Frequency is the number of times $n_i$ a specific value or member of a [[Class|class]] or a [[Variable|variable]] appears in a [[Math/Probability and Statistics/Dataset|dataset]]. It's helpful to divide a dataset into a set of [[Class|classes]] to reduce the number of The relative frequency ($f_i = \frac{n_i}{n}$) is the ratio of the value or class to the total [[Count|count]] of the sample. The cumulative relative frequency ($F_i = \sum_{k \le i}f_k$) represents the proportion of a sample that has been tallied *up to* a particular value. ### Visualisation Frequency can be directly visualised using **bar graphs**, where each bar represents data categories from any levels of measurement. The bar graph is arranged mostly vertically, but may be horizontal if the names are long. Each bar should be of equal width and have even spacing. The height of the bars must at least meaningful as **interval** [[Variable|variables]]. A **Pareto** chart is a bar graph where the bars are sorted from tallest to lowest, whose height represent frequency/percentages, and provide a sorting criteria for nominal variables. A **histogram** is a bar graph that has *no gap* between each bar, typically used to represent quantitative data. A **pie chart** is a particularly useful visualisation for the relative frequency of a data set, using each wedge to represent a component of the total. ### Analysis A **pivot table** from Excel is useful for discrete data with a limit number of values. A **histogram** from Excel can be done directly, but requires parameters set on the [[Class|classes]].