1 / 28

Geography 463 GIS Workshop

Geography 463 GIS Workshop. April 10, 2006. Outlines. Project management principles Issues in system requirements Consider issues intrinsic to spatial data Choose a logical data model UML Class Diagram. Part I. Project management principles. What is project? What is not?.

ivrit
Download Presentation

Geography 463 GIS Workshop

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Geography 463GIS Workshop April 10, 2006

  2. Outlines • Project management principles • Issues in system requirements • Consider issues intrinsic to spatial data • Choose a logical data model • UML Class Diagram

  3. Part I. Project management principles

  4. What is project? What is not?

  5. The Triple Constraint • Every project is constrained by its: • Scope: what’s included and what’s not • Schedule: anticipated time to complete goals • Cost: budget allocated for project completion • Project manager’s primary duty is to balance these competing constraints, while satisfying intended business need

  6. The Triple Constraint of Project Management System requirements report helps you define the scope, cost, and timing constraints in advance. The image is retrieved from Tom Nolan’s lecture note

  7. Part II. Issues in system requirements

  8. System requirements and project management • Can be thought of as implementation plan that allows the triple constraints to be recognized and balanced given information products identified from needs assessment report • System requirements report will be composed of • Data requirement • Software requirement • Hardware requirement • Staffing requirement • Timing requirement

  9. Master Input Data List How is MIDL useful? Help you recognize requirements for data, software, hardware, staffing, and timing. Data: accuracy, data ready? Options for putting data into system (Buy? Outsourcing? In-house data creation?) Software: which functions are needed to put data into system? Hardware: decision on which workstation purchase is determined by data volume and computing complexity …. From p. 88 in Tomlinson (2005)

  10. Data availability vs. data readiness • Data availability: the date on which you can receive data from its source • Data readiness: the date on which the data is in your system, processed, and ready for use in the creation of an information product • Identify functions needed to put data available into the system (data readiness) • Examples: to be ready for the system • Satellite image requires file transfer and ID creation • Paper map requires scanning, digitizing, topology editing, attribute editing, and so on • Digital data would require a very little work

  11. Priorities • Help you manage the scope of projects • What are important and what are less important information products and data? • Occurs due to time/cost constraints • This decision will be affected by the organization’s strategic plan/goals, thus consultants are not usually invited for this process

  12. Before you estimate timing requirements, consider • Data input timing • Various options for putting data into system will affect timing • Information product application programming timing • Any scripting effort, what are steps involved? Required work hours? • System acquisition timing • Some hardware should be purchased/waited • Training and staff issues • Any training need? • Make the use of existing staff or hire new persons

  13. Part III. Consider issues intrinsic to spatial data

  14. Choosing the appropriate map projection • Intended uses of information products • Be aware of distortion of geometric properties (e.g. area, direction, distance) due to map projections • Geographic area of interest • Location of standard line/parallels • Difference in datum? • e.g. NAD27 vs. NAD83 • Difference in measurement units? • e.g. SPC vs. UTM

  15. Choosing the appropriate map scale • The appropriate map scale can be chosen based on resolution and error tolerance • Resolution: the minimum amount of detail that can be mapped at a given scale; related to data requirements for information products • Error tolerance: how much do your clients tolerate error? related to cost • See p. 115 for the matrix for the choice of map scale

  16. Topology editing • Arc/info coverage • Stored in feature attribute table • Use BUILD/CLEAN command • Shapefiles • Non-topological model • Geodatabase • Topological rules are assigned to feature data set • Topological rules within/between feature class can be defined • Use topological editing tool; help you identify potential topological errors given predefined rules

  17. Part IV. Choose a logical data model

  18. Relational • A collection of table (simple, ad-hoc) • Table (entity) has attributes • The relation between tables is created through a common attribute (remember how you perform table join?) • Logical linkage: which keys needed? • Most common logical model in GIS as well as hybrid form (file for features + relational table for attribute): e.g. arc/info coverage, shapefile

  19. Object-oriented • A collection of object (complex, natural) • Object has attributes and operations • Class is a collection of similar objects • Database can be built using class hierarchy (e.g. some attributes of city database can inherit from supper class such as spatial objects) • Model complex relationships • Association: any • Aggregation: weak form of whole-part relation • Composition: strong form of whole-part relation • Generalization: (e.g. secretary and employee)

  20. Object-relational • Compromise between relational model and OO model • Extend relational data model so that good thing about OO (e.g. complexity, integrity) can be incorporated • Maintain table with abstract data type • Can use SQL • E.g. Geodatabase

  21. Choosing a logical data model • Given continuous popularity of relational data model, relational data model can support the information products that involves a simple conceptual design • If database requires handling of complex relationship, integrity and versioning, OO or OR will work better (e.g. facilities management) • See p. 145 for which logical data model is suggested given data modeling situation

  22. Part V. UML Class Diagram

  23. What is UML? • Acronym for Unified Modeling Language • Standardized efforts that combine many years’ work on object-oriented analysis and design • Use graphic notations (e.g. diagram) • Intuitive; somewhere between natural language and programming language • Anything to do with project management? Yes, user requirements modeling, planning, system design, database design, and so on whose tasks can be chosen from different kinds of diagram

  24. Class Diagrams • Associations • Physical location • (e.g., is next to) • Direct actions • (e.g., drives) • Communication • (e.g., talks to) • Ownership • (e.g., has) • Satisfaction of a • condition • (e.g., works for)

  25. Class Diagrams Association class Aggregation

  26. Class Diagrams Generalization / Inheritance

  27. Class Diagrams aggregation association generalization

  28. UML Resources • Software to use in the department • Microsoft Visio or other drawing tools • ESRI’s “Building Geodatabases with Visio” • IBM Rational Rose literature on design w/ UML • Introduction to UML • Entity-relationship modeling with UML • UML tutorials on the web

More Related