Context – The structure of interrelationships between data and how data is collected, processed, used and understood in an application – Data only works in context
Data – Organized measurements
Database – collection of data arranged in a way the computer can efficiently store and retrieve it
4 elements for a useful database: 1.Reliable 2. Correct and contained 3. Technology proof 4. Secure
Other database elements: 5. User views (levels of access) 6. User interface (ease of use)
7. Data Independence (encapsulation) 8.Self-describing (metadata) 9.Currency (multi users) 10. Distributed (remote access) 11. Performance (rapid retrieval)
Dbms – Database Management System – system to create and maintain databases
6 DBMS components: 1. Query language 2. Query compiler 3. Runtime database processor 4. Constraint enforcer 5. Stored data manager 6. System catalog/data dictionary
4 Types of DBMS 1. Heirarchical 2. Network 3. Relational 4. Object-Oriented
Application Domain Model –describes core requirements of users in a particular application domain – interface made for who will use it
Conceptual Model – tailored to a particular type of implementation
Physical computational model – the result of a process of programming (actual setup)
Transaction – unit of transaction between the user (with access) and the database
4 types of access (high to low level access) – 1.Insertion of new data 2. Modification of existing data 3. Deletion of data 4. Retrieval of data
Tuple – aka Row aka Record
Relation – aka table
Attribute – aka Column aka Field aka Item (in INFO)
Cardinality – Number of rows (Degree = # columns)
Schema – overall scheme/design for a table) – description of attributes (ie form – text, int, float pt, etc
Primary Key – liks a relation to another relation
Foreign Key – attribute in one table that can serve as a primary key in another table
Candidate Key – uniquely identifies each record in a table
Project – identified by pi symbol ( ) – Method of selection – produces a subset of attributes of the original relation – duplicate tuples are dropped in the smaller table (relation)
Restrict – identified by theda symbol ( ) – Method of selection with a condition – produces a subset of attributes of the original relation based on the condition (ex pop > 50K)
Union – relational operation – retains spatial and attribute data of bothojects/relations
Intersection – relational operation – retains spatial and attribute data of
Object – Thing/ entity that has attributes in a model system – an object has identity that is independent of its attribute values
Class – super/sub are levels of hierarchy that ‘end’ with an ‘instance’ that inherits properties of all class levels above it
Inheritance – very important attribute of object-oriented – allows objects to share common properties: single iheritance – zero or one super class; multiple inheritance – zero or multiple parent classes (requires protocol for resolving behavioral conflicts)
Polymorphism – Inheritance that allows objects to perform different roles within specific contexts inclusion polymorphism – where subclass is substituted for superclass; overloading polymorphism – where subclasses implement their own specialized version of general behaviors (es try to ope .tiff in ‘open vector’ application – will ‘on the fly’ id .tiff as raster and open it anyway)
Subclass – class that is ‘lower’ in hierarchy and has more specialized behaviors
Superclass – class that is ‘higher’ in hierarchy and has more generalized behaviors
Disjoint –
Overlapping –
Encapsulation – Internal structure of the object hidden from the user
Association – groups objects together to model phenomena with complex internal structure (homogenous – formed from objects of the same class) (can also be ordered – where order of component objects is important ex construction of a poly)
Aggregation – type of association concerned with part/whole relationships (partonomy – the whole is made of these parts)
Constructor – behaviors that are activated when an object is created (ex file, open) (destructors – activated when an object is destroyed – ex “are you sure you want to deledte”)
Accessors – behaviors used to examine the state of an object
Transformers – behaviors that change the state of the object
File Geodatabase – New version of personal db – supports many readers and a single editor- max out at 2G file size; container for other elements
Personal Geodatabase – many readers/one editor – uses Microsoft Access (.mdb) as dbms. Can be used with Arcview 9.0 without addtl software, but software making it obsolete.
ArcSDE Geodatabase (aka Arc GIS Server) – Multiple readers, multiple users, – requires ArcSDE editor for schema management/’end of day’ conflict resolution.
