For CS Masters Program at CWU
About: This application displays multi-dimensional data in 2D using OpenGL for rendering. The plot can be dragged and zoomed in/out. Classes can be hidden, as well as specified attribute markers. A box-clipping algorithm is included to clip lines and return samples of the dataset.
This visualization tool features four plots, including Parallel Coordinates, Paired Coordinates, DSC1, and DSC2. Additional multidimensional plots can be added with ease as the plot context class uses general vertices.
--Dataset Information: Dataset must be in .txt or .csv format Dataset must include headers Dataset should only include class and attribute columns Dataset class header must be labeled as "class" without quotations Dataset features besides "class" must be in numeric representation --Example dataset: length,width,height,class 2.7,3.5,3.2,dog 1.2,5.5,2.1,cat 2.5,4.1,1.6,dog --Language Used: Python 3.9 --Required Packages: *version number can be different if classes and functions are present PyOpenGL 3.1.5 PyQt5-sip <- install this first on anaconda. (Not needed on Pycharm) PyQt5 5.15.6 numpy 1.21.4 pandas 1.3.4 scikit-learn 1.0.1 --Not Required but the .UI file for PyQt5 was designed with: QTDesigner -- Functions: Left click on plot and move mouse to drag the plot. Scroll the mouse wheel to zoom in/out of the plot. Right click twice to make a box clipping rectangle. The 1st right click is the upper right corner, and the 2nd right click is the bottom left corner. The cells in the class and attribute tables can be dragged and dropped to switch their orders. The slider below the attribute table will change the transparency of the attribute markers that are not selected in the highlight column. --Dataset Links: Iris dataset - https://archive.ics.uci.edu/ml/datasets/iris Breast Cancer Wisconsin (Original) dataset - https://archive.ics.uci.edu/ml/datasets/breast+cancer+wisconsin+%28original%29
MNIST on DSC2 using t-SNE as scaffold origin points. Image contains 3,120,000 data points (60,000 * 52).
Showing Zoom and Drag capabilities of App
