Description
At a railway station, an operations manager studies a report showing that the average train delay has fallen from twelve minutes to nine. The number looks promising until a colleague asks a quieter question: have fewer passengers been delayed, or have a small number of extremely late trains simply disappeared from the sample? Averages can be correct and still conceal what people actually experience. Descriptive analytics begins at precisely this point, when a plausible summary is treated as the start of an inquiry rather than its conclusion. The discipline is sometimes introduced as the simplest branch of analytics: descriptive work reports what happened, predictive work anticipates what may happen, and prescriptive work recommends action. That description is useful but incomplete. In serious data science, sound description is often the most difficult intellectual work. It requires decisions about the unit of observation, the population of interest, missing records, denominators, visual scales, comparisons across groups, and the consequences of reducing a complicated distribution to a single number. This book develops descriptive analytics as a practical craft. Mathematics appears where it clarifies judgment; code appears where it makes a method reproducible. The examples move across commerce, public services, health systems, software products, and manufacturing. The intended reader may be a university student encountering exploratory data analysis for the first time, an analyst preparing operational reports, a manager reading dashboards, or a data scientist who wants stronger foundations before building models. Familiarity with spreadsheets is sufficient for the opening chapters. Later sections introduce SQL and Python, but the ideas remain understandable without programming. Readers can follow the chapters in order or use the applied sections as a reference.