School of Population Health


SPAN - instruction manual

A Manual for SPAN


© Roger J. Marshall

Section of Epidemiology and Biostatistics, School of Population Health, The University of Auckland, New Zealand

Table of contents


1 Introduction
2 Installation and requirements
  2.1 SPAN on the Internet
  2.2 Version
3 SPAN methodology
  3.1 Partition representation
  3.2 Attributes
  3.3 SPAN Boolean  partitions
  3.4 Regular  partitions: positive attibutes
4 Windows SPAN       
  4.1 Mouse input in windows
  4.2 Extracting contents of windows
  4.3 I'm thinking...
  4.4 The View function
5 Data preparation
  5.1 Free format
  5.2 Fixed format
  5.3 Excel spreadsheet input data
6 Main menu items
7 File and Edit
  7.1 File
  7.2 Edit .SPN and Edit file
  7.3 Edit:Split sample
  7.4 Edit:Select ...
8 Control file: Data input            
  8.1 Control file
  8.2 Variables
  8.3 Format
  8.4 Grouped data
  8.5 Missing values and missing attributes
  8.6 Data restrictions
  8.7 Character Data
  8.8 Boolean combination lines
9 Control file: Creating attributes
  9.1 Attribute representation
  9.2 Interval and ordinal variables
    9.2.1   Range attributes
    9.2.2   Percentile and mean cutpoints
    9.2.3   Multiple cuts
  9.3 Nominal or binary variables
    9.3.1   Category representation
  9.4 Labels
  9.5 Direction of association
  9.6 Multiple attributes
  9.7 Using SPAN to edit or create a control file
  9.8 Missing value attributes
    9.8.1   Nominal variables
    9.8.2   Interval variables
  9.9 Special attributes
    9.9.1   The null attribute
    9.9.2   Index attribute
    9.9.3   Random attribute
    9.9.4   all_ attribute
    9.9.5   The cross-validation specification
  9.10  Continuation indicator
10 Y menu
  10.1 OK
  10.2 Distribution
  10.3 Scatterplot /cross-tab
    10.3.1   Z: Adding a third variable. Rotating 3D plots
    10.4.1   Creating attributes from binary transformations
  10.5 Select Y
  10.6 By-group/cross-validate
11 Criteria          
  11.1 Criterion
  11.2 Balancing parameter gamma
  11.3 Penalty parameter beta
  11.4 Complexity
  11.5 Priors for Entropy/Gini indices
  11.6 Costs for Entropy/Gini and Quality indices
12 Rank       
  12.1 Attributes Rank Plot
  12.2 Fix positive attributes
    12.2.1   Multiple Y 
  12.3 Effectiveness v Cuts plot
  12.4 ROC curves
  12.5 QROC curves
13 Strategy            
  13.1 Constrain X space
  13.2 Force in attribute
  13.3 Iterative mode
  13.4 Tree
  13.5 Top m
  13.6 Decide positivity
  13.7 Set Extent parameters
14 Extent             
  14.1 Size p_1 ....p_q
  14.2 Up to stated size
  14.3 A  and A-
  14.4 Number of float levels
15 Search              
  15.1 The Search 
  15.2 During the search
  15.3 The end of a search: complexity hull
  15.4 Penalty parameter setting
  15.5 Searching in iterative mode
  15.6 Search interrupts
  15.7 Search with by-group/ cross-validate  
  15.8 Search in tree mode
16 Process                     
  16.1 Rectangle Diagrams
    16.1.1   Rectangles construction method
  16.2 Lists
  16.3 Distribution
  16.4 Statistics
    16.4.1   2x2 table statistics
16.4.2   Chi-square, generalised error etc
16.4.3   Risk matrix
16.4.4   Log-rank: incidence rates
16.4.5   Table of prior adjusted pseudo-counts
  16.5 Random
  16.6 Tree
  16.7 Test [training] sample
  16.8 Create added attribute
  16.9 Manual partitions [simplifying Boolean expressions]
17 Options              
  17.1 Full Boolean reduction
  17.2 Attributes are primitive
  17.3 Tied partitions: use last found
  17.4 Missing attributes
  17.5 Detailed output
  17.6 Display graphics while searching
18 View and Window   
  18.1 Output log
  18.2 Viewing data and .SPN files
  18.3 Added attributes
  18.4 Window
19 Help
  19.1 Website
  19.2 Topics
  19.3 Enter Registration #
  19.4 Manual
  19.5 FAQs
20 References