IE5342 DOE Assignment 2

Analyzed 10 seasons (2014–2024) of NBA schedule data to measure the impact of travel and game density on team performance. Built interactive R visualizations (ggplotly, geosphere) to highlight back-to-backs, 4-in-6 stretches, and excessive travel periods. Designed a regression model to estimate wins gained or lost due to schedule factors, identifying the most advantaged and disadvantaged teams. Delivered a professional brief comparing Oklahoma City’s 2024–25 draft schedule with Denver’s, outlining key stretches and recovery strategies.

The accompanying R code generates a visualization of the Expected Value and Median as functions of p, using ggplot2 to plot both curves on the same graph. The results show that as p approaches 1, both statistics converge to 1 (success almost always occurs immediately), while for small p, the Expected Value grows rapidly like 1/p while the Median increases more moderately, reflecting the skewness of the distribution. Together, the blog and code provide both a mathematical derivation and a graphical interpretation that clarify how the mean and median behave under different probabilities.

um exemplo atualizado a partir de https://rpubs.com/Selcuk/1088675

The sample size is n=1599. There are no missing data points. There are outliers for residual sugar and chlorides. There were some bottles where sugar was high for red wine compared to other wine types, and therefore the data is right skewed. This is also true for chlorides, because there are a few high values, but most are in the same range. The chloride levels for wine vary by region, and high levels can make the wine taste too salty. The median is a better measure for these variables than the mean. Having taken into consideration, I do not have problems with the data quality. I’ve created histograms and boxplots to summarize the data for each variable. The histograms show the distributions. The alcohol variable is almost symmetrical. The quality scores are left skewed, but not much. More wines scored above the mean than below.

Welcome to the Library Event Attendance and Performance Analysis! This report was designed to provide a clear, data-driven look into our library's adult event programming. The goal is to help us better understand what makes an event successful, so we can make more informed decisions about future programming. We began by taking our raw event data and cleaning it up. To focus on the adult audience, we filtered out all events intended for children or youth, as well as any events that were cancelled, had zero attendees, or lacked a valid date or registration count. We also removed any events that started before 9 a.m. This process left us with a clean dataset of our true adult programs. The core of our analysis is a series of visual heatmaps that show us when people are most likely to register for events. By looking at a simple grid of days and times, we can immediately spot patterns and trends. We created a large, overall heatmap to see a general picture of attendance across all adult programs. To dig deeper, we also generated subject-specific and room-specific heatmaps. This allows us to answer questions like, "Do our technology workshops perform better on a certain day of the week?" or "Is a room more popular in the morning or evening?" We then supported these visuals with summary tables that highlight the top-performing subjects and rooms based on the number of events and total registrations, giving us a quick reference for our most successful programs and locations. This analysis provides a new way to look at our event data, offering clear insights into when and where our adult programs are most successful. By leveraging this information, we can strategically plan our future events to better meet the needs of our community and maximize attendance.