Working with NetDraw to Visualize Graphs

Slides:



Advertisements
Similar presentations
KompoZer. This is what KompoZer will look like with a blank document open. As you can see, there are a lot of icons for beginning users. But don't be.
Advertisements

Social network partition Presenter: Xiaofei Cao Partick Berg.
Clustering.
Network Matrix and Graph. Network Size Network size – a number of actors (nodes) in a network, usually denoted as k or n Size is critical for the structure.
Clustering Categorical Data The Case of Quran Verses
PARTITIONAL CLUSTERING
Label Placement and graph drawing Imo Lieberwerth.
Image Indexing and Retrieval using Moment Invariants Imran Ahmad School of Computer Science University of Windsor – Canada.
Feb 20, Definition of subgroups Definition of sub-groups: “Cohesive subgroups are subsets of actors among whom there are relatively strong, direct,
The Basics of Network Computing Michael T. Heaney University of Michigan August 31, Hour lesson This material is distributed under an Attribution‐NonCommercial‐ShareAlike.
Annotation & Nomenclature By Corey Fortezzo for PG&G GIS Workshop, 2010.
Using Surfcam to Produce a Numeric Control (NC) Program An Introduction to the CAD/CAM Process Instructions for 3 Axis Programming Using the D&M CNC Milling.
Chapter 6: Pivot Tables Spreadsheet-Based Decision Support Systems Prof. Name Position (123) University Name.
Section 2.3 Gauss-Jordan Method for General Systems of Equations
HCI 530 : Seminar (HCI) Damian Schofield. HCI 530: Seminar (HCI) Transforms –Two Dimensional –Three Dimensional The Graphics Pipeline.
Communities in Heterogeneous Networks Chapter 4 1 Chapter 4, Community Detection and Mining in Social Media. Lei Tang and Huan Liu, Morgan & Claypool,
Describing the Relation Between Two Variables
Social Networks Visualization Who’s the popular kid?
Centrality and Prestige HCC Spring 2005 Wednesday, April 13, 2005 Aliseya Wright.
Social Position & Social Role Lei Tang 2009/02/13.
1 of 5 This document is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS DOCUMENT. © 2007 Microsoft Corporation.
Dimension reduction : PCA and Clustering Christopher Workman Center for Biological Sequence Analysis DTU.
Google SketchUp Castle
Quark QuarkXPress 4 Intermediate Level Course. Working with Master Pages The Document Layout Palette allows you to add, delete, and move document and.
PowerPoint: Tables Computer Information Technology Section 5-11 Some text and examples used with permission from: Note: We are.
Introduction to Management Science
Microarray Gene Expression Data Analysis A.Venkatesh CBBL Functional Genomics Chapter: 07.
ADVANCED MICROSOFT POWERPOINT Lesson 6 – Creating Tables and Charts
Clustering Unsupervised learning Generating “classes”
Creating a MagicInfo Pro Screen Template
Information theory, fitness and sampling semantics colin johnson / university of kent john woodward / university of stirling.
Social Science Research Design and Statistics, 2/e Alfred P. Rovai, Jason D. Baker, and Michael K. Ponton Factor Analysis PowerPoint Prepared by Alfred.
Manifold learning: Locally Linear Embedding Jieping Ye Department of Computer Science and Engineering Arizona State University
Exploratory Data Analysis. Computing Science, University of Aberdeen2 Introduction Applying data mining (InfoVis as well) techniques requires gaining.
