The following is a listing of demos of student projects done in the Fall 2022 Principles of Computing course. The students found their choice of data source from resources online and decided how they want to examine, analyze and visualize the data.
We took 5 csv's containing information about mass shootings in the United States from 2018-2022 and uploaded them into python to analyze. We combined them into one large set of data to further analyze topics such as trends of mass shootings per state in America, during months from 2018-2022, and the amount of deaths and injuries that result from the mass shootings.
We wanted to use this project opportunity to combine what we learned in computer science to our economics majors by using questions of income inequality as a means to do this. We were curious about what income inequality looks like across the world and how countries compare. We would hope for someone who is interested in economic questions of income inequality, such as an economist or a social scientist, to use this project as, maybe more so, an inspiration to combine concepts of python to visualize economic research. We chose data sources of the Gini Index and Gapminder to compare the Gini Index to other country measures like Population Counts, GDP per capita. While working on this project, we discovered that many countries rarely measure the Gini Index, Brazil was one of the only countries that measured it frequently. So, we used the first two visualizations to put the GINI index into context of measures of population and GDP per capita. The second two visualizations, we compare Brazil’s GINI index to Brazil's population and GDP per capita.
In our project, we collected five datasets from online which hosted the mass shooting incidents in the United States from 2018-2022. In these datasets we were able to collect information such as the state in which the incident occurred, the number injured in the incident, the number killed in the incident, and a brief description of the incident. We ultimately brought this data together into one source which we were able to look at and couple with related statistics like education rates and population rates which could potentially help us paint a better picture of the trends and reasoning for this alarming amount of gun violence in the United States.
We used a very large dataset from multiple sources that allowed us to be able to gather some useful insights from our visualizations. For example, through using poverty data from the Census Bureau, we were able to find that there is a positive correlation between poverty and discipline of Hispanic males. We also explored the number of counties with agreements, which has changed a bit over time, and the concentration of school discipline of Hispanic males across various regions of the country.
To start off, we uploaded the airlinesatisfaction.csv file to google colab. We then turned the file into a list and did some data cleaning in order to make the file into a dataframe for easy use and data retrieval. We then went through and made visualizations that we thought would be interesting and necessary to analyze in our project and throughout each visualization we generally had to turn the data into an int datatype. We focused on the relationship between delays/flight promptness and satisfaction. We created a scatter plot displaying arrival and departure delay times and satisfaction. We then divided the data into those satisfied and not and made a box plot using the flight distance in miles. After that, we focused on a few demographic traits to gain insight into the dataset and what contributes to passenger satisfaction. We analyzed gender and satisfaction in a pie chart and age and satisfaction in a histogram. Then we looked at combined total rating and satisfaction to see if improving the service ratings actually contributed to passenger satisfaction which it did. After creating all the visualizations we looked at each to gain our insights and wrote these insights down.
Our project started by wanting to synthesize and visualize public data from the Federal Election Committee's website of financial Congressional campaign information. We took steps to clean the data and separate it into lists, using those lists to manipulate the data into dataframes. Then, we utilized the Matplotlib and Plotly.express libraries to create visualizations. Finally, we built a comprehensive website to display this data and our insights for a larger audience using Bootstrapr for the base HTML code and GitHub to host our live site and final product.
Video link missingOur project illuminates trends in global assassination attempts in 2019-2020. We examined the effect of countries’ politics, economics, and national security—among other factors—on the number of assassination attempts as well as the survival rate and prevention rate. The first data source we used was a dataset from Global Initiative’s “Global Assassination Monitor” that included an entry for every assassination attempted in 2019 and 2020. The dataset included information about each attempt as well as the country in which it occurred. Based on the country in which the assassination attempt occurred, we were able to cross reference the dataset with data on individual countries’ economies, political states, and quality of national security. From various research institutes such as the Heritage Foundation and the United Nations Development Programme, we extracted data on the individual countries. With the primary data source and our additional country-specific data sources, we created a master dataset. In this new Excel file, each row was a country and each column included information about the country as well as its number of assassinations, prevention rate, and survival rate among other statistics calculated in Excel based on the primary data source’s information. We then converted this Excel file to a CSV and imported it into Google Colab. Reading the data from the CSV file, we compiled the data into lists and then into a DataFrame. We used the DataFrame to create visualizations—ranging from dot plots and bar graphs to choropleth charts—using Plotly Express. From these visualizations, we were able to deduce insights about which macro factors had a sizable impact on the number of assassination attempts in each country. In addition, we were able to make conclusions about what factors yielded higher prevention and survival rates.
