Earlier today, I used for the first time the Power BI REST API to access some reports in my company’s workspace. This API can be used with Windows PowerShell. After spending a couple hours familiarizing myself with both PowerShell and the API, I think I’ve finally got the hang of it.

The applications for hosting an R Shiny web application on the cloud are huge, and learning to work with cloud services in general is critical for data analysts, engineers, and scientists. Dashboards that rely on cloud-hosted data do not need to be constantly redeployed; as long they point to the right file on S3 (or Redshift or RDS database), your Shiny application will use the most recently available data. This has the potential to reduce overhead and provide near-real time access to information for your clients.

This collaborative project between myself and Priyanka Verma seeks to gain insights from particulate matter air pollutant trends in the New York Metropolitan Area. Specifically, we look at particulate matter that is less than 2.5 micrometers (PM2.5), and only data obtained during winter months, due to the higher level of PM2.5 during that season. We developed this in a way that should be easy to reproduce and understand.

Last fall semester, I was able to take an independent study where I expanded my knowledge of SQL beyond its traditional use of querying relational databases for text and numeric values. Instead, I began using it as a standalone GIS, using it to store and query geometries (points, lines, and polygons).

This semester, my peers Aaron Eubank, Bryce Stouffer, Samuel Watson, Wei Hong Loh, and I took on a directed study under our professor Dr. Lyndon Estes on the implementation and use of PostGIS. While I had prior experience with SQL Server and some of its spatial functions, I had never used SQL to perform a full range of geoprocessing tasks. I found my workflows to be more reproducible than using ArcGIS and QGIS, and found processing speeds to be faster than R’s {sf} package.

Data fusion is a method of combining different modes of data (whose degrees of heterogeneity can vary) into a single output. While the purpose of data fusion varies from project to project, the unifying goal is to extract information from the fused result that could not be easily obtained from either source alone. Bear in mind the quote, “The whole is greater than the sum of its parts.”