For a set of grouped data, the frequency density of a class is defined by \[\text{frequency density}=\frac{\text{frequency}}{\text{class width}}.\] It gives the frequency per unit for the data in this class, where the unit is the unit of measurement of the data. This allows for a meaningful comparison of different classes where the class widths may not be equal.
When drawing a histogram, the axes are the measurement and the frequency density:
A related idea is the relative frequency density. This is the relative frequency of the item divided by its class width, or alternatively, the frequency density divided by the total number of data items: \[\text{relative frequency density}=\frac{\text{relative frequency}}{\text{class width}}=\frac{\text{frequency density}}{\text{total number of data}}\] If a histogram is drawn with relative frequency density instead of frequency density, then its total area will be \(1\).
