Visualizing monotone interleavings

This repository implements ParkView, a method for visualizing monotone interleavings.

To run ParkView, do the following.

1. Download the repository.
2. Open the repository directory in the terminal.
3. Execute .\gradlew run for Windows or ./gradlew run for a Unix-based operating system (make sure Java is installed properly; we use OpenJDK version 20). This should open a window with ParkView.
4. A pair of merge trees is loaded by default. To visualize custom merge trees, drag and drop a pair of .txt files representing merge trees into the application.

Key F11 opens and closes a GUI panel, which can be used to export the trees to SVG.

Merge tree format

We represent ordered merge trees using a plain text format. Specifically, an ordered merge tree is represented by a pair of parenthesis that has within it first the height of the root of the tree, and then a sequence of subtrees. For example, consider a tree rooted at height 10, with two subtrees: 1. a leaf at height 5, and 2. an internal node at height 3 with two leaves at heights 2 and 0. This is encoded as follows:

(10(5)(3(2)(0)))

We use The Topology Toolkit to generate merge trees whose arcs and nodes we export to CSV files. We transform the CSV files into ordered merge trees in our format using the Python script ttk-parser/main.py. The script expects the following arguments:

main.py {nodes} {arcs} {minX} {maxX} {minY} {maxY} {scale}

Here {arcs} and {nodes} are the paths to the CSV files containing the corresponding parts of the merge tree output by TTK. The remaining arguments are needed to create the space filling curve used for determining an order on the leaves. The four arguments {minX}, {maxX}, {minY}, and {maxY} specify the bounding box of the scaled scalar field using scaling factor {scale}. Scaling may be necessary for some scalar fields as the cells of the space filling curve are each a unit square.

Directory data/merge-trees contains the ordered merge trees in txt format used in the paper.