160 likes | 263 Views
Trees. cs5984: Information Visualization Chris North. Review. Data space: Multi-dimensional 1-D space 2-D space Interaction strategies: Dynamic Queries Multiple views, brushing & linking Visual overviews Zooming, overview+detail, focus+context Design guidelines Empirical Evaluation.
E N D
Trees cs5984: Information Visualization Chris North
Review • Data space: • Multi-dimensional • 1-D space • 2-D space • Interaction strategies: • Dynamic Queries • Multiple views, brushing & linking • Visual overviews • Zooming, overview+detail, focus+context • Design guidelines • Empirical Evaluation
Next • Data space: • 3-D • Trees • Networks • Document collections • Workspaces • Theory • …
Trees (Hierarchies) • What is a tree? • Items + structure • Add parent pointer attribute • Examples • Family trees, Directories, Org charts, biology taxonomy, menus • Tasks • All previous tasks plus structure-based tasks: • Find descendants, ancestors, siblings, cousins • Overall structure, height, breadth, dense/sparse areas
Tree Visualization • Example: Outliner • Why is tree visualization hard? • Structure AND items • Structure harder, consumes more space • Data size grows very quickly (exponential) • #nodes = bheight
2 Approaches • Connection (node & link) • Containment (node in node) • Structure vs. attributes • Attributes only (multi-dimensional viz) • Structure only (1 attribute, e.g. name) • Structure + attributes A B C A B C
Outliner • Good for directed search tasks • Not good for learning structure • No attributes • Apx 50 items visible • Lose path to root for deep nodes
Mac Finder Branching factor: Small large
Today • Rao, “Hyperbolic Tree”, book pg 382 • Joy, maulik
Nifty site of the day: X-Files • http://www.thexfiles.com/main_flash.html
ConeTree / CamTree • Video CHI’91
WebTOC • Website map: Outliner + size attributes • http://www.cs.umd.edu/projects/hcil/webtoc/fhcil.html
PDQ Trees • Overview+Detail of 2D layout • Dynamic Queries on each level for pruning
Assignment • Read for Thurs • Johnson, “Treemaps”, book pg 152 • Stasko, “Sunburst”, web • Marcus, marty • Homework #2 due Thurs • Spring Break! • Read for Tues (Mar 13) • Beaudoin, “Cheops”, web • Satya, sumithra • Furnas, “Fisheye View”, book pg 311
Scenario: Visualizing Biotech Data • Database of experiments on DNA • 1000 experiments? • DNA = long sequence of letters A,C,T,G • 100,000 – 1,000,000 letters • Experiment = data values for set of sub-sequences • 1000 sub-sequences, 10-100 letters / sub-sequence • Tasks: • Find experiments given criteria • Find patterns between known set of experiments • Find related experiments • Find trends in experimentation DNA: AAGTGTTCCGAAATGCAAAAATAGACCCAAAGA… Experiment: (5-50)=1.4, (72-112)=0.2, …