Building a UMAP using Seaborn

In this recipe, we’ll learn about a newer and very visually appealing clustering algorithm called UMAP, using our breast cancer dataset!

UMAP stands for Uniform Manifold Approximation and Projection. It is useful to understand the structure of higher-dimensional data, even when it is non-linear. UMAP essentially takes a high-dimensional space and represents it using the most equivalent lower-dimensional graph it can find. It is also fast and efficient.

Getting ready

The code for this recipe can be found in Ch04/Ch04-6-seaborn.ipynb.

You will need to install the seaborn and umap-learn packages if you don’t already have them. You can do this from the terminal by typing the following:

pip install seaborn
pip install umap-learn

Or you can install these from the notebook like this:

! pip install umap-learn
! pip install seaborn
! pip install ipywidgets

This installs the umap package. We also install seaborn here, although...