PSTAT 100: Data Science Concepts and Analysis
Selected Solutions
Important
Below are solutions to selected assignment. Some things to keep in mind:
- There may be typos, so please proceed with caution (feel free to ask during Office Hours!)
- There are often many different ways to approach a problem, each of them “correct”! As such, if you work through a problem in a different way than the solutions that’s fine - your way is not necessarily wrong.
Labs
Lab04 Note: There was a typo in Question 9: you were meant to overlay the density over your histogram from Question 8, not Question 7. (The grader graded very leniently, as a result.)
Lab07 Note: I know people had some trouble getting Question 6 to work. In a sense, this was to be expected and one of the reasons why multicollinearity is so dangerous!
- For the purposes of an ICA, you should just make sure you’re comfortable detecting multicollinearity using a correlation matrix, and be generally aware of some of the implications multicollinearity has on a model (if it is present).