Using the statistical programming language R, I analyzed a dataset of sales transactions for an online superstore. I performed exploratory data analysis and found a possible logistical area for improvement. Using unsupervised clustering methods I isolated the busiest times of the year and used location data to pinpoint the best locations for product distribution centers.
The data for this project, original and cleaned, is in the Data folder.
For this project, I took zip codes from the original sales data and joined the dataset with a list of latitudes and longitudes. Using those coordinates I was able to plot the data points across the U.S. and get a sense of the types of products being sold and shipped in different locations. Using different clustering methods and factors I pinpointed locations for distribution centers that made the most sense.
I performed the data cleaning and EDA in analytics_sales.R. I performed some simple linear regression and logistic regression and made a few line plots to explore the data.
The clustering was performed in the Coffey_Lukas_Final_Project_Report.Rmd file.