
Smart Mapping: Counts and Amounts (color)
Use color to reveal significant patterns in your numerical or ranked data
A layer of points, lines or polygons lets anyone see the location of these features relative to one another. Very often, these features on the map have numeric attributes associated with them. How can you discover and visualize meaningful patterns from these numerical attributes? How do you ensure the colors relate to meaningful numbers in the real world, so that the patterns provide useful comparisons?
Color on a map is used to help readers immediately see where places are similar and where they are different. A logical use of color on a map does more: it also encodes how big or small those differences are, something we can see at a glance.
The Counts & Amounts (color) drawing style in ArcGIS Online and ArcGIS Enterprise applies color to your layer's features, based on a numerical attribute you choose. You can vary the features' colors on the map when you want to compare them based on an attribute containing an average, mean, median, rate, ratio, percent, index or other normalized data.
Smart mapping helps you create beautiful and informative maps, quickly, using color. It sets the cartography of your layer based on the significant values within your data. Smart mapping lets anyone quickly discover patterns from attributes in their data and make meaningful maps from them.
This guide introduces easy ways to use color to discover and emphasize interesting stories in your numerical data. If your data has a numeric attribute field, it’s easy to color each feature on the map based on that field's value. Within five or six mouse clicks you will be looking at a first draft of your map.
Production Expected - Hydropower Plans and Dams in Switzerland
Daily Traffic on Florida Roads
Smart mapping helps you create beautiful and informative maps, quickly. The map's defaults jump start your exploration of the data, helping to discover meaningful values that shape the story of the data. Let's get started.
Two Steps
There are two basic steps to using this drawing style on a numerical attribute you have selected.
- Explore the numbers first. Set the breaks on meaningful values that bring out a story in the data.
- Choose colors last that emphasize the important part of the data, and de-emphasize the rest.
Smart mapping helps you follow this simple pattern. Most people choose colors first, because they get immediate visual feedback on their choice. They spend less time thinking critically about adjusting the breaks which control how the colors are applied. A perfectly constructed color ramp can only reveal a story if you explore your data first.
You can use this ACS Educational Attainment Variables - Boundaries layer to follow along.
The layer is already mapping a topic, which you will change shortly. Zoom in until the map looks similar to the one at right, and select the County layer.
ArcGIS Online and Enterprise users who select a layer can choose Styles from the configuration toolbar.
Choose it and scroll through the drawing styles to select Location (single symbol).
Smart mapping reacts as soon as you choose attribute(s) to map. Click + Field to see a list of attribute fields.
Just under Add fields, you can search for the field you wish to map. Today, you want to map what percent of the population have a bachelor's degree or higher, so you search for "bach" and scroll through the results until you find the field shown at right.
Before you add it, you can investigate its metadata. Click the Information button next to the field name.
The field information pane shows useful information about the field type, field alias and field description (if provided by the layer's owner). This layer's owner provided everything, including the formula used to calculate this field, so you don't have to go find a metadata document to figure out whether this is the field you want to use.
Other information shown at right is automatically provided, including a sample value and number of records in the layer. You don't have to go to the layer's attribute table to get this information.
Scroll down in the field information pane to see more details. For numeric attributes, you see the number of decimal places in use, average value, standard deviation, minimum and maximum values, and more.
Close the field information pane and hit the "Add" button to add this field to the map.
As soon as you hit "Add," smart mapping looks at the characteristics of the field you chose, and suggests an initial drawing style based on what your data contains.
In this example, because your field is numeric and is not an integer), the Counts & Amounts (color) style is suggested. This is a commonly used drawing style for a percent, ratio, average, mean, or any other normalized attribute.
The style it suggests is only a starting point, to get you started thinking about how you might map this data. Additional drawing styles are one click away, so it is easy to try out different options quickly. Since every style has its biases, strengths and weaknesses, it's important that you can explore each option easily.
Note how the legend is immediately useful. It shows the name of your layer, and the field alias of the attribute you chose to map. The suggested colors are applied using an unclassed color ramp whose breaks are set to one standard deviation above and below the mean of your attribute. We will see how to adjust these later.
A consideration for this drawing style is that it rewards bigger polygons with more color. Bigger polygons get more pixels of color than smaller polygons. In the map at right, there are counties whose area is so small that they don't merit a single pixel on this map. It's a common pattern in census geographies of many nations. Smaller states, provinces, enumeration areas, tracts, postal codes, etc. struggle to be seen on maps like this. The Color and Size style ( discussed here ) helps level the playing field, visually, for polygons of all sizes.
The basemap is pretty colorful and all its colors will compete with the thematic layer's colors for attention. The basemap's labels are also hidden by the thematic layer. Ten years ago, you might simply set transparency on your thematic layer as a compromise, but today there are far better options.
Watch your map improve significantly by simply switching the basemap. Click the "Basemaps" button on the left toolbar to see what basemaps are available to you. Choose the "Human Geography Map" or the "Light Gray Canvas" basemap.
Why? Each of these basemaps have what you want when making a thematic map:
- Each has almost no color in them, instead using shades of grey, black and white. When a basemap has lots of color in it, those colors compete with your thematic layer for attention.
- These basemaps put city labels and other features like water on a separate layer within the basemap, appearing on top of the thematic layer's content. This means city labels appear on top, where they can be seen. The Human Geography basemap puts water on top as well, so lakes and rivers appear on top of your thematic layer content.
You can easily try out other basemaps with one click. Hover the mouse over each to see its description.
If you don't see a neutral basemap in your list, talk to your organization's administrator to request access. They can edit the list of basemaps that appear in the pane shown at right.
You can find "Human Geography" basemap and hundreds of other basemaps in Living Atlas, at the bottom of the basemaps pane.
1) Explore the Numbers First
Once you have a neutral basemap underneath your thematic layer, you can now focus on which story you want to tell about the data. Before you tune this map into a particular story, let's evaluate what story it is already telling.
In the Counts & Amounts (color) style, click Style options if you don't see the histogram yet.
Smart mapping defaults to a "High to low" theme, which emphasizes the high values with a stronger color, and de-emphasizes the low values with a weaker color. You can reverse the color ramp to put emphasis on the low values if desired. The values in the middle near the mean of the data are not an important part of the story visually.
The histogram at far right tells the data's story. We see the average of the county layer's data is 22, indicating 22% of the adult population have a Bachelor's Degree or Higher in an "average" county. Areas lower than average trend toward