Peer Review (Andrew) - Githubissues

Hey Kivalina (+ @datalorax),

First off, you went totally above and beyond to create some super stunning visualizations! leaflet also seems like an amazing package. It creates so much with only a small amount of code. Since I'm not too familiar with packages like leaflet, tidytext, or shiny, I'm afraid I won't be able to give the most guided feedback about how you implemented them.

But what I can is that you've certainly inspired me to look into those packages in more detail moving forward. leaflet and shiny complement each other so nicely. I loved being able to click on an individual station and see information about it. Putting interactivity into those map-based plots makes relatively complex information more accessible, as the audience can attend to individual components of the plot with ease. These visualizations aren't just good for the requirements of the final project — they are sleek and professional presentation-ready pieces!

A general comment I do have is to make sure in your blog post that you provide some context about what the leaflet visualizations are designed to examine. For instance, in Viz 1, are we looking at train stations and their respective routes in Boston? It's a bit unclear. There is nothing I would change about the visualization itself, but some contextualization will help the audience derive meaning from what they see. The same applies for Viz 2, where you color the stations differently by rides started. I see that the individual data points are sized according to the number of docks divided by three. Be sure to indicate this in your plot somewhere, and also communicate in your write-up why this distinction was important to make.

In Viz 3, where you plot density distributions of average ride length on Saturday and September overall, I would again be a bit clearer about what is being communicated. Average ride length using what mode of transportation and in what area? Additionally, I might consider using a different color than black to represent the overall average ride length. The reason for this is that your light blue and black blend together to create a darker blue, which I thought was actually a third density distribution for a while. Experiment with different colors to see what makes this visualization as clear as possible. Alternatively, you might consider using ggridges to showcase the same information, but without the two density distributions on top of one another. However, I also recognize that the overlay makes differences between the two a bit clearer. For example, the audience can see that rides tend to be longer on Saturday than the overall.

Finally, the word cloud (Viz 4) is awesome! My only comment on that is I see both "issue" and "issues" in the plot. Perhaps "issues" is worth replacing with whatever the next most-mentioned word was on the list.

Overall, again, this is sensational work, and I will be referencing your code as I delve into (hopefully) my own journey with leaflet, tidytext, and shiny.

Best of luck with your final portfolio!

Andrew

kgrove10 / Final-Project-EDLD610

Peer Review (Andrew) #2