Geographic Data Science - Lecture V Space, formally Dani - - PowerPoint PPT Presentation

geographic data science lecture v
SMART_READER_LITE
LIVE PREVIEW

Geographic Data Science - Lecture V Space, formally Dani - - PowerPoint PPT Presentation

Geographic Data Science - Lecture V Space, formally Dani Arribas-Bel Today The need to represent space formally Spatial weights matrices What Why Types The spatial lag The Moran Plot Space, formally For a statistical method to be


slide-1
SLIDE 1

Geographic Data Science - Lecture V

Space, formally

Dani Arribas-Bel

slide-2
SLIDE 2

Today

The need to represent space formally Spatial weights matrices What Why Types The spatial lag The Moran Plot

slide-3
SLIDE 3

Space, formally

slide-4
SLIDE 4

For a statistical method to be explicitly spatial, it needs to contain some representation of the geography, or spatial context

slide-5
SLIDE 5

For a statistical method to be explicitly spatial, it needs to contain some representation of the geography, or spatial context One of the most common ways is through Spatial Weights Matrices

slide-6
SLIDE 6

(Geo)Visualization: translating numbers into a (visual) language that the human brain "speaks better" Spatial Weights Matrices: translating geography into a (numerical) language that a computer "speaks better".

slide-7
SLIDE 7

Core element in several spatial analysis techniques: Spatial autocorrelation Spatial clustering / geodemographics Spatial regression

slide-8
SLIDE 8

W as a formal representation of space

slide-9
SLIDE 9

W

N x N positive matrix

slide-10
SLIDE 10

W

N x N positive matrix that contains spatial relations

slide-11
SLIDE 11

W

N x N positive matrix that contains spatial relations between all the observations in the sample

slide-12
SLIDE 12

W

N x N positive matrix that contains spatial relations between all the observations in the sample wii = 0 by convention

= { } wij x > 0 if i and j are neighbors

  • therwise
slide-13
SLIDE 13

W

N x N positive matrix that contains spatial relations between all the observations in the sample wii = 0 by convention ...What is a neighbor???

= { } wij x > 0 if i and j are neighbors

  • therwise
slide-14
SLIDE 14

Types of W

A neighbor is "somebody" who is: Next door Close In the same "place" as us ...

slide-15
SLIDE 15

Types of W

A neighbor is "somebody" who is: Next door → Contiguity-based Ws Close In the same "place" as us ...

slide-16
SLIDE 16

Types of W

A neighbor is "somebody" who is: Next door → Contiguity-based Ws Close → Distance-based Ws In the same "place" as us ...

slide-17
SLIDE 17

Types of W

A neighbor is "somebody" who is: Next door → Contiguity-based Ws Close → Distance-based Ws In the same "place" as us → Block weights ...

slide-18
SLIDE 18

Contiguity-based weights

Sharing boundaries to any extent Rook Queen ...

slide-19
SLIDE 19
slide-20
SLIDE 20

Distance-based weights

Weight is (inversely) proportional to distance between observations Inverse distance (threshold) KNN (fixed number of neighbors) ...

slide-21
SLIDE 21
slide-22
SLIDE 22
slide-23
SLIDE 23

Block weights

Weights are assigned based on discretionary rules loosely related to geography For example: LSOAs into MSOAs Post-codes within city boundaries Counties within states ...

slide-24
SLIDE 24
slide-25
SLIDE 25

Other types of weights

Combinations of the above Kernel Statistically-derived ... See for an in-detail discussion. Anselin & Rey (2014)

slide-26
SLIDE 26

How much of a neighbor?

No neighbors receive zero weight: wij = 0 Neighbors, it depends, wij can be: One wij = 1 → Binary Some proportion (0 < wij < 1, continuous) which can be a function of: Distance Strength of interaction (e.g. commuting flows, trade, etc.) ...

slide-27
SLIDE 27

Choice of W

Should be based on and reflect the underlying channels of interaction for the question at hand. Examples: Processes propagated by inmediate contact (e.g. disease contagion) → Contiguity weights Accessibility → Distance weights Effects of county differences in laws → Block weights

slide-28
SLIDE 28

Do your own (contiguity) weights time!

slide-29
SLIDE 29
slide-30
SLIDE 30
slide-31
SLIDE 31

Standardization

In some applications (e.g. spatial autocorrelation) it is common to standardize W The most widely used standardization is row-based: divide every element by the sum of the row: where is the sum of a row.

= wij ¯ wij wi⋅ wi⋅

slide-32
SLIDE 32

The spatial lag

slide-33
SLIDE 33

The spatial lag

Product of a spatial weights matrix W and a given variably Y

slide-34
SLIDE 34

The spatial lag

Product of a spatial weights matrix W and a given variably Y Ysl = WY ysl − i = ∑jwijyj

slide-35
SLIDE 35

Measure that captures the behaviour of a variable in the neighborhood of a given observation i. If W is standardized, the spatial lag is the average value of the variable in the neighborhood

slide-36
SLIDE 36

Common way to introduce space formally in a statistical framework Heavily used in both ESDA and spatial regression to delineate neighborhoods. Examples: Moran's I LISAs Spatial models (lag, error...)

slide-37
SLIDE 37

Moran Plot

slide-38
SLIDE 38

Moran Plot

Graphical device that displays a variable on the horizontal axis against its spatial lag on the vertical one Usually, variables are standardized ( ), which divides the space into quadrants Tool to start exploring spatial autocorrelation

y − mean(y) std(y)

slide-39
SLIDE 39

Moran Plot

slide-40
SLIDE 40

Moran Plot

slide-41
SLIDE 41

Recapitulation

Spatial Weights matrices: matrix encapsulation of space Different types for different cases Useful in many contexts, like the spatial lag and Moran plot, but also many other things!

slide-42
SLIDE 42

Geographic Data Science'16 - Lecture 5 by is licensed under a . Dani Arribas-Bel Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License