Feature Class – Group of points, lines or polys representing similar geographic features – may contain one or more features, but all features have same geometry (.shp = only one feature class; ArcInfo coverages, CAD files, and geodatabases may have several feature classes) *Like a layer
Feature Dataset – Feature classes that are spatial dependant on each other can be organized into larger units called feature data sets. Contain objects, features and their relationships.*Share common coordinate system * Like a sub-chunk of a geodatabase
Relationship Class – Stores relationships between entities in two object or feature classes.
*Object, feature and relationship classes can be contained inside or outside feature datasets
Range Domain – Allowable values for fields in feature or object classes *creates intelligent features (ex, road can’t cross a stream without a bridge)
One-to-Many Relationship – Record in one table refers (via primary/foreign key) to multiple records in a second table
One-to-One Relationship – Record in one table refers (via primary/foreign key) to single record in a second table
1. FOUR reasons to have a GIS: 1)Cost reduction – efficient use/allocation of resources 2)Cost avoidance – informed decision making 3)Increased revenue – ex. Ensuring people are taxed fairly 4)Intangible benefits – consistency, fairness, documentation
2. NINE conditions under which GIS could fail 1) Lack of awareness or commitment at executive level 2) Lack of or too much oversight of key participants (need for analysts to be able to analyze) 3)Inexperienced managers 4)Unsupportive organizational structure (need to view as a system) 5)Political or personal pressure 6) Lack of demonstrated benefits 7)Intervening non-GIS responsibilities (other office network/copier/etc tasks) 8)Poor planning 9)Insufficient long-term funding
3.List 6 steps going from RW to understanding 1) Milieu – infinitely complex real world 2) Measurement – accurate observation of one aspect of a phenomenon 3)Data Organized measurements 4) Information – summary and basic analysis of data 5)Knowledge – interpretation of information 6) Understanding – linking with other knowledge (7 = modeling/predicting!)
4. FOUR steps from RW to information on computer 1) Milieu – infinitely complex 2) Data Model – used to represent certain elements of the real world 3) Data Structure – how we arrange this representation conceptually in a computer 4) File structure – How data is placed on the computer disc or memory (1,0,0,1,0,0,1 – sos!)
5.File processing vs. computer as data repository: FP = Input, process, output – duplicates both data and processing, don’t necessarily need huge continuous data storage. DR = Users deposit, store and retrieve data in a way that avoids duplication. FP is like a printer and DR is like a library.
6. SEVEN characteristics of an effective database: : 1.Reliable 2. Correct and contained 3. Technology proof 4. Secure 5. User views (levels of access) 6. User interface (ease of use) 7. Data Independence (encapsulation) 8.Self-describing (metadata) 9.Currency (multi users) 10. Distributed (remote access) 11. Performance (rapid retrieval)
7. Sketch how dbms components are connected:
8. Describe how ArcInfo one-to-one requirement can be limiting – If there are multiple tuples in the destination table, only the first tuple would be shown and therefore not all of the information would be conveyed. This problem makes normalization essential.
9.Differentiate between Join and Relate – Relate operanions are a temporary linkage between the records in one table to another through a key field. Join operations add the fields from one table to the display of the other. ALWAYS join before relate, because a join will delete any relates
Surface or Field – Representation of plane with unique (one and only one) Z value for every X,Y location
Ontology – the branch of metaphysics that studies the nature of existence or being as such
Tuple – aka Row aka Record
Scalar – One number representation (elevations of a topographical surface are scalar)
Vector – geographical data model in which attribute is controlled and space (location). Can have multiple numbers representing the same location (x, y, z)
Manhattan Distance – Along a road or other surface network
Euclidean Distance – As the crow flies
Polyline – Set of edges (line segments) where each edge (except for the ends) is shared by exactly two end points (vertices) – simple polylines=no intersections except at ends, complex polylines = have lines that cross it at one or more location
*Convex polygon – Every interior angle is greater than 180 deg therefore every point is visible within the polygon from every other point within the polygon – useful for visibility studies
*Star-Shaped Polygon – at least one point is visible from every other point in the polygon
*Skeleton Plot – A network of lines in a polygon constructed so that each point on the network is equidistant from the nearest two edges on the polygon boundary and bisects angles (corners) – produces a useful center for labeling because it always lies within the polygon, unlike the mean center
*Monotone – A line is monotone if a projection of the vertices onto another line retains the ordering of the vertices
*Function Image –A type of relation in which each member of the first set relates to exactly one member of the second set “f(x)” domain = bigger set of values have function result of codomain and image
