Basic Visualization in Rstudio

 


For this assignment, we did some basic visualisations in RStudio. Here we can first see a pie chart showing counts and colors. For the pie chart and the boxplot, we did 40, 30, 20, 10. This analysis is quite simple, as its main focus is to summarize and compare instead of making predictions or testing a hypothesis. 

The pie chart fits less strongly with  Few's typical critique of pie charts, while it still communicates a valid idea of part-to-whole, since the total adds up to 100. In the pie chart, people might have a hard time comparing the sizes accurately, especially since some values are close to each other. For example its hard to see an accurate difference between 30 and 20, while piecharts are good at showing the different proportions, they can be hard to accurately compare. 

On the other hand, I think the boxplot fits well with the ideas by Few and Yau. The boxplot makes the comparison clear as they all share a common baseline, so it's easier to see the difference between them. This supports their goal of descriptive analysis, quickly being able to communicate what's bigger or smaller and by how much, with minimal effort for the reader. 

Overall, both charts summarize the same information, but the boxplot best matches those principles by Few and Yau. It gives the viwer and easy chart where the difference is noticeable and accurate to compare. The pie chart can be good at showing parts of a whole, but it's not as readable for comparing categories. I personally think the box plot is the better chart to show comparisons, and it will be easier for the reader to view and compare the differences between the different objects in the chart. 

Comments

Post a Comment

Popular posts from this blog

My Geographic Map of Miami-Dade County Public Libraries

Image
  Map of Public Libraries in Miami-Dade County https://docs.google.com/spreadsheets/d/1T17POMNsTbxoCZmJmeO4m2vuOLzXWeINip18Q1LwdNE/edit?gid=1338048746#gid=1338048746 This is a map of the public libraries in Miami-Dade County that I created using a dataset and Tableau. The primary challenge I faced when creating this map was selecting a suitable dataset. The first one, I found some cities and locations were not showing, it was because Tableau was not recognizing them properly, even though they were spelled correctly. After switching to a different dataset, it all worked much nicer.  I used the same consistent color and size across all the points ot reate similarity between them and convey that to the viewer. Making all points the same tells the viewer they all represent public libraries. The principle of proximity allows us to see clusters of libraries in those heavily populated areas. I think this map could be further improved by adding more detail, like increasing the point s...
Read more

Visualizing Distribution in RStudio

Image
  For this weeks assignemnt, we were tasked with showing visual distribution using RStudio. I chose the built-in dataset mtcars, which contains many different attributes about 32 different cars. To show the distribution, I created a histogram using ggplot 2 to visualize the distribution of one of the most important things when searching for a new car, miles per gallon (mpg). Looking at the histogram, we can see that most vehicles fall between 15-25 mpg, with a few cars achieving above 30mpg. While designing this visualization, I followed Few's recommendations by just keeping the layout simple, using consistent coloring, and labeling the axis clearly. I do agree that sometimes distributions can be overcomplicated, which can cause them to be as clear to read. In my case, I try to make it as simple as possible while remaining visually engaging and appealing. 
Read more

Using Plotly for Data Visualization

Image
  Using the given dataset, I went with Plotly to analyze the dataset as its simplicity to use and customization grabbed my attention. I have to say, after using it for a bit and playing around with it, Plotly is a very intuitive tool to have in the data science field. I used these two charts to compare the difference between part-to-whole charts and ranking charts. The ranking chart helped me visualize closely how each ID compared to the average position. By having them in an ascending order, we can see which ones are performing better relatively to others. This type of visualization is best put to use when comparing data within the same dataset. On the other hand, the part-to-whole chart shows how each ID contributes to the total average position. The main focus of this visualization is to show proportion and distribution. We can see how using the same dataset can be interpreted differently depending on which type of visualization we apply to it. 
Read more