Portfolio

Personality Generator Dashboard

In a Python file, character traits are generated using multiple frameworks from psychology and pseudoscience. Some traits are duplicates but have different percentage values representing their influence on overall personality. The traits are grouped up in a Pandas dataframe and psychology frameworks are reapplied. The traits data is normalized and delivered through Flask to express the personality in multiple radar charts using ChartJS library from Javascript. To generate a new personality, JQuery is used to request data from the backend.

Job Title data scraped from the U.S. Bureau of Labor Statistics and the DSM-V Diagnostic Codes was scraped from Psych Central thanks to beautifulsoup and regex.

Gym Data Tracking

Used Google Sheets API to automatically display gym progress. Data is grabbed from a log of data containing date, exercise name, weight, repetitions (reps), and set number. With the pandas library, the data is transformed, mimicking the pivot table I had on Google Sheets itself. Rows are grouped by date and exercise. For each exercise, a ChartJS chart is created. At the bottom of the page, a table is dynamically created with vanilla Javascript.

Analysis: Spotify Music Genres

R analysis. How can predictive data and linear regression be used to increase Spotify streams? A song can have multiple music genres associated with it. What combinations of music genres is recommended to increase it's chances to be popular? A dplyr dataframe displays what audio features (i.e. energy, danceability, tempo, etc..) are more prevalent in each music genre and how it affects the genre's popularity. Seven visualizations are created using Excel. The analysis concludes that a song with elements of Dance, Pop, and Rock has the highest likelyhood to be the most popular.

Analysis: Common Traits of Presidents of the United States

Python and Jupyter Notebook used. Presidential demographic data and historian rankings of presidents were webscrapped with beautifulsoup. Data was sourced from potus.com, wikipedia.org, britannica.com, and c-span.org. After extraction and transformation, the data is loaded in a Pandas dataframe called potusData. Nineteen visualizations are created with Seaborn which is library built on top of Matplotlib. At the bottom, an interactive Tableau dashboard explores presidential performance and sibling count.