This week, Ellis Hughes and I use pro beach volleyball data, provided by TidyTuesday, to visualize winning percentage for different teams over time. We then show a simple Bayes adjustment for winning percentage when teams have small sample sizes (similar to what I discussed in an older blog article). Finally, we plot distributions of winning percentage to compare different teams. But before that, we go over code from Eric Ekholm, who produced a really cool plot of the performance metrics of Kerri Walsh-Jennings and Misty May-Treanor across their respective careers.

To watch the screen cast, CLICK HERE.

For our code, CLICK HERE.

This week, instead of working through someone else’s code, Ellis Hughes and I decided to use the TidyTuesday data set on volcano eruptions and walk through how we might start to approach looking at the data from an exploratory data analysis perspective.

Along the way, we not only show how we approach cleaning the data and creating new features but also touch on the value and ease of using the ggplotly() to quickly construct interactive visualizations from the ggplot2 visuals you’ve already built. These types of visuals provide a useful and interesting way to exploring your data are can be valuable when sharing information with decision makers to assist discussions around how to proceed with further analysis.

To watch the screen cast, CLICK HERE.

To view our code, CLICK HERE.

In this episode, Ellis Hughes and I discuss the use of data tables to accompany your visuals in rmarkdown reports.

First we go over code provided by Ted Laderas, who used the gt package to create some nice data tables on the Animal Crossings data set from the TidyTuesday Project.

Then, we apply a similar approach to NFL Kicker data and show how you can add conditional formatting to your table to aid in data presentation.

For the screen cast, CLICK HERE.

For our code, CLICK HERE.

This week, Ellis Hughes and I discuss plotting lines for a single group with grey traced lines in the background representing the rest of the population.

First, with data provided by the TidyTuesday Project, we go over how Jake Kaupp used this approach to visualize some of the top grossing Broadway shows over the past couple decades.

Then, we apply this same approach to historic NBA data to show how different teams have trended in a variety of game statistics and produce a plot like this:

To watch our screen cast, CLICK HERE.

To access our code, CLICK HERE.

In today’s screen cast, Ellis Hughes and I discuss how to get started using git.

We go over:

1) How to get started with git.
2) The value of git when working on projections with colleagues.
3) How version control works when yourself or colleagues make updates to the project.
4) How to push your updates to github.

To watch the screen cast, CLICK HERE.