If you style your layer using color or size, you have the option of classifying your data—that is, dividing it into classes, or groups—and defining the ranges and breaks for the classes. Depending on how much data you have in your layer, you can also choose the number of classes: 1 through 10. The more data you have, the more classes you can have. The way in which you define the class ranges and breaks—the high and low values that bracket each class—determines which locations fall into each class and what the layer looks like. By changing the classes, you can create very different-looking maps. Generally, the goal is to make sure locations with similar values are in the same class.
Natural breaks classes are based on natural groupings inherent in the data. Class breaks that group similar values and maximize the differences between classes—for example, tree height in a national forest—are identified. The locations are divided into classes with boundaries that are set where there are relatively big differences in the data values.
Because natural breaks classification places clustered values in the same class, this method is good for mapping data values that are not evenly distributed.
Equal interval divides the range of attribute values into subranges of equal size. With this classification method, you specify the number of classes, and ArcGIS for Power BI automatically determines how to divide the data. For example, if you specify three classes for a field with values ranging from 0 to 300, ArcGIS for Power BI creates three classes with ranges of 0–100, 101–200, and 201–300.
The equal interval classification is best applied to familiar data ranges, such as percentages and temperature. This method emphasizes the amount of an attribute value relative to other values. For example, it could show that a store is part of a group of stores that make up the top one-third of all sales.
With quantile classification, each class contains an equal number of locations, for example, 10 per class or 20 per class. There are no empty classes or classes with too few or too many values. Quantile classification is well suited to linearly (evenly) distributed data. If you need to have the same number of locations or values in each class, use quantile classification.
Because locations are grouped in equal numbers in each class, the resulting map can often be misleading. Similar locations can be placed in adjacent classes, or locations with widely different values can be put in the same class. You can minimize this distortion by increasing the number of classes.
Standard deviation classification shows you how much a location's attribute value varies from the mean. By emphasizing values above and below the mean, standard deviation classification helps show which locations are above or below an average value. Use this classification method when it is important to know how values relate to the mean, such as population density in a given area, or comparing foreclosure rates across the country. For greater detail in your map, you can change the class size from 1 standard deviation to 0.5 standard deviation.
If you want to define your own classes, you can manually add class breaks and set class ranges that are appropriate for your data. Alternatively, you can start with one of the standard classifications and make adjustments as needed. There may already be certain standards or guidelines for mapping your data-for example, an agency might use standard classes or breaks for all maps, such as the Fujita scale (F-scale) used to classify tornado strength. Place the breaks where you want or need them.