We rushed a little bit to find the data. Then a substantial amount of time passed before we revisited the data. Cleaning the data consisted mostly of creating data frames and querying the data which we found very helpful. We made some visualizations independently and then, during the second week, came together to make a plan. It was during the cleaning and planning that we realized we weren't sure we had enough data for the comparisons we wanted to make and began looking for other sources. After some time looking, we gave up and decided to do what we could with the data we'd initially chosen. We came up with a plan, worked on the visualizations and website together, and recorded the videos once done.
Josh and I first found a dataset that included the world’s top 100 hotels according to “Travel + Leisure”, a reputable magazine published out of New York City focused on coverage and evaluation on travel destinations across the globe. We were both entertained by the idea of figuring out which countries are the best counties to travel to right now since we are both studying abroad next semester. On a more important note, we hoped that by answering questions such as: Do hotels with more rooms tend to outperform smaller hotels? How does the modernness of the hotel/the year it was built affect the rankings or the rating? Which themes are currently preferred by luxury travelers? How does a country's quality of life correlate to its average luxury hotel rating? we could help travel agencies figure out where to send there clients. Additionally, with Covid-19 restrictions loosening, we hoped that people who are eager to travel and find a luxury getaway can gain a better understanding of the best options available, and find what may meet their preferences and styles. Along with the data from “Travel + Leisure” we also, outsourced quality of life data for the countries that all had ranked hotels. This was done in an effort to find if there is correlation between the overall quality of life that a country enjoys and the scores of their luxury get away for people hoping to travel to these countries.
Our data mostly categorized registration and census data about Florida counties, categorizing everything from demographic data to boat registry data, and by comparing these factors on scatter plots, violin graphs, and even heat maps, we were able to use the data to create a picture of each county's unique characteristics. Then, we could use those characterizations and apply them to similar counties in terms of boat ownership and learn from those trends.
The journey started in class, deciding on what to learn more about. Jake and Jake decided to explore the movie sector, particularly what makes some movies successful. Jake and Jake started searching the web for data and stumbled across three datasets that fit the project's needs. Two of these datasets were found on Github, and the other was found on Kaggle. From there, Jake Irons started building the website, while Jake Drwal started building the visualizations. After a few nights of merging DataFrames, Jake could start building the visualizations, and Jake could gain a better understanding of HTML/CSS. As a result of this hard work, Jake and Jake were able to create a beautiful website demonstrating what makes a movie successful.
Our process for turning the raw online data into useable data frames involved scouring the data site for the information that we were searching for and then manually entering that information into individual local csv files. Each visualization has its own corresponding csv file, so we created 6 csv files in total for the project. Once a csv file was created, we set about breaking it down into separate lists with for and while loops. This was done so that the information could be reunited again in a larger list, just in a different format that could be understood by the plotly express DataFrame function in order to create data frames that would serve as the foundations for each graph.
We were curious what affects a country's olympic medals. So we thought of different factors that might affect it such as the country's GDP, crime rate, health care performance and whether or not they have hosted the olympics. We gathered the data into a DataFrame table and created visualizations based on this data.
We chose to examine a CSV cataloguing speed dating, and what traits led to matches. There obviously quite a few different categories to sort through, so we had to write code to turn this csv into data frames that we could then plot and visualize.
We started by looking at raw csv, and slowly were able to clean it so we could work with it. We worked with World Cup results data as well as Individual World Cup Match data.