Injection – Any two different points in the domain are transformed to two distinct points in the codomain
Ex -2,4;2,4;2,-4: squared can’t be an injection because ≠ result in distinct points in codomain (all same)
Surjection – When the image equals the codomain
Bijection – A function that is both a surjection and an injection
Usual topology – property that describes relative spatial relationships. Reduces redundancy in file storage because each arc starts and ends with a node (vertices give them shape, listed in arc-coordinate list, but not in arc-node list). In an ArcInfo coverage, the spatial relationships between connecting or adjacent features in a geographic data layer (for example, arcs, nodes, polygons, and points). Topological relationships are used for spatial modeling operations that do not require coordinate information
Closure – a series of lines that return to an original node creating a polygon
Interior – Original representation of islands in topology. Interior boundary is denoted in Arc-node list by a “0” which indicates a ‘step-off’ the exterior poly boundary to describe the interior poly boundary
Geodatabase Topology – In geodatabases, the arrangement that constrains how point, line, and polygon features share geometry. For example, street centerlines and census blocks share geometry, and adjacent soil polygons share geometry. Topology defines and enforces data integrity rules (for example, there should be no gaps between polygons). It supports topological relationship queries and navigation (for example, navigating feature adjacency or connectivity), supports sophisticated editing tools, and allows feature construction from unstructured geometry (for example, constructing polygons from lines).
Tree – Hierarchical structure. Causal arrangement (parent, child inheritance relationship)
Graph or Network – A finite non-empty set of nodes or junctions together with a set of unordered pairs of distinct nodes called edges – “complicated way of saying-a graph that describes how things are connected”
*Edge – Unordered pairs of distinct nodes that represent links in a network. Simple edge features are associated with
a single edge in a logical network; complex edge features
are associated with any number of edges
*Junction – nodes through which links of a network are connected.
Simple junction features work as one-to-one location; complex junction features are associated with a collection of
junctions and edges and therefore requires a custom feature type to handle discrepancies of rule orders
Geometric Network – Set of features that participate in a linear system (connected system of edges and junctions) stored in a feature dataset with common coordinate system
Logical Network – A pure network graph (schematic) consisting of edge and junction elements that stores the connectivity information (what lines connect to others and how) not the coordinates – Encapsulated from the user (user interacts with geometric network)
Network Contexts – Transportation – direction of flow along edges determined by ‘driver’ therefore can have bi-directional flow; Utility – Direction of flow along edges (links) determined by the configuration of the sources, sinks and switches in the network
Source – A junction in a network from which a commodity flows
Sink – A junction in a network where commodity flows terminate
Flow – Flow may be indeterminate if the network configuration does not provide enough information to establish the direction of flow for an edge. Disabled features do not participate in the network flow (a closed street or a switched off circuit). Uninitiated flow to features is the case when a disabled junction feature cuts off flow to downstream edges. Barriers can be used during network configurations to temporarily disable a network element
Cost Raster An input dataset necessary to run the cost weighted distance function – identifies the cost of traveling through each cell in the raster. The cost weighted distance function uses this cost raster to calculate the accumulative cost of traveling from every cell in the raster to a source or a set of sources. cost-weighted distance An ArcGIS Spatial Analyst function that uses a cost grid to assign a value—the least accumulative cost of getting back to the source—to each cell of an output grid. cost-weighted direction
An ArcGIS Spatial Analyst function that provides a road map from the cost weighted distance grid, identifying the route to take from any cell, along the least-cost path, back to the nearest source. cost-weighted allocation An ArcGIS Spatial Analyst function that identifies the nearest source from each cell in a cost-weighted distance grid. Each cell assigned to nearest source cell, in terms of accumulated travel cost.
Weight – Express the cost (impedance) of traversing an edge or crossing a junction – exist as a function of the edge length, and there can be multiple weights associated with an element
Path – Shortest minimum impedance through a network where stops are visited in a predetermined order
Tour – 2 problems – Both the order in which to visit the stops and the shortest path between the stops is needed to solve. Heuristic procedure to find the most efficient path to a series of locations (aka Traveling Salesman Problem) (also, lawn mowing crew example ‘tired at end of day so want to end route close to the shop’). Large data sets result in optimal (not optimum) solution.
