Advanced Machine Learning with R

Descriptive statistics

Traditionally, we could use the base R summary() function to identify some basic statistics. Now, and recently I might add, I like to use the package sjmisc and its descr() function. It produces a more readable output, and you can assign that output to a dataframe. What works well is to create that dataframe, save it as a .csv, and explore it at your leisure. It automatically selects numeric features only. It also fits well with tidyverse so that you can incorporate dplyr functions such as group_by() and filter(). Here's an example in our case where we examine the descriptive stats for the infantry of the Confederate Army. The output will consist of the following:

• var: feature name
• type: integer
• n: number of observations ...