Portfolio
Select a project from the list below to view, or browse through all projects by scrolling
Highlighted Projects
Customer Churn Analysis
In this report, we take a surface-level look into the customer churn of the global telecommunications company Telco. The goal of the project is to identify and visualize some key characteristics that drive churn. This information could be used to take a deeper dive into the specific areas identified or to develop predictive models.
- After transferring the data to Jupyter Lab, it was cleaned and formatted using Pandas. A few subsets of the data were created based on field attributes for easier reference later.
- Data exploration began by calculating churn rates based on each field containing categorical data
- ‘Impact’ metrics were calculated for each field. This measured the difference in churn rates of the categories in the field. This helped to get an idea of which fields influenced churn most
- I then began analyzing the continuous variables (tenure and monthly charges.) After plotting the average churn against tenure using the seaborn library, I noticed a distinct relationship. Upon further inspection, I saw that the relationship between the fields looked almost logistic. I did some regression analysis using logistic regression and the scipy library. I found that my model accounted for essentially all of the data.
- Using the plotly express library, I made some appealing and interactive charts to create a better user experience.
![](http://alexanderimurray.com/wp-content/uploads/2022/02/tenure_regression-768x384.png)
Read about the findings!
Go to ReportSkills Used: Python, Pandas, EDA, Data Visualization
Pharmaceutical Stock Analysis
In this project, we take a surface level look at stock prices compared to covid-19 cases. Three main U.S. companies who have produced vaccines were analyzed (Pfizer, Moderna, and Johnson & Johnson). We also look at the S&P 500 index as a baseline measure of the economy. The goal is to find out whether these stock prices are correlated with Covid cases and by just how much.
- The stock data was fetched using the Yahoo! Finance API.
- The data was imported into Jupyter Lab, where it was cleaned and formatted.
- I looked through the correlation matrix to compare stock prices against covid data.
- Seaborn was used to visualize some regressions between cumulative cases.
- Plotly express was then used for the more useful visualizations and subplots.
![](http://alexanderimurray.com/wp-content/uploads/2022/02/stock_prices-768x674.png)
Read about the findings!
Go to ReportSkills Used: Python, Pandas, EDA, Data Visualization
U.S. Diversity Dashboard
In this project, we look at the diversity of The United States through different metrics. We visualized racial diversity, income distribution, as well as levels of education for every state.
You can explore average income of every state, as well as breakdowns of race, income, and education for a specific state.
- The data was retrieved using the government Census API.
- The data was cleaned and formatted into tables using Pandas.
- The tables were then exported to CSV files for easy upload to Tableau.
- Once in Tableau, the data was visualized and organized into a concise Dashboard
- Additional diversity visualizations were created using Plotly Express in Python.
![](http://alexanderimurray.com/wp-content/uploads/2022/02/Economic-Racial-Diversity-in-U.S.-768x416.png)
See this highly interactive visual in action!
Try it OutSkills Used: Tableau, Python, APIs, Data Visualization
Interactive Covid-19 Dashboard
In this project, I create a U.S. Covid-19 Statistics Dashboard on Tableau using various datasets of Covid data through 2021. The top right section contains an interactive U.S. map of cases broken down by state. In the top right under the legend, there is a slider to filter out states based on the number of cases. You can also filter according to a certain date or specific states. These two filters are applied to the entire dashboard.
I started by designing and creating a database in MySQL. After formatting the data in Excel a bit, It was imported to MySQL. There I finished cleaning the data as well as creating a few views and calculated fields. Then it was on to Tableau for visualization. This being my first full tableau project, I wanted to make a variety of visuals to gain experience creating all types of visualizations.
![](http://alexanderimurray.com/wp-content/uploads/2022/01/covid-dash-768x467.png)
See this highly interactive visual in action!
Try it OutSkills Used: Tableau, SQL, Data Cleaning, Data Visualization, Database administration
Plotly Covid-19 Map
In this short project, I create an interactive map of U.S. Covid-19 data using the Python library Plotly. The map includes a dropdown menu to switch between the 4 measures included in the visualization.
![](http://alexanderimurray.com/wp-content/uploads/2022/01/PlotlyCovidMap.png)
See this interactive visual in action!
Try it OutSkills Used: Python, Data Cleaning, Data Visualization