Allocation – Assign portions of the network to a resource supply location (Fire station). Locations and networks are given – as we travel we use the ‘supply’ of time available to get to a location
Tracing – Determine whether one location in a network is connected to another (ex – sewer line example in ESRI module – ‘find common ancestry’) ex- which facilities serve a particular customer, or which tributaries feed a particular extent of a river
Spatial Interaction – Involved in more complex studies – ex which hubs get most use from other hubs
Distance Matrix – Representative network of shortest paths between a point and all others in study area
Location-Allocation – 2 problems: where things go and what gets assigned to them (ex where to locate new fire stations); Goal is to locate facilities designed to supply a good or service for which there is a demand in the most efficient manner possible – these models optimize efficiency by simultaneously determining the configuration of the facilities and assigning the demand to the facilities. Determine the locations for centers and the allocation of demand to these centers according to a specific objective
Usually based on 3 main ESRI categories: Private sector (max profit), Public sector (max efficiency with ethical constraints), Emergency service (max efficiency with ethical and temporal constraints)
Impedance – Cost associated with traversing an entire network link – most important factor influencing path finding, allocation and spatial interaction operations. Negative impedance indicates a prohibited direction (ex time, loss of electricity, resistance of flow, one-way streets, etc.)
Stops – Locations visited on a path or a tour (ex customers on a delivery route or cities in a highway system) – where goods, people or resources may be transferred to and from some form of transportation system (ex- bakery orders route to deliver heaviest load first and loads truck to correspond with stops). Stop impedance the time it takes for a stop to occur used to compute the impedance of a path or tour. ex when a school bus drops off/picks up children at homes, the stop impedance may be 2 min at each stop
Turns – Represent relationships between network links (ex turning left is more difficult than turning right) Dependant on the properties of edges – n^{2} possible turns at any node where n is the number of edges connected at a junction (ex 4 edges = 16 possible turns)
Centers – Discrete locations where a resource supply exists. (ex cities, distribution hub, fire stations, hospitals
Mindistance – Minimize total distance from all demand points to centers (P-Median Problem – heuristic to get optimal –not optimum- solution) – objective is to minimize cost or maximize efficiency. Private sector location model – store owner looking for new location, no notion of fairness or ethics, pure profit max
Maxattend – (Attendance maximizing problem) maximize the assignment of demand to each center – likelihood of assignment decreases linearly with distance (most useful for school zoning problems) Public sector location model – efficiency often clouded by fairness
Mindistancepower – (Minimize total powered distance problem) minimize total distance traveled, where distance is subject to a power function – “sometimes doubling the length means quadrupling the impedance.” Public sector location model
Maxcover – (maximal covering location problem) maximize the demand that is covered within a specified time or distance “get the most service to people within 5 minutes” Emergency service location model
Model – A simplified abstraction of reality. An artificial construction in which parts of a source domain are represented in target domain. A set of rules and procedures for representing a phenomenon or predicting an outcome. In geoprocessing, a model consists of one process or a sequence of processes connected together. It is created in a toolbox and built in a ModelBuilder window. A model can be exported to a script file.
Source Domain – Real world
Target Domain – Where the real world is modeled
Grid – Spatial relationships are implicitly stored as part of the structure – regular arrangement of cells (tessellation) – a square grid is arranged in regular rows and columns. Grid assumes all areas within each cell have the same z value. Raster controls for space, measures attribute.
Lattice – Regularly spaced z values arranged in a raster pattern (ex DEM) – lattice mesh points are only spot values, and no assumption is made about the locations in between (therefore interpolation is used to measure values that fall between the lattice mesh)
*Tessellation – form of small squares or blocks
arranged in a checkered or mosaic pattern.
3 most common for GIS: square, triangle, hexagon
DEM – Digital Elevation Models can be represented as a raster lattice or as a TIN. DEMs are commonly built using remote sensing techniques, or from land surveying. Note that the contour line data or any other sampled elevation datasets (by GPS or ground survey) are not DEMs, but may be considered digital terrain models. A DEM implies that elevation is available continuously at each location in the study area.
