Module 5 - Part to Whole and Ranking

This week we were introduced to the visualization dashboard application Plot.ly. I created several plots using the data provided by the instructor, which had two dimensions or variables, average position and time. 

First, I created two pie charts to represent the relationship of each data point of both the average position and time in relation to the sum of each set of data: 





While these pie charts represent the proportion of each value to the whole or sum of the values, it does not give us a very good representation of the two dimensions in relation to each other. Additionally, the categories in the legends do not correspond to meaningful aspects of the data. The pie chart function of Plot.ly therefore does not express meaningful relationships between the graph and the legend.

Further, a drawback of the export function is that the legend is cut off near the bottom, so it cannot be used to effectively interpret the pie charts. Otherwise, the colors of the pie chart are very vibrant, and it is easy to tell the difference between each component of the chart. 

Next, I created a bar graph and a line graph to represent the relationship between the x value, time, and the y value, average position. 



I found these visualizations to be more useful in identifying relationships between the values of each dimension.

I wanted to experiment with using multiple graphs to better elucidate the relationships between the data, so I created a scatter plot and a line graph visualization, as well as a Pareto chart, to better elucidate the relationships of the values.



Finally, I created a dashboard to showcase all of my visualizations. Here is the link.

Overall, the scatter, line, and bar graph visualizations reveal the same insights, that there is a strong, positive, linear relationship between the time variable and the average position variable. With a more contextualized data set, Plot.ly is potentially a powerful tool to generate visualizations quickly that express relationships between many different types of data in many different ways.


Comments

Post a Comment

Popular posts from this blog

Module 2 - Geographic Visualizations in Tableau

Image
The following visualization (which can be interacted with here: https://tinyurl.com/publichousinginflorida) was developed based on Housing and Urban Development data for public housing developments in the United States, updated in December of 2023. The visualization represents the average household income (size of the marks) and average rent (shade of the marks) for each development in the state of Florida.  The dataset used for this visualization contains many variables, including number of occupied dwellings and vacancies, percentage of demographic groups in each development (Native American, Hispanic, female head of household with and without children), among many others.  Initially, I found it difficult to organize the visualization, because my impulse was to create pie chart markings based on occupancy. However, this proved difficult. This visualization is the result of a compromise I made with myself to represent some aspect of the data without thinking too hard about ho...
Read more

Final Project

Image
 For the final project, I created a correlated graph a selection of data from the Orlando Police Department. View here. Between 2009 and 2024, suspect fatalities were fewer than suspects shot. Suspect fatalities were alarmingly high.
Read more

Module 7 - Distribution Analysis

Image
 For this module, we were assigned to create a visualization based on distribution analysis. I created a bar graph based on the mpg table of the mtcars dataset in R. Below is my graph: This graph shows the miles per gallon of each vehicle in the mtcars dataset. A drawback of this graph is that the vehicles are not identified on the x axis, therefore, the graph does not demonstrate the best information on distribution for this dataset.  Here is the code:  data("mtcars") head(mtcars) counts <- table(mtcars$mgp) counts2 <- table(mtcars$wt) counts View(mtcars) barplot(mtcars$mpg, main = "Miles per Gallon", xlab = "MPG", ylab = "count", col = "orange")
Read more