Characterizing Layers

1
Characterizing Layers

• Displaying, Viewing, Describing, Communicating

OR What have I got here?
Plus a brief review
2
The RASTER of raster systems

• A raster is a
• table like array of values
• a grid of cells or pixels
• This is a 6x5 (C,R) ARRAY
• Cells can have different attributes

Symbolization of the attribute values
3
The Raster

• The Geographic position is determined by the
  location of the cell in the raster
• Row 1 Column 5 (R1C5 or, sometimes, C5R1)
• The attribute data is the single numeric value
  stored in the cell
• The raster image has an extension of .rst
• The metadata is in a document file associated
  with the image file and has an extension of .rdc
• BOTH files have to be present. If no .rdc file
  then IDRISI cannot even find the image

4
Navigation in the grid

• To get around in a grid or raster you have to
  know
• Column and row address AND/OR
• X and Y distances
• How do you know X and Y?
• Because X and Y max and min AND cell size are
  part of the data set!
• X min (cell size cells) X (roughly)

5
Addressing, Column, Row
6x5 Grid
C0
C1
C2
C4
C3
C4
R0
R1
R2
R3
R4
6
Addressing, R,C (C,R)
A 6x5 grid (C,R format)
C0
C1
C2
C4
C3
C5
R0
0,0
R1
R2
R3
5,4
R4
7
Addressing, X,Y
C0
C1
C2
C4
C3
C4
R0
R1
R2
R3
R4
8
Coordinate Systems in IDRISI
Since Row and Column counts start at 0 The NUMBER
OF rows or columns in the image is always 1 more
than the maximum values. OR A r5,c6 image has max
row 4 and max column 5!
Booby Trap!
9
IDRISIs
Data Type File Type
10
Data Types

• Data Type according to IDRISI is the type of
  data which IDRISI will find in the file - what
  IDRISI sees
• Integers are whole numbers within the range -
  32768 to 32767.
• Real numbers have a fractional part, or are whole
  numbers outside the integer range.
• Byte values are positive integer numbers ranging
  from 0 to 255.

11
File Type

• File type is the kind of file format used by
  the computer what the computer system sees
• Image files may be stored as ASCII, Binary, or
  Packed Binary,
• Vector files as Binary,
• Attribute Values files as ASCII or ACCESS
  (database)
• DO NOT GET FILE TYPE AND DATA TYPE CONFUSED!

12
Attribute Data Categories

• Data types
• Nominal
• Ordinal
• Interval
• Ratio
• When we do analysis -- Boolean (1,0) later....

13
Metadata

• The who, what, where, when, how and why of data
  set creation
• The document file - .rdc or .vdc
• Very, very important.
• First it contains the information that describes
  the image.
• Second, it keeps references for others to use.
• Third, it tells people how you created/changed
  data
• IT IS METADATA!

14
Creating New Images?

• You should think about --
• Data Type
• Nominal, ordinal, ratio, interval
• Classification system -- Legend
• What do you want to do with the data?
• Data Entry -- Coding Rules
• Be Consistent
• Why you chose that method

15
Metadata

• The who, what, where, when, how and why of data
  set creation
• The document file - .rdc or .vdc
• Very, very important.
• First it contains the information that describes
  the image.
• Second, it keeps references for others to use.
• Third, it tells people how you created/changed
  data

16
Coding rules
?

• Point
• Majority
• Average
• Prominence
• Connective

17
OK, Symbolization

• Or
• How is data seen
• By
• Us

18
Layer Symbolization

• How we explain and illustrate...
• What it is
• Where it is
• What is important and significant about it
• To others and to ourselves

19
Two Methods

• Graphic (Viewing on screen or on paper)
• Charts, Maps, Displays, Symbols, etc
• Visual! -- intuitively makes sense when you look
  at it
• You can see the spatial relationships
• Descriptive
• Statistics, Narratives

20
Combinations

• Graphic and Descriptive
• Tables, Maps, and Charts with descriptions and
  legends
• Both should work together to effectively
  communicate the importance and relevance of your
  information
• In this course Memos with Tables, Maps, and
  Histograms the spatial models from your logbooks

21
The Spatial Model

• A graphic of how the process was carried out

• Check the Syllabus for complete description

22
Why?

• The spatial model is a record of how you carried
  out the required process
• You include it in your reports so I can figure
  out what you did so you can fix it if incorrect.
• When we get to part where you use the modeling
  software it will provide your spatial model
• In real life it is the basis for your defense of
  what you did in a hearing!

23
Symbolization

• Using color, shapes, style, and size to
  represent or symbolize the data
• Initially we will concentrate on the use of color
  to represent numeric values
• The relationship between value and color in a
  given pallet can have a large effect on the
  communication of the data to you and any other
  viewer.

