- Performs Linear Regression with ZERO import statements (Except if you want visualization, then matplotlib is imported)
- Considers the last column as the dependant variable and all non-numeric data as catagorical.
- Automatically performs One-Hot Encoding on all non-numeric columns.
- Also calculates and displayes the R^2 value for the test set.
This is more of a programming exercise than it is a data science exercise, but helpful to anyone trying to get into either. As no libraries are imported (not even pandas or numpy :) ), even the most trivial tasks like generating random numbers and reading CSV files are handled organically.
All suggestions/PRs welcome.