[rahimnathwani]

This is a deliberate pedagogical choice, and one which will familiar to those who did one of Jeremy Howard's deep learning courses.

  One of the challenges of learning Linear Algebra is where to start. Most textbooks start with vector arithmetic, which make senses if you are working with paper and pencil, but they take a long time to get to something useful.
  
  With a computational approach, we have the option to proceed top-down -- that is, we can start with libraries that implement the core algorithms of linear algebra, and wait until later to see how they work. With this approach we can can get to the good stuff faster.