Lesson 7 Guide for Software Design Description (SDD)
© Commonwealth of Australia 2009 Using OneNote in the Classroom.
Tutorial 1 Getting Started with Adobe Dreamweaver CS3
Chapter 14: SEGMENTATION BY CLUSTERING 1. 2 Outline Introduction Human Vision & Gestalt Properties Applications – Background Subtraction – Shot Boundary.
A Clustering Algorithm based on Graph Connectivity Balakrishna Thiagarajan Computer Science and Engineering State University of New York at Buffalo.
Welcome to MM305 Unit 6 Seminar Larry Musolino
MICHAEL T. HEANEY UNIVERSITY OF MICHIGAN AUGUST 31, HOUR LESSON The Basics of Network Computing.
ArcGIS: ArcMap Tables. Agenda Opening tables The interface Working with columns Working with records Making selections Advanced table tools ▫Add fields.
Basic Editing Lesson 2.
A to Z Math Project BY: AUSTIN WAHL. A is for Algebra Tiles  Algebra Tiles are used to represent variables and constants. Also The tiles help you visualize.
Learning How to Make a PowerPoint Presentation A tutorial for novice users.
Susan O’Shea The Mitchell Centre for Social Network Analysis CCSR/Social Statistics, University of Manchester
Chapter 13: AutoCAD DesignCenter. After completing this Chapter, you will be able to use the following: Overview of DesignCenter Container, Content Type.
Neural Networks - Lecture 81 Unsupervised competitive learning Particularities of unsupervised learning Data clustering Neural networks for clustering.
Chapter 11: Block References and Attributes. After completing this Chapter, you will be able to use the following features: Create and insert block references.
1.To begin click: File, New, then select Design 2.Draw a circle, then use the dimension tool to set it’s size to 84 diameter or 42 radius. 3. Next draw.
Clustering.
Graphs & Matrices Todd Cromedy & Bruce Nicometo March 30, 2004.
University of Kentucky – Gatton College of Business LAB 1 – Intro to Ucinet & Netdraw Virginie Kidwell Travis Grosser Doctoral Candidates in Management.
Selected Topics in Data Networking
© Tan,Steinbach, Kumar Introduction to Data Mining 4/18/ Data Mining: Cluster Analysis This lecture node is modified based on Lecture Notes for Chapter.
Block modeling and equivalence Christopher McCarty October 23, 2012.
1 Microarray Clustering. 2 Outline Microarrays Hierarchical Clustering K-Means Clustering Corrupted Cliques Problem CAST Clustering Algorithm.
Example Apply hierarchical clustering with d min to below data where c=3. Nearest neighbor clustering d min d max will form elongated clusters!
Network Partition –Finding modules of the network. Graph Clustering –Partition graphs according to the connectivity. –Nodes within a cluster is highly.
Topical Analysis and Visualization of (Network) Data Using Sci2 Ted Polley Research & Editorial Assistant Cyberinfrastructure for Network Science Center.
Clustering [Idea only, Chapter 10.1, 10.2, 10.4].
Complex Numbers and Equation Solving 1. Simple Equations 2. Compound Equations 3. Systems of Equations 4. Quadratic Equations 5. Determining Quadratic.
GUILLOU Frederic. Outline Introduction Motivations The basic recommendation system First phase : semantic similarities Second phase : communities Application.
Unified Modeling Language
PowerPoint: Tables and Charts
Session 1b Decision Models -- Prof. Juran.
Text Categorization Berlin Chen 2003 Reference:
Clustering.
Presentation transcript:

