MultiViz

Web interface with linked visualizations of multi-dimensional data. Source code accompanying my master thesis.

Dependencies

For running:

npm install

For compiling documentation:

npm install markdown-pdf
plantuml

Running

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.

Documentation

For further development, see docs/HACKING.pdf.

Compile it with docs/build.sh.

Gallery

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.