*Local Function – cell by cell computation of a new value for a
single cell as a function of one or more existing values at that
location in the inputs or as an operand
*Focal Function – Neighborhood operations – Moving window compute each location’s new value as a function of the existing values/distances and/or directions of neighboring (but not necessarily adjacent) locations on a specified layer (focalmax, focalmin, focalmean, focalmedian, etc.) Simple (density) calculations sum of all features that fall within search area divided by search area size. Kernel calculations are similar to simple density calculations, except proximal features to the center are weighted more heavily than those near the edge. Kernel calculations smooth the data more than simple calculations
Global Function- work on all cells within a raster – used for calculating Euclidean or weighted distances for every cell in a raster dataset
*Zonal Function – A zone is the set of cells in a raster that
share the same value – not necessarily contiguous. The connected
set of cells that share the same value is called a region. Additional
information about a location or address, used to narrow a geocoding
search and increase search speed. Address elements and their related
locations such as city, postal code, or country all can act as a zone. The calculation of a statistic for each zone of a zone dataset based on values from another dataset, a value raster. A single output value is computed for each cell in each zone defined by the input zone dataset
Arithmetic Operator – Add/subtract/multiply/divide/modulus (“MOD” – remainder)/ Negation (“-“ unary minus. All or combination can be applied to perform spatial analysis on raster datasets
Relational Operator – Return values of T/F based on less than, greater than, equal, not equal, etc.
Boolean Operator – Return values of T/F based on logic: Logical AND, OR, XOR, NOT (compliment)
Logical Operator – Build logical tests on a cell-by-cell basis – returns binary output (0,1s). Non-zero values are true (1), zero are false (0) and NoData returns NoData. Logical difference (“useful for change detection”), Contained in List (“a way of masking”), Replace (“filling in values when merging things”)
Combinatorial Operator – Combine attributes of multiple rasters – finds all unique combinations of values and assigns a unique ID to each in the output grid (CAND, COR, CXOR) – good for site suitability (overlay soil type, slope, vegetation, land value, output = best place to build)
TIN- Triangulated (3 pts = min, 4 pts = redundant) – depicts z-values as simplest facet of a planar surface. Irregular – nodes can be at any x,y location to represent complex curvilinear features better than a grid (of course, not as well as vector). Network – based on relative relationship between three corner nodes. TIN used to display and analyze terrain and other types of surfaces. When building a TIN, breaklines, mass points and exclusion polygons should be established first, and allow model builder to fill in intermediate (between these predefined features) terrain variations
Node – is the fundamental building block (corner point) connected to nearest neighbor by edges
Edge – like arcs – have L/R topology that identifies adjacent triangles
Hull – Outer boundary of the area represented
*Breakline – define linear features by lining up the triangle sides and nodes
*Mass Point – Important locations where z values come together to represent peaks or valleys – triangles create facets similar to cut diamonds.
*Exclusion Polygon – represent totally flat areas – like lakes, and edges of triangles don’t cross – triangle edges line up outside of polygon
Discretization – Locating intersections where the precision of coordinates is limited: coordinates in an integer or single precision real number domain may not include the actual intersection location – usually you must select closest coordinate or have a rule for ties. Greene-Yao algorithm splits line segments to that intersections must be within a specified distance (similar to buffer) of the original lines)
Delaunay Triangulation – Ideal triangles in TINs are equilateral. This is a proximal method that satisfies the requirement that a circle drawn through the three nodes of a triangle will contain no other point. “You don’t want little sliver triangles as they are a sign of a poorly made TIN – you can’t always avoid them”
*Voronoi Diagram – aka proximity polygons or Thiessen Polygons. Simplest form of interpolation (with nominal or ordinal data this is the only way to interpolate). Polygons are created by an area of points that are closer to its centroid than any other.
Transformation – aka reprojections -4 types: Euclidean (preserve shape/size) “just moving origin or rotating”; Similarity (preserve shape, not size) “scaling up or down”; Affine (preserve parallelism, rotations, reflections shears, possible) “May stretch or shift, but parallel lines remain parallel”; Projective (move from spherical to Cartesian coordinates) “sphere to flat.”
