Visualize, explore, and compare multi-dimensional data on the web.
TODO: Screenshot
SAAV offers:
-
A Quality view, which visualizes quantitative, comparable dimensions of your data. It uses nested parallel coordinate visualizations to explore data dimensions across hierarchical levels. Mean values are calculated ("aggregated") for each hierarchical level, eventually producing an overall ranking of your data entities. Customizable weights control how dimensions contribute to the ranking.
-
A Profile view, which visualizes dimensions of your data for which a mathematical comparison does not make sense. Using a bubble visualization, this view allows to explore data dimensions with respect to similarity.
-
An Expert Configuration panel to assign data dimensions to views (Quality vs. Profile) and control aggregation weights.
-
An exportable PDF report of your analysis
If you just want to get a feel for SAAV, use it with demo data right here and now.
If you would like to load your own data, read on to learn about supported input data.
SAAV expects a CSV data matrix with the following columns (order matters):
"Entity","Hierarchy","Reviewer","Value"
-
Entity: The names of the things you are comparing (e.g. "Project 1", "John Doe", "Age Group < 20")
-
Hierarchy: The hierarchical path; three levels, separated by ":::" (e.g. "Language:::French:::Grade")
-
Reviewer: An identifier of an individual measurement (e.g. "Reviewer 1")
-
Value: The actual value (parsed as
Double)
TODO: Example Input Data
In addition to loading data from file via "drag & drop", SAAV supports auto-loading data from any URL:
|
Tip
|
This is particularly convenient if you want to embed SAAV on your website (e.g. in an iframe), with all data already loaded.
|
SAAV optionally supports a file-based configuration to provide application defaults and control input data validation. Specifically, it allows to
-
Customize an analysis title
-
Constrain the range of allowed values (aborting the import process upon violations)
-
Specify default weights ("Quality" vs "Profile" view; aggregation weights)
-
Specify an expected hierarchy (detection of input data mismatches)
A SAAV config file can be loaded from any URL:
The latest milestone is deployed via GitHub Pages:
This project originally provided software to analyse and visualise research output in the humanities and social sciences (SAAV). The application is quite generic now, which hopefully makes it re-usable for all kinds of data. Development was a joint effort between the University of Lucerne and the University of Applied Sciences and Arts Northwestern Switzerland FHNW. The software was presented at the final SUK P3 Conference (November 2016; Poster).
All source code is available under the MIT License.