Graduated symbols are used to show a quantitative difference between mapped features by varying the size of symbols. Data is classified into ranges that are each then assigned a symbol size to represent the range. For instance, if your classification scheme has four classes, four symbol sizes are assigned. The color of the symbols stays the same.
Symbol size is an effective way to represent differences in magnitude of a phenomenon, because larger symbols are naturally associated with a greater amount of something. Using graduated symbols gives you a good degree of control over the size of each symbol, because they are not related directly to data values as they are with proportional symbols. This means you can design a set of symbols that have sufficient variation in the size that represents each class of data to make them distinguishable from one another.
Graduated symbols can be based on an attribute field in the dataset, or you can write an Arcade expression on which to generate numeric values to symbolize.
When graduated symbology is based on a single field, the symbols are drawn in a sorted order in which the larger features draw first and the smaller features draw above them. When the symbology is based on an expression, this sorting does not occur, and some smaller symbols may be obscured by larger ones.
The Primary symbology tab has two subtabs to establish graduated symbol symbology:
To draw a layer with graduated symbols, follow these steps:
and click Graduated Symbols 
.
In the Symbology pane, on the Primary symbology tab , choose the numeric field for the data to be mapped, or write an expression.
To use an expression, click the expression button 
to open the Expression Builder dialog box. Write an expression and click Verify 
to validate it. Note that although an expression is valid, it may not return a valid numeric value. You can filter 
the Expression Builder dialog box to show only numeric fields to help prevent this.
This can be an effective way to generate a smaller range of values if the dataset includes significant outliers. Normalization is available only when the graduated symbology is based on a field. If it is symbolized on an expression, the Normalization field is unavailable.
Click the Template symbol to open the Format Symbol pane to modify the symbol or choose a different symbol. The template symbol defines the appearance of the symbols that are sized for each class. For point and line layers, the template symbol is a point or a line symbol, respectively. In the case of polygon layers, the template symbol is a point symbol that is sized for each symbol class. Additionally, a background polygon symbol is defined to draw the polygon features themselves. Click the Background symbol to change it.
This option does not apply when the layer is in the 3D Layers category of a scene. In this case, the point symbols always honor the layer draw order.
The Primary symbology tab of the Symbology pane includes the Classes and the Histogram subtabs, both of which also have a More drop-down menu that contains additional commands.
The Classes tab includes an interactive table that lists the graduated symbol classes with their corresponding symbols and headings, as defined by the symbology parameters in the upper portion of the pane.
In the graduated symbol classes table, you can do the following:
Change the symbol for a single class
Click a symbol in the Symbol column of the table to open the Format Symbol pane and modify the symbol or choose a different symbol.
You can modify common properties of more than one symbol at a time. Highlight the symbol classes in the table, right-click, and click Format symbol .
Click the back button to return to the Symbology pane after.
Edit the label of a class
Right-click a non-symbol cell in the table and click Edit label , or double-click a cell in the Label column, and type a new value.
Edit the upper value of a class
Right-click a non-symbol cell in the table, or double-click a cell in the Upper value column and type a new value.
Edit the description of a symbol class.
First, ensure that the Description column is visible in the table. If not, from the More menu, check Show description to add the column to the table.
To modify the description, right-click any non-symbol cell in the table, or double-click a cell in the Description column, to open the Edit Description dialog box to enter or modify the description of the symbol class.
Symbol class descriptions are included in layout legends, depending on the legend settings.
Add additional symbol classes
Use the Classes drop-down menu at the top of the pane to control the number of symbol classes. If the classification method is set to Manual Interval, you add additional symbol classes manually within the table itself. Right-click a non-symbol cell in the table and click Insert break above 
or Insert break below 
to position the new symbol class.
Combine symbol classes
To combine two or more symbol classes, highlight the rows in the table, right-click, and click Combine breaks .
Remove symbol classes
Use the Classes drop-down menu at the top of the pane to control the number of symbol classes. To removed a symbol class, right-click a non-symbol cell in the table and click Remove .
