Get startedGet started for free

Alternatives to the Regular Matrix Inverse

In the last video, we talked briefly about generalized or pseudo-inverses.

In this exercise, you'll use the Moore-Penrose generalized inverse in the `MASS' package and find that it produces the regular inverse if the matrix you're working with is already invertible!

The command ginv() computes the Moore-Penrose generalized inverse in R.

For more information on the Moore-Penrose inverse, see the following link.

This exercise is part of the course

Linear Algebra for Data Science in R

View Course

Exercise instructions

  • The WNBA Massey Matrix M and vector of point differentials f are loaded for you. Print M again.
  • Apply the regular inverse to M to f and print the result.
  • The MASS package is already loaded for you. Apply the generalized inverse ginv(M) to `f' and print the result.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Print M
print(M)

# Find the rating vector the conventional way
r <- ___%*%f
colnames(r) <- "Rating"
print(r)

# Find the rating vector using ginv
r <- ___ ___ ___
colnames(r) <- "Rating"
print(r)
Edit and Run Code