1. 4 levels of abstraction – 1) Real world milieu 2) Data model 3) Data Structure 4) File structure
Control |
Measure |
Type |
Location |
Attribute |
Raster |
Attribute |
Location |
Vector |
Relationships |
Loc/Att |
TIN |
Composite |
Loc/Att |
Choropleth |
2. 3 main components of geographic info – 1) Space – 2D or 3D representation 2) Time – ray or vector 3) Attribute – descriptive info; one must be fixed, one controlled one measured. In a raster, space is controlled by the tessellation and attributes vary over the grid. In vector, attribute is controlled in tables and the spatial relationships vary.
In a TIN, both the locations and attributes vary, but their relationships are controlled by the design of the network.
3. Match data model with data type: Vector – points, lines and areas. Raster – continuous data. TIN – terrain
*4. Describe 3 forms of surface continuity w/ex
1) Piecewise continuous surfaces – each location has unique value and abrupt jumps are possible, ex pop density
2) Once differentiable surfaces – no abrupt changes in value, but abrupt changes in surface are possible, ex young mountain ranges
3) Twice differentiable surfaces – both value and rate of change are continuous, ex mature mountain ranges
4) Describe how to detect and fix topological inconsistencies in geodatabases – Use the validation tool on selected features or an entire dataset. Also, use the Error Propagation Wizard.
5) Be able to format a logical expression to select by attribute: AND = must be in both; OR = can be in either or both; XOR = must be in one or the other, not both; NOT = opposite of result is returned
6) Use a diagram to illustrate Overlay Operations True overlay operations change the schema of a table, the alternative (boundary operations – update (in ‘overlay’ tools in Arc), erase (donut), clip (cookie-cutter), split (1=2), merge and append add datasets to existing targets)– only change the cardinality of the attribute table.
UNION – “OR” Overlay that includes all polygon features from multiple feature classes. Output attribute table has fields from both inputs. Keeps all new polys – even ‘slivers’
IDENTITY – All input features are retained, and all overlapping identity features are added to input layer in output. Schema is enlarged to include new attributes “Give me all of the input layer and only part of the identity layer that intersects the input layer”
INTERSECT – “AND” Can work with multiple point, line or area features. Features that overlap in all inputs are written to the output. “Give me things in this layer and things in that layer but nothing outside both”
SYMMETRICAL DIFFERENCE – “XOR” Only works on pairs of polygon layers. It’s the opposite of intersect only outputs features of the input and update feature layers that do not overlap
UPDATE – not a true overlay – it is a boundary operation because the schema is not changed. “cut and paste” – replaces existing polygons and old records in the attribute table – so cardinality is increased
7) Describe tabular output for boundary operations vs overlays – In boundary ops the cardinality changes, in overlays both schema and cardinality change
8) Use Simpson’s Rule to find the area of polygon
Area(A) = S [(x _{i+1} – x _{i})(y _{i+1} + y _{i}) / 2]
(5-2)(8+6)/2 = 21
(7-5)(7+8)/2 = 15
(6-7)(6+7)/2 = -6.5
(4-6)(6+6)/2 = -12
(3-4)(4+6)/2 = -5
(2-3)(6+4)/2 = -5
A = 7.5
9) List and describe 3 simple quantitative shape descriptors compute these indices for a sample poly
These ratios don’t uniquely define the shape, they just give you parameters that represent it related to well known shapes. Compactness ratio – gives radius if it were a circle; Elongation ratio – divide the longest length by the shortest length gives an idea of shape ‘range’; Form ratio – longest length squared and divided by the area gives you an idea of how ‘far away’ the shape is from a square
10) Distinguish between weakly connected and strongly connected sets from weakest to strongest(?) Binary relation, reflexive relation, symmetric relation, transitive relation, equivalence relation
11) Use Euler’s formula to compute a missing parameter from a given polyhedron (Face, Edge or Vertices)
12) Use a venn diagram to illustrate overlay operations
UNION INTERSECT IDENTITY SYMMETRIC DIFFERENCE
13) Discuss how line feature representations are more complex than simple point objects
14) Describe the differences between a geometric and logical network in the context of geodatabases
Geometric networks are the spatial relationships displayed and manipulated in Arc – set of features that participate in a linear system (connected system of edges and junctions) stored in a feature dataset with common coordinate system. Logical Networks are a pure network graph (schematic) consisting of edge and junction elements that stores the connectivity information (what lines connect to others and how) not the coordinates – Encapsulated from the user (user interacts with geometric network)
15) Differentiate between simple and complex edge and junction features Simple edge features are associated with a single edge in a logical network; complex edge features are associated with any number of edges
Simple junction features work as one-to-one location; complex junction features are associated with a collection of junctions and edges and therefore requires a custom feature type to handle discrepancies of rule orders
16) Distinguish between the transportation and utility contexts for networks Transportation – direction of flow along edges determined by ‘driver’ therefore can have bi-directional flow; Utility – Direction of flow along edges (links) determined by the configuration of the sources, sinks and switches in the network
17) List and describe 7 major categories of networking tools
Tracing, Path, Tour, Allocation, Location-Allocation, (6, 7?)