Working with NetDraw to Visualize Graphs Chapter 4 Working with NetDraw to Visualize Graphs Hanneman, R. A., & Riddle, M. (2005). Introduction to social network methods. (Some of the data are from the reference materials.) Presented by Minzhe XU 1301110888

EMPHASIS How to draw graphs based on social network data using NetDraw How to visualize node attributes, relation properties, change position, and highlight certain parts How to draw preliminary propositions from the graph after visualization What is a good drawing of a graph

CATALOG 1. Introduction 5. Location Note that some names of the original sections have been changed. 1. Introduction 5. Location 2. Data Input 6. Highlighting Parts of the Network 3. Node Attributes 7. Output 4. Relation Properties 8. Summary

INTRODUCTION What is the contribution of a good drawing of a graph? Suggest some important features of overall network structure Help in understanding how a particular node is embedded

DATA INPUT Import data from UCINET or Pajek: “File> Open” Create a random network and revise: “File> Random”; “Transform> Link Editor”; “Transform> Node Attribute Editor” Use an external editor to create a NetDraw Dataset Q1: How to differentiate (in properties) the same tie representing different relations using a text file?

NODE ATTRIBUTES What main node attributes can be visualized? Attributes based on “external” information Whether the actor in a social network is male or female; is a 1st/ 2nd/ 3rd –grade student; majors in arts or sciences etc. Attributes based on “internal” information Whether the actor in a social network is in clique 1 or 2; has a high or low level of power etc.

NODE ATTRIBUTES How to visualize attributes based on “external” information? 1. Open the data file: “Netdraw>File>Open>Uclnetdataset>Network” 2. Edit attribute data: “Transform>Node Attribute Editor>…>File>Update and Exit” or Create an attribute data file: “UCINET>Data>Spreadsheets>Matrix>…>Save” and Open the attribute data file: “Netdraw>File>Open>Uclnetdataset>Attribute Data” 3. Visualizing the attributes: “Properties>Nodes>…”

NODE ATTRIBUTES How to visualize attributes based on “internal” information? 1. Open the data file: “Netdraw>File>Open>Uclnetdataset>Network” 2. Calculate/ use UCINET to identify internal attributes (For k-core: “Analysis>K-core”) 3. Edit attribute data: “Transform>Node Attribute Editor>…>File>Update and Exit” or Create an attribute data file: “UCINET>Data>Spreadsheets>Matrix>…>Save” and Open the attribute data file: “Netdraw>File>Open>Uclnetdataset>Attribute Data” 4. Visualizing the attributes: “Properties>Nodes>…”

NODE ATTRIBUTES Q2: What conjectures can you raise based on this graph?

NODE ATTRIBUTES 1. Not as institutional theory suggested, the information exchange among governmental and non-governmental organizations seem also very common. 2. Not as ecological theory of organizations suggested, a division of “generalists” and “specialists” seems not to affect information-sharing patterns. 3. These are simply conjectures since the number of actors is rather limited. Q3: How to cancel and revise previous actions of visualizing node attributes?

RELATION PROPERTIES What main relation properties can be visualized? Relation types (in a multiplex graph) Whether the relation is between roommates, or classmates, or both, etc. Relation types (based on node attributes) Whether the relation is between similar or different actors etc. Tie strength/ “value” of the relations Whether the relation is very strong (5), strong (4), moderate (3), weak (2), very weak (1), or even absent (0) etc.

RELATION PROPERTIES How to visualize relation types (in a multiplex graph)? 1. Open the data files (each file including one relation) 2. Visualizing the attributes: “Properties>Lines>…”

RELATION PROPERTIES How to visualize relation types (based on node attributes)? 1. Open data file 2. Open the attribute file 3. Visualizing the attributes: “Properties>Lines>…”

RELATION PROPERTIES How to visualize reciprocal ties? 1. Open data file 2. Visualizing reciprocal ties: “Analysis> Reciprocal Ties> …”

RELATION PROPERTIES How to tie strength/ “value” of the relations? 1. Open data file (with each tie measured with an ordinal/ interval variable) 2. Visualizing the attributes: “Properties>Lines>…”

LOCATION How can node position (in a 2 or 3 dimensional space) be changed (with arbitrary distances between the nodes)? “Drag and drop” method”: Move by hand Random drawing: “Layout> Random”

LOCATION Assigning the X and Y dimensions to attribute scores: “Layout> Attributes as Coordinates” To see how patterns of ties differ within and between “partitions”

LOCATION Q4: By comparing the two graphs below, what conjectures can you raise?

LOCATION Circle graphs “Layout> Circle” To visualize which nodes are most highly connected.

LOCATION Q5: What does the “optimization” option mean in the dialog box to draw a circle graph? A Guess: to put the nodes which are more connected to one side (at the top left), and those which are less connected to the other (at the bottom right).