24
Color Palettes in IDRISI

• What is a palette
• An index that relates the presented color of a
  pixel to an attribute value

0 1 2 3 4 5 -------
25
Color Palettes in IDRISI

• What is a palette
• An index that relates the presented color of a
  pixel to an attribute values

• Range of colors in palettes (how many colors)
• 2, 16, 256 or any number lt 256
• The human eye can only discern about 16 different
  colors when they are not close in location

26
16 Color problem
27
Assignment of Symbols

• Last images had 1 symbol per value in the image
• Because I made a palette for 0-20 values
• How does the system assign colors if the number
  of values does not match the number of colors in
  the palette???
• Not trivial since this strongly influences the
  perception of what the image is telling you!

28
Assignment of Symbols

• There are a number of ways in which systems apply
  a palette to an image
• Wrapping
• Truncating
• Stretching
• Reclassing

29
Assignment of Symbols

• Wrapping
• If you have 4 symbols and more than 4 values the
  symbols repeat
• Color range of 4 symbols with 8 attribute values
• Blue, Green, Red, Yellow, Blue, Green, Red,
  Yellow, Blue, etc...
• 0,1, 2, 3, 4, 5, 6, 7, 8, etc
• IDRISI DOES NOT WRAP!
• (actually I had to fool IDRISI to create the
  following image)

30
Wrapping
The software provides symbols for values 1-4 and
then just cycles through the values for values gt 4
31
Assignment of Symbols

• Truncating
• The attribute values gt palette max are
• Integer data type- given the value of pallet max
  so that you cannot discern them
• Real data type- image is autoscaled like it or
  not
• This is create a problem when trying to
  understand the image

32
Using IDRISI16 display which has symbols for
values 0 through 15 Any values gt 15 are given
the symbol for 15, In this case dark green.
33
Assignment of Symbols

• Stretching
• The color range is stretched over the attribute
  values
• Color range of 4 symbols with 8 attribute value
• Blue, Red, Yellow, Green
• 1-2 Blue, 3-4 Red, 5-6 Yellow, 7-8 Green
• Autoscaling stretches but stretches between the
  minimum value and the maximum value in the image
• Autoscaling does NOT change the data values in
  the image being viewed!

34
Stretching
Here the IDRISI16 set if symbols has been
stretched over the values 1 to 20 with the
result that some symbols have been duplicated.
There is no longer a 1 to 1 relationshipbetween
symbol and value
35
Autoscaling
Autoscaled
Stretched
36
Assignment of Symbols

• Reclassing
• Textbook explains that stretching is a form of
  reclassing
• We could reclass the image so that R0C0-R0C1 were
  given values of 0 and R5C2-R5C3 were given values
  of 15 and the intermediate cell other values
• In this chapter youve read about reclassing and
  assigning. We will come back to this soon.

37
Reclassification
38
Reclassification
39
Arrangement of Colors

• Qualitative palettes have sharp contrast between
  each symbol in the array
• Black to Yellow to Blue to Orange
• Great for Nominal and Ordinal Data
• Discrete Values
• Quantitative palettes have ramped colors
• Dark to Light
• Great for Ratio and Interval data
• Continuous Values

40
Quant and Qual
41
Quan256 palette applied to continuous, ratio data
42
Qual256 applied continuous, ratio data is ugly!!!
43
Qual256 palette applied to discrete, nominal data
44
Qual256 palette applied to discrete, nominal data
very hard to read!
45
In IDRISI

• If you are having trouble understanding the
  display of your data check ..
• 1. Palette selection
• Range of palette-16 vs. 256 colors re range of
  attribute values
• Quantitative vs. Qualitative
• 2. Autoscale option (on or off)
• Actual numbers in image using Data Structure
  display

46
Orthographic View
Soils draped on elevation
BUT Expanded images by 5 so are 320x320
otherwise you could not see the pattern
47
(No Transcript)
48
(No Transcript)
49
Starting Projects - Exercise 3

• You have completed your grids after coding and
  assigning values
• Bring Exercise 2 to lab with youalso your Zip
  disk, a 3.5 floppy disk and your log book.

50
Starting Projects - Exercise 3

• The first thing you usually have to do with most
  any project is to establish folders with MS
  Windows Explorer -- One working and several
  resource folders

51
Data Capture - Exercise 3
Enter your data into spreadsheet
52
Starting IDRISI Projects - Exercise 3

• Setting the Project Environment

53
Then What?

• Convert to an IDRISI raster file (.rst) and
• Complete the document file (.rdc)

54
SSTIDRIS
55
Complete Metadata
56
What else?

• Decode your coded elevation using math modules
• Try a map composition
• Display features!
• Help System
• This one is a big onea drink with a fire hose
• Wednesday -- Elevation! - Quiz 3