Final Project: Visual Analytics


For this project, I was not sure which dataset to use, so I went through the R built-in datasets and chose the ToothGrowth dataset. It was easy to work with, and it has enough observations for me to make different visuals and compare patterns. The main question I wanted to go over in this project was how different supplement types and dose amounts affect tooth length. I also wanted to see if there was a big difference when increasing the doses. My goal was to understand the patterns in the ToothGrowth dataset.

Related works I looked back on was the mtcars dataset we worked on for some of the assignments. We used boxplots, scatterplots, and other visuals to compare variables, which helped me decide which graphs to use for this project and gave me ideas for the visuals I made for the ToothGrowth dataset.

For my solution, I used RStudio to make all the visuals and used a dataset that was already built into R. I created three graphs that helped me explore the dataset further. The first graph I made was a boxplot comparing tooth length between the two supplement types (OJ and VC). I chose this because it clearly shows the differences between the two groups. My second graph is a scatterplot showing the relationship between dose and tooth length, which helped me see how tooth length changes as the dose increases. My third graph is a histogram that shows the overall distribution of tooth lengths across the dataset.





This visual compares tooth length for two different supplements types which are OJ and VC.

OJ = Orange Juice

VC = Vitamin C

I created a boxplot that represents the distribution of tooth length when the supplement is used and colored to make it better and easier to understand which is which. From the visual, the OJ group shows a higher tooth length overall. The overall distribution and the median of OJ are much higher than those of VC, which means OJ leads to more tooth growth compared to VC.








I created a scatterplot showing how tooth length changes with dose. From the scatterplot, we can see that dose is on the x-axis and tooth length is on the y-axis, and on the side, OJ and VC are shown as the supplements. The main thing I noticed from this plot is that when the dose increases, tooth length also increases, and that higher doses have higher points on the graph. Also, at 0.5 mg, OJ has a longer tooth length than VC. At 1 mg, OJ is still higher; at 2 mg, OJ and VC are similar, but VC is slightly higher.






For my last visualization, I chose a histogram to showcase the distribution of tooth lengths. From the visual, there is a spread from 5 to 35, and most tooth lengths fall in the middle range, which is around 15-25. There are also very few small teeth or very large teeth. Looking at the shape, it is neither symmetric nor skewed.



Comments

Post a Comment

Popular posts from this blog

Module #9: Visualizing Multivariate Data with ggplot2 & Corrgram

Image
  I used the Iris, which has many flowers and includes details on sepal length and width, as well as petal length and width. I used this dataset because it measures different flower types, such as setosa, versicolor, and virginica.  My visualization measures sepal length and compares it across the three species of iris flowers. The boxplot shows that setosa has the smallest sepal length, virginica has the largest, and versicolor is in between. This visualization helps tell the differences and similarities.  Yes, the boxplot is a multivariate visualization because it compares sepal length across three species. Making it easy to compare them side by side.  In my boxplot, I used alignment by keeping everything easy to read and compare. I also used contrast by using a gray background with red boxes so the boxplots would stand out, and I used balance by spacing the boxplots evenly so the chart looks neat and easier to read and understand. 
Read more

Assignment # 12 using Python with netwokxx and plotnine

Image
  The code ran successfully. It was difficult at first because I don’t usually use Python, so I had to install it and figure out how it works. At first, I was typing the commands into the Python shell and the code didn’t run. After doing some research, I realized I needed to use Command Prompt instead to install everything correctly. Once I ran the code there, it worked and generated the ggplot output confirming the graph was created. I personally would not use this method because it was confusing to set up and the graph did not show up in the command prompt.
Read more