse


Recommended