18) Compare and contrast Path and Tour, and why their difference is analogous to the difference between the Allocation/Location-Allocation tools – a PATH tool solves 1 problem – finding the shortest minimum impedance through a network where stops are visited in a predetermined order and a TOUR solves 2 problems – both the order in which to visit the stops and the shortest path between the stops is needed to solve. Similarly, ALLOCATION solves 1 problem by assigning portions of the network to a resource supply location (Fire station). Locations and networks are given – as we travel we use the ‘supply’ of time available to get to a location. And LOCATION-ALLOCATION solves 2 problems: where things go and what gets assigned to them (ex where to locate new fire stations); Goal is to locate facilities designed to supply a good or service for which there is a demand in the most efficient manner possible – these models optimize efficiency by simultaneously determining the configuration of the facilities and assigning the demand to the facilities.
19) For a city street describe how to compute a travel time impedance for an edge and how to make a particular edge one-way Impedance weights can be assigned based on amount of traffic the street (edge) gets at particular times in the day, impedance weights can be assigned based on speed limits, impedance weights can be assigned based on the presence of bike lanes and other considerations for multi-modal forms of transportation (which cause automobile traffic to slow down). Negative impedance indicates a prohibited direction so we would use this function to denote a one-way street
20) List and describe 3 main categories of location-allocation models in Arc/INFO Private sector (max profit), mindistance/P=Median Problem; Public sector (max efficiency with ethical constraints),maxattend, mindistance power, mindistance constrained; Emergency service (max efficiency with ethical and temporal constraints) maxcover, maxcover constrained
21) Given a diagram of a linear network, indicate where flow might be indeterminate – Flow may be indeterminate if the network configuration does not provide enough information to establish the direction of flow for an edge. Disabled features do not participate in the network flow (a closed street or a switched off circuit). Uninitiated flow to features is the case when a disabled junction feature cuts off flow to downstream edges. Barriers can be used during network configurations to temporarily disable a network element
22) Briefly describe four types of transformations performed on spatial data aka reprojections -4 types: Euclidean (preserve shape/size) “just moving origin or rotating”; Similarity (preserve shape, not size) “scaling up or down”; Affine (preserve parallelism, rotations, reflections shears, possible) “May stretch or shift, but parallel lines remain parallel”; Projective (move from spherical to Cartesian coordinates) “sphere to flat.”
23) Describe how continuous phenomenon like topography can be represented in a vector model, and why this makes slope and aspect difficult to calculate – Creating contours or other isolines in vector would produce a representation similar to a wedding cake, where each line is threaded through all elevation points. This is ok for human perusal, but interpolation of slope and aspect between these lines involves calculus (taking the first derivative, setting it to zero and then rotating it 90 degrees). This is a lot of work with much room for error that could easily be avoided by selecting a more appropriate model to represent the phenomenon.