From the More drop-down menu, you can perform the following tasks:
Show values that are out of range
Values can end up out of range in the classification scheme when they are added after the classification was established, when they fall within removed classes, or when they contain null values. You can show values that are out of range in the table and assign a symbol to them to show them on the map.
Click the More menu and check Show values out of range to add an symbol class.
Show excluded values
To specify which features are to be excluded from classification, on the Advanced symbology options tab in the Symbology pane, enter a clause under the Data Exclusion heading.
On the More menu, check Show excluded values to add an symbol class.
Show the feature count of each symbol class
Click the More menu and check Show count to add the Count column to the table.
If excluded values or out of range values are not shown in the legend, their respective total counts are listed at the bottom of the Symbology pane.
To update the count of features in each classification range (and values out of range), right-click the Count column in the table and choose Refresh count .
Click the More menu and check Show count to add the Statistics section below the table.
Statistics include the overall feature count, the minimum and maximum values of the data, and the mean and standard deviation of the values.
Show class ranges using feature value
Click the More menu and check Use feature values in labels to use only values that are present in the field as boundaries in the symbol class labels. This can help you identify significant gaps in the data.
If the source data has changed or updated, click More , and click Refresh values to update the class ranges using the current settings at the top of the pane.
Include a description of each symbol class.
Click the More menu and check Show description to add the Description column to the table.
Double-click a cell in this column to edit or update the description on the Edit Description dialog box and click Update to apply the description.
The description appears with its symbol class in layout legends, depending on the legend settings.
Format all symbols
Click the More menu and click Format all symbols to modify the common properties of all the symbols in all the symbol classes. Verify the change to each symbol visually in the symbol preview.
Click the back button to return to the Symbology pane after.
Regenerate all symbols
Click the More menu and click Regenerate all symbols to regenerate all symbols to reset each symbol class to its default symbol based on current symbology parameters.
You may want to do this to discard individual symbol edits and return all symbol class symbols to a default state.
Reverse symbol order
Click the More menu and click Reverse symbol order .
Reversing the symbol order retains the symbol class ranges but flips the order of the symbols assigned to each class, so that the symbol that was initially assigned to the lowest symbol class is now assigned to the highest, and vice versa.
Click the More menu and click Reverse values .
Reversing the values flips the order of the symbol classes, leaving the symbols themselves in the original order, so that the highest class is not at the top of the table, assigned to the symbol that was previously used with the lowest class, and vice versa.
From the Advanced symbology options tab , you can do the following:
Expand the Format labels heading to change the rounding, alignment, and other formatting of the symbol class labels.
See the Format numbers in legend labels topic for detailed instructions.
Adjust the sample size
Expand Sample size to change the maximum sample size used to calculate the symbol class ranges and statistics. The default sample size is 10,000 records.
Limiting the sample size can improve performance but may inadvertently omit important outliers in the dataset. Generally, the larger the dataset, the larger the sample size you should use.
To exclude data values from the classification, expand Data exclusion to define the query.
To visualize excluded values on a map, on the More menu, check Show excluded values to add an symbol class with a unique symbol.
Set up feature level masking
Expand the Feature level masking heading to set up masking per feature
Override the default feature drawing order
To draw features in a different order—usually to avoid smaller symbols getting obscured by larger ones—expand the Feature drawing order heading. Choose one or more fields, and set a sorting method. See Set the feature drawing order for detailed instructions.
The histogram provides a visual tool for editing the classes and understanding how the data is represented by different classification methods. Access it by clicking the Histogram tab on the Primary symbology tab .
Making dynamic edits to the histogram switches the classification method to Manual .
In addition to specifying the magnitude of features with graduated symbology, you can also symbolize additional attributes by varying the transparency, rotation, and color of the graduated symbols. These variations are sometimes called visual variables. While all of these visual variables can be applied simultaneously, be aware that too many variations make the layer difficult to interpret. It is recommended that you apply secondary visual variable symbology sparingly.