LOCATION How can node position (in a 2 or 3 dimensional space) be changed (with distances between the nodes interpreted in a meaningful way)? Multi-Dimensional Scaling (MDS): “Layout> Graph-Theoretic Layout> MDS” Nodes that are “more similar“ (many reasonable definitions; in this example, referring to similar shortest paths (geodesic distances) ) are closer together. Direction interpretation No single “correct” interpretation

LOCATION Q6: What conjectures can you raise based on this graph (MDS solution) ? Q7: Do we only get “one” result using MDS ? Note that node “Wro” is missed; maybe due to its long distance from other nodes.

LOCATION Spring-embedding: “Layout>Graph Theoretic Layout>Spring Embedding” The algorithm uses iterative fitting. Nodes with smallest path lengths to one another are closest in the graph.

LOCATION Q8: How is the graph below (spring-embedding solution; on the left) similar to and different from the previous one (MDS solution; on the right) ? Similar in shape and interpretation (distance, and direction); but different in easiness to read.

HIGHLIGHTING PARTS OF THE NETWORK It is hard to visualize large networks in useful ways since they contain too much information. Therefore, we need to clear away some to see important main patterns more clearly. Simplify complex diagrams Locate interesting sub-graphs/ “local sub-structures” Ego Networks (Neighborhoods): To see how the complicated network arises from the local connections of individual actors

HIGHLIGHTING PARTS OF THE NETWORK How to simplify complex diagrams? Combine multiple relations into an index: “Transform> Matrix Operations> Between Datasets> Boolean Combinations” (in UNICET) Q9: How to use this file in UNICET or NetDraw more easily?

HIGHLIGHTING PARTS OF THE NETWORK How to simplify complex diagrams? Select relations you want to display Hide isolates and/ or pendants Use button-bar tools or a menu item (“Analysis> Isolates”)

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Components “Analysis> Components” To locate the parts of graph that are completely disconnected from one another.

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Blocks and Cutpoints: “Analysis> Blocks & Cutpoints” To locate parts of the graph that would become disconnected components if either one node or one relation were removed.

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? K-cores: “Analysis> K-Cores” To locate parts of the graph that form sub-groups such that each member of a sub-group is connected to N-K of the other members.  Q10: What do you think “K” means here?

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Block-based: “Analysis> Subgroups> Block-based”

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Hierarchical Clustering of Geodesic Distances: “Analysis> Subgroups> Hiclus of Geo Distances” To put nodes that are most similar in their profile of distances to all other points are joined into a cluster. 

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Factions: “Analysis> Subgroups> Factions” To form the number of groups that you desire by seeking to maximize connection within, and minimize connection between the groups. Q11: Will you always get the number of groups you desire?

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize local sub-structures? Block Modeling: “Analysis> Subgroups> Girvan-Newman” With functions similar to “Fractions” Providing measures of goodness-of-fit when partitioning different numbers of clusters

HIGHLIGHTING PARTS OF THE NETWORK How to find and visualize ego networks? “Layout> Ego Networks (New)” or “Layout> Ego Networks (Simple)” Answering questions like “who's most connected”, “how dense are the neighborhoods of particular actors”, “if one node is the ego, with what geodesic distance can the whole network be developed” etc.

DATA OUTPUT Save Diagram Save Data

SUMMARY Example of drawing a graph based on data “JMS school” Q12: How to cluster the schools to maximize goodness-of-fit? Q13: Which schools are the most connected? Q14: What other important analyses can you think of?

SUMMARY Q15: What is a good drawing of a graph/ how to properly visualize graphs? No single “right” way Easy to read (not too complicated) and draw meaningful patterns After a thoroughly consideration of data features, function, terminology, methods and tools Two remarks: When facing considerable data, one can choose to combine numerical and graphical approaches and include important nodes only. When drawing patterns from the graph, one must be very cautious.

RECOMMENDED PAPER Freeman, L. C. (2000). Visualizing social networks. Journal of social structure, 1(1), 4. Five fairly distinct phases in the development and use of point and line displays in social network analysis. Hand drawn images Images grounded in computation Early machine generated images Screen oriented images Network images in the era of web browsers

Other Questions ?

Thank you !