24) List 4 additions to Steven’s list of measurement levels (nominal, ordinal, interval, ratio) Nominal – definitions of categories; (1) Graded Membership – fuzzy classes or definition of categories plus degree of membership or distance from prototype; Ordinal – used to run non-parametric tests – definitions of categories plus ordering; Interval – unit of measure plus zero point, arbitrary zero means 40 isn’t necessarily double 20 – like in temperature (Fahrenheit not Calvin); Extensive ratio – unit of measure with additive rule applied (money, distances); (2) Cyclic ratio – unit of measure plus length of cycle (like hours, days, years, etc); (3) Derived ratio – unit of measure with weighting (like relative humidity, as temp increases so does the amount of water vapor the air can hold); (4) Counts – definition of objects counted; (5) – Absolute – type (probability, proportions, etc.)
25) Differentiate between simple and kernel focal functions Simple Density Functions sum of all features that fall within search area divided by search area size. Kernel Functions are similar to simple density calculations, except proximal features to the center are weighted more heavily than those near the edge. Kernel calculations smooth the data more than simple calculations
26) Differentiate between a regular and irregular tessellation – this is in reference to the shape that is being tessellated – regular shapes are consistently repeated, irregular tessellations vary the shape as they’re repeated
27) Sketch 3 types of regular tessellation
SQUARE TRIANGLE HEXAGON triangle nested in hexagon
27) Distinguish between global, local and zonal map algebra operations Local Functions are cell by cell computation of a new value for a single cell as a function of one or more existing values at that location in the inputs or as an operand; Focal Functions, or neighborhood operations, use a moving window compute each location’s new value as a function of the existing values/distances and/or directions of neighboring (but not necessarily adjacent) locations on a specified layer (focalmax, focalmin, focalmean, focalmedian, etc.) Simple (density) calculations sum of all features that fall within search area divided by search area size. Kernel calculations are similar to simple density calculations, except proximal features to the center are weighted more heavily than those near the edge. Kernel calculations smooth the data more than simple calculations; Global Functions work on all cells within a raster and are used for calculating Euclidean or weighted distances for every cell in a raster dataset; Zonal Functions are zones designated by an input zone dataset that establishes a set of cells in a raster that share the same value – not necessarily contiguous. The connected set of cells that share the same value is called a region. Additional information about a location or address, used to narrow a geocoding search and increase search speed. Address elements and their related locations such as city, postal code, or country all can act as a zone. The calculation of a statistic for each zone of a zone dataset based on values from another dataset, a value raster. A single output value is computed for each cell in each zone defined by the input zone dataset
28) Describe the differences between grids and lattices –For both, spatial relationships are implicitly stored as part of the structure – regular arrangement of cells (tessellation). A square grid is arranged in regular rows and columns. Raster controls for space, measures attribute. They are different in that a Grid assumes all areas within each cell have the same z value and a lattice mesh points are only spot values, and no assumption is made about the locations in between (therefore interpolation is used to measure values that fall between the lattice mesh) In the example of a DEM, regularly spaced z values arranged in a raster pattern.
29) How does one register a grid to real world coordinates Because of the spatial relationships implicitly stored within the grid model, one can create a header, in which the x,y coordinates that correspond to the (usually upper/left-most – although some software starts in the lower/left) origin cell (0,0) in the grid, along with the resolution (or real world correspondence to the cell size). The rest of the real world coordinates in the file can then be placed based on the origin and the given resolution of each cell (usually anywhere from 30 to 1 meter resolution). In the event that the coordinates are not oriented N/S, one must also specify the direction of the y-axis.
30) What type of grid has a value attribute table, and which does not Integer grids (no decimals) have a VAT that lists all of the values in the grid and how many cells have that value – you can add items to the vat (ex 20 = residential). Real number grids have no table since each grid cell value is theoretically unique. The VAT is a quick way to determine the overall area of a category because the cell resolution is known and this can be combined with the VAT count.
31) List the major components of TINs and explain how they can be used to represent terrain When building a TIN, breaklines, mass points and exclusion polygons should be established first, and allow model builder to fill in intermediate (between these predefined features) terrain variations. Breaklines define linear features by lining up the triangle sides and nodes and represent rivers and coastlines well. Mass Points are placed at important locations where z values come together to represent peaks or valleys. Exclusion Polygon represent totally flat areas – like lakes – because triangles don’t cross – triangle edges line up outside of polygon
32) Use the Semi-Line algorithm to determine a point as inside or outside a polygon
Odd number of borders crossed are inside, and even number are outside
33) Apply the Peuker algorithm to a curved line for simplification