Geography 280b
Lecture #3
Thursday, January 27, 2000
Basic Operations I
Chapter 1 in the book.....
What does the “event” part of Dueker’s definition of a GIS allude to?
What are the Laboratory Assignments for this chapter?
Reading Corner
For next week: Chapter 2: “GIS’s Roots in Cartography”
Announcements
Ombuspersons
Lecture notes website: http://publish.uwo.ca/~pazner/
What the dickens is the “OLH” Folder?
The Structure of Today's Lecture:
- Explicit Renumbering (Recode)
- Automatic Renumbering (Slice)
- Break
- GIS 2000 Brief
- Adjacency Assessment (Clump)
- Basic Measurement (Spread)
Basic Operations I
Reference: Geographic Information Systems and Cartographic Modeling, C. Dana Tomlin, Prentice Hall, NJ, 1990.
Explicit Renumbering
- Renumber by user-specified values.
- Renumber by explicit value assignment.
- (A look-up table (LUT) method)
Useful for: relabeling, reclassification, isolating, grouping
Explicit Renumbering in MF Works is mostly done using: Recode
- although a short-cut exists in some cases
- in the form of the Change Value(s) dialog box
The common Recode and its modifiers
- NewLayer = Recode map Assigning new value To existing value range list [CarryOver] [BuildText]
A word or two about the Syntax of the operation statement
Explanation of the operation and its modifiers.
- Note that values can be specified as a range and/or as a list (incl. a list of ranges, etc.).
- CarryOver allows us to partially Renumber by user-specified values while retaining the rest of the information.
What does BuildText do? (please check in the OLH...)
An Example. The use of Recode to isolate and group:
- /* Isolating and Grouping bodies of water */
- ponds = Recode HydrologySub Assigning 1 To 2013 … 2015 ;
Additional Example. The use of Recode for reclassification:
- /* distinguishing between A particular body of water versus the rest */
- WhichPond = Recode "clumped ponds" Assigning 1 To 1 Assigning 2 To 0 , 2 Assigning 0 To VOID ;
Automatic Renumbering
Interval renumbering—Slice
An additional shortcut to Recode can be used when the desired grouping is regular or substantial.
The renumbering is performed by a “(data) density slicing” operation.
- This operation is Slice in MFworks.
- The data or legend categories are reduced systematically according to user specified parameters using the modifiers below.
Slice and its modifiers
- Slice map Into value By increment value Range value range UseTopValue UseBottomValue UseMiddleValue MakeRangeLabels
- Explanation of the operation and its modifiers.
An Example. The use of Slice:
- /* grouping distances automatically */
- ThisFar = Slice "pre ThisFar" Range 1… By 100 ;
Break :-)
GIS 2000 Brief
- The detailed program is now up on the web
- http://www.gis2000.com/
- One day: Wednesday, March 15, 2000
- Lets take the train: VIA Rail (Use this link to book your tickets)
- Note: You can save money if you buy an ISIC card at TravelCuts in the UCC
- Total trip cost around $75
Adjacency Assessment
Assessing adjacent same-zone cells — Clump
What it does:
- Assesses adjacent and similar cells. Clump evaluates and flags the spatial adjacency of similar (i.e. belonging to the same zone) cells.
The operation and its modifiers.
Syntax is exceedingly simple:
- Clumps = Clump map At distance ;
Explanation of the operation and its modifiers.
- This operation uniquely identifies groups of adjacent cells. There are 2 criteria; the cells must have the same value and be within a specified distance of each other. Each group is assigned a new value in the output map. VOID cells are not ignored by the clump operation.
- For each cell, the operation asks: Is this cell within range of another, previously evaluated same-value cell? If yes, it gets the same label on output as that other cell did. Otherwise the result is a new value. The values act as a label, or clump-ID or tag, etc.
The distance is specified as number of cell side lengths (1 and 1.5 are very common).
- Use 1.5 if you want to include diagonal cells
***** Bring a calculator to the midterm that can do square roots. *****
The order by which Clump does its work.
- From left to right, and top to bottom.
There is a limit to the number of generated clumps.
Questions:
- What do you think it is?
- Is it enough?
A simple case involving only 1-zone.
An example with n-zones.
Recode can be used to control a Clump.
- A Recode can be used to reclassify the data into a desired number of categories for Clump to operate on. Recode is often used to collapse and isolate the data prior to clumping. The paired Recode-Clump occurrence is extremely common within models. Example: tagging houses in a scanned and classified air-photo with a unique building ID.
Clump assesses the property of the spatial adjacency of objects to one another
- Clump touches on the most basic spatial relationship—Distance, and focuses on nearness or closeness—one of the most important spatial relations.
In more general terms, adjacency indicates spatial ‘belonging’; membership in a group, set, cluster, or clump
- A fundamental logic relationship of being part vs. apart(isolated, insular), in our case, incorporating the spatial dimension.
There are many aspects of adjacency or proximity:
Clustering (into “clumps”)
- Things in nature tend to cluster. Given a salt-and-pepper sprinkle of points, Clump checks which occurrences form and belong to a given cluster.
Containment and inclusion
- Within an imaginary boundary and area determined by the threshold criterion. Clump can be used in a model to make these boundaries more explicit.
Insularity
- Things in nature tend to separate. The focus of the operation may be to maximize separability. Eg, in a scanned air-photo of a thundering herd of caribou.
Natural continuity
- Are there gaps in the road lines due to input/editing errors?
And discontinuity
- Examining sub-network structure: Which water-ways and water-bodies in a hydrologic network are not connected on the surface?
Finally, believe it or not:
- Some MAJOR GISs don’t support a Clump operation and/or many instructors don’t teach it.
- A black hole in GIS functionality and use.
Basic Measurement
- Measuring single straight distances.