inst/gallery/BreastCancer/README.md

In matloff/prVis: Polynomial-Based Visualization

Data

Data can be found here. Features are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. The first column us the patient ID, the second column is the classification in diagnosis: M = malignant, B = benign, and the remaining columns are real-valued features computed for each cell nucleus.

Graph

prVis

prVis shows a good separation of data

When we try to further zoom in the graph by removing 50 outliers (about 12% of the original dataset), prVis shows a clearer distinction between red(malignant) and black(benign), which may provides us the insight on what distinct them from each other.

umap

t-sne

matloff/prVis documentation built on Dec. 3, 2019, 9:27 a.m.