Web interface with linked visualizations of multi-dimensional data. Source code accompanying my master thesis.
For running:
npm install
For compiling documentation:
npm install markdown-pdf
plantuml
Generate data with data/templates/build.sh
.
Start a local web server, for example:
python -m http.server 8888
# or
python -m SimpleHTTPServer 8888
Open index.html
with Google Chrome browser.
For further development, see docs/HACKING.pdf
.
Compile it with docs/build.sh
.
Layout of the interface: ActionPanel (A), Panes (B), MainPanel (C).
Co-Occurance matrix, arranged for all pairs of attribute categories.
Co-Occurance matrix, arranged for distinct pairs of attribute categories. The ActionPanel contains user set values, while the PatientDistributions Icicle Plot shows the hierarchical application of filters.
Spiral visualization. User applies brushing to the time interval, reducing the number of sectors displayed.
Heatmap with summarized frequencies of attribute occurrences.
Timelines for time periods between attribute occurrences.