Lecture 7 + Chapter 6 Flashcards
what is the purpose of a point pattern analysis
examines the spatial arrangement of locations - objects and events
- all within a single theme - disregards attribute variation
EX: distribution of burglaries near the river in London - your focusing on the location of where burglaries happen not species attributes like what stoles items and their values
what are the 3 times of point pattern analysis
- Random - points distributed without correlation to each other
- taking a random survey in random locations for an unbiased representation - Uniform
- even distribution through space
- fire stations across the city for the best coverage - Clustered
- locations close together
- burglaries targeting a rich region
- disease spread
what are the 5 types of analysis
point pattern
autocorrelation
proximity
correlation
combining
what does autocorrelation mean
examines both location and attribute distribution over areas - often with census demographics
it examines if these similar values are more likely to occur near each other
EX: Checking if high-income areas are clustered together in a city map
what is a point map
specific locations for individual things
what is an area map
Aggregated data over an area - burglary rates by neighbourhood
Determines high-crime neighborhoods for resource allocation
what are 3 autocorrelation patterns
- Negative
Uniform distribution - no clustering - Positive
Clustering of similar attribute values - No Autocorrelation
No clear pattern of relation between values
what is proximity analysis
analyzes spatial relationships between 2 themes - exploring how proximity impacts patterns over time and distance
EX: public health - John Snow Cholera Outbreak
- showed concentrations of cases near the sewer system pump proving that the disease was waterborne
- city layouts use proximity for ambulance routes
- using straight line distance to connect two points to see the distance crows fly
- using network distance - the distance over a transportation network to reach somewhere
what is correlation analysis
measures spatial relationships between multiple attributes
- This type of analysis helps to understand how different variables might be interrelated within the same space
- correlation does not mean the cause - it is the additional data that is needed to make causation claims
- data must be comparable - needs to have good spatial, attribute, temporal and interoperability
EX: Areas with high income often correlate with high educational attainment
EX: demographic shifts in LA for black and Hispanic populations using census categories and tract
what does combining analyses mean
mixing methods and different analytical approaches can be combined to have a better product
explain a type of map that has combined analyses
police shooting map
- point pattern to show the location of fatal incidents
- autocorrelation of Hispanic population density distribution
- correlation of the relationship between Hispanic residents and income levels
why are overlapping analyses helpful
Overlap is often seen - both analyses help understand spatial relations in complex scenarios like shootings and demographic distributions