Java Parallel Coordinates Visualization Tool

Java Parallel Coordinates Visualization Tool, visualizing multidimensional/multivariate CSV data with Java Swing.

What are Parallel Coordinates

"Parallel Coordinates plots are a common method of visualizing high-dimensional datasets to analyze multivariate data having multiple variables, or attributes." Wikipedia

Screenshots

Iris dataset render

Heart disease dataset render

Wisconsin breast cancer diagnosis render

MNIST render

Prerequisites

Make sure Java is installed on your system to compile and run this application.

How to Compile

To get started, open a command line interface (CLI) and navigate to your project directory. Run the following commands:

# Compile all Java files in the javaPC directory
javac javaPC/*.java

How to Run

# Run the main Driver class
java javaPC.Driver

How to Build the Jar

jar cvfm Parallel-Coordinates-Vis.jar manifest.txt javaPC/*.class

Makefile Script

The Makefile in this project simplifies the process of compiling Java files and building the JAR file. Below are the provided recipes:

• Compile Java Files and Build the JAR: This single command compiles all .java files within the javaPC directory and packages the compiled .class files into a JAR file, including the specified manifest.

  make

• Clean: Removes all compiled .class files and the generated JAR file to clean the project directory.

  make clean

How to Use

To plot a dataset make sure the class/id column is the last column.

• Start program with Java by running JavaPC/Main.java or by running the compiled JAR file.
• Click 'Load CSV' to open the file picker and select a dataset to visualize.
• Click 'Render Plot', rerendering the plot will generate a new color scheme if preferred.
• Click 'Toggle Labels', to toggle on/off the visibility of class, attribute, and ranges.
• Click 'Histogram', to toggle on/off histogram correlated density sized vertices.

Datasets

• breast-cancer-wisconsin.csv - UCI's Wisconsin breast cancer dataset with 30 features. Classes: Malignant, Benign
• breast-cancer-wisconsin-9f.csv - UCI's Wisconsin breast cancer dataset with 9 features. Classes: Malignant, Benign
• diabetes.csv - UCI's diabetes dataset. Classes: Negative, Positive
• fisher_iris_SVe.csv - Subset of Fisher's Iris dataset for Versicolor and Setosa. Classes: Versicolor, Setosa
• heart.csv - UCI's heart disease dataset. Classes: Absence, Presence
• ionosphere.csv - Dataset of radar data. Classes: Good, Bad
• iris.csv - Fisher's Iris flower classification dataset. Classes: Virginica, Versicolor, Setosa
• iris_S_vs_VW.csv - Subset of Fisher's Iris dataset for Setosa vs. others. Classes: Setosa, Other
• iris_setosa.csv - Subset of Fisher's Iris dataset, only Setosa class. Classes: Setosa
• iris_SVe_vs_Vi.csv - Subset of Fisher's Iris dataset, comparing Setosa/Versicolor against Virginica. Classes: Setosa/Versicolor, Virginica
• iris_SVi_vs_Ve.csv - Subset of Fisher's Iris dataset, comparing Setosa/Virginica against Versicolor. Classes: Setosa/Virginica, Versicolor
• iris_V_vs_V.csv - Subset of Fisher's Iris dataset, comparing Virginica against others. Classes: Virginica, Other
• mnist_letters.csv - MNIST's capital letter dimensions of handwriting dataset. Classes: Capital letters A - Z
• sin_cos.csv - Dataset containing sin and cos values for various angles. Classes: Sin, Cos
• wheat_seeds.csv - UCI's wheat seeds dataset. Classes: 0, 1, 2
• wine.csv - Dataset with chemical analysis of wines. Classes: Class_0, Class_1, Class_2