open spatial analysis with ron buliung assistant professor department of geography university of...

13
Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga [email protected]

Upload: bruce-martin

Post on 17-Dec-2015

216 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Open Spatial Analysis with

Ron BuliungAssistant Professor

Department of GeographyUniversity of Toronto at

[email protected]

Page 2: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Outline

What, Why, Where?

Some Key Projects

What Have I Been Up To?

Resources

Page 3: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

What is “R”?

Integrated software suite for data manipulation, calculation and graphical display:

Data handling and storage.

Operators for calculations on arrays and matrices.

Collection of tools for data analysis.

Graphing and data visualization.

Complete programming language

Domain specific sub-projects

Page 4: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

What does R Look Like?

Page 5: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Can also Look Like This

Rcmdr – John Fox

Page 6: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Why Adopt “R”?

Analytically Comprehensive FREE, and open source Multi platform support Flexibility Vast array of learning resources Continuity of Use Culture of Use Analytical Clarity Introducing, my international

team of collaborators!!

Page 7: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Drawbacks?

User interface R-GUIs: http://www.sciviews.org/_rgui/

Problems with packages Will it always be around? Problems with large problems? A non-

starter (see object.size, memory.size, memory.limits)

Can’t leave home without it?

Page 8: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Where can you learn more?

http://r-spatial.sourceforge.net/

http://sal.uiuc.edu/csiss/Rgeo//

Page 9: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Some Key Projects

sp splancs Spatstat maptools Shapefiles RpostGIS Spgrass6 Aspace adehabitat GeoDa?? STARS

OS OS

OS Proprietary

Page 10: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Teaching

Working Closer to the Data “Looking under the hood” Knowledge Spillovers Skills are in demand

GGR276: Spatial Data Analysis and Mapping

GGR370: Planes, Trains and Automobiles: The Geography of Transportation

580000 600000 620000 640000 660000 680000 7000004

78

00

00

48

20

00

04

86

00

00

49

00

00

0

GTA Landfill Sites

X-coords

Y-c

oo

rds

Page 11: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Teaching

#next two lines should be written into R as a single line

mpb <- read.table("C:/bc/mpb.txt", header=TRUE, sep=",", na.strings="NA", dec=".", strip.white=TRUE)

#next two lines should be written into R as a single line

plot(mpb$X_coord,mpb$Y_coord,xlab="X-coords",ylab="Y-coords",main="Pine Beetle Data")

n <- nrow(mpb[1])

n

mean(mpb$X_coord);mean(mpb$Y_coord)

mc.x<-sum(mpb$X_coord)/n

mc.y<-sum(mpb$Y_coord)/n

points(mc.x,mc.y,'p',pch=22,cex=2,col="blue",bg="blue")

wmc.x<-sum(as.numeric(mpb$Tree_no*mpb$X_coord))/sum(mpb$Tree_no)

wmc.y<-sum(as.numeric(mpb$Tree_no*mpb$Y_coord))/sum(mpb$Tree_no)

points(wmc.x,wmc.y,'p',pch=22,cex=2,col="red",bg="blue")

Deconstruct, Modify, Implement

Page 12: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Research: Human Spatial Behaviour

605000 610000 615000 620000

48

26

00

04

83

00

00

48

34

00

04

83

80

00

Easting (m)

No

rth

ing

(m

)

Graphic Results: Household MCP

570000 590000 610000 630000

47

90

00

04

81

00

00

48

30

00

0

Easting (m)

No

rth

ing

(m

)

Graphic Results:Household SDE

605000 610000 615000 620000

48

25

00

04

83

00

00

48

35

00

04

84

00

00

Easting (m)

No

rth

ing

(m

)

Graphic Results: Household SDD

Aspace:

Open Source toolkit for geographic visualization and analysis of activity spaces – i.e., the set of locations visited by a behavioural agent (e.g., person, household) over time.

Page 13: Open Spatial Analysis with Ron Buliung Assistant Professor Department of Geography University of Toronto at Mississauga ron.buliung@utoronto.ca

Resources R-download http://www.r-project.org/

R-packages http://probability.ca/cran/ R-Spatial http://sal.uiuc.edu/csiss/Rgeo/

R-GUIs: http://www.sciviews.org/_rgui/

John Fox, Sociology, McMaster University: http://socserv.socsci.mcmaster.ca/jfox/

“R for Beginners” Emmanuel Paradis http://cran.r-project.org/ {click on Contributed

on left panel.}