wk04-tidytuesday-commentary-hs

[sotohaven]

The visualizations by @NicciPotts caught my attention because of how clean and organized it appears. The way she shows the change in median salary for nurses from 2009 to 2018 in each state is clear and easy to understand even with minimal text to explain the graph (Figure 2). Using the figure caption as a legend was an interesting way to omit a legend in an already busy graph. Figure 1 in comparison was a bit more difficult to understand due to her choice of colour scheme as it made it difficult to read. It was also difficult to understand what the percentages meant relative to the state salary upon initial inspection. To improve Figure 1, I would change the colour scheme and add a bit more explanation in the figure caption to explain the data. I did enjoy the way she created a map with the boxes, showing which region the states approximately fall in.

A few things that had stood out to me from her code that we had touched upon in our lecture:

• library()
• extracting specific columns from datasets (nurses <- teusdata$nurses)
• %>% (other students have mentioned that this is used in many codes, likely due to its usefulness)
• ggplot() (we have yet to go into detail about how to use ggplot, but I'm excited to learn about it. I have some experience using ggplot, but I find that base R is more straightforward when creating plots even if they aren't as visually exciting)

Some codes that I have seen variations of, but I'm not sure if they perform the exact same function:

• The way Dr. Potts loaded data is different from what we have covered (tuesdata <- tidytuesdayR::tt_load('2021-10-05'))
• She also used a code to filter for specific years and states to be a part of her data set, which looks different from what I am familiar with in base R

e.g., tot = subset(tot, Age !="") is a line of code from my own work that I used to remove rows of data that did not contain a value for age. Dr. Potts' code to choose specific years to be part of her analysis: nurses_df1 <- nurses %>% filter(Year %in% c('2009', '2015', '2018'))

[lcorlett]

Good observation on Dr. Potts' alternative method for uploading data! Makes me want to go back and open the code on each individual post :100: :clap: