Scope out functions to create

[ttimbers]

No description provided.

[ttimbers]

Here is my idea on how we can modularize some of our code to functions to reduce redundancy:

1. A function named summarize_col that takes a data frame, and name of a column and returns the median, mean and standard deviation for that column. We would call it to create listings_train_summary and listings_test_summary

2. A function named scatter_lm that takes a data frame, a col to map to the x-axis, a column to map to the y-axis. It would return a scatter plot with a linear regression line. It would reduce the repetitive code current in our EDA section - especially if we paired it with mapping it across the data frame columns to generate a list to plot with plot_grid.

3. We could make our repetitive and verbose use of str_remove(...) |> as.numeric() inside mutate more readable by abstracting this to a function. We could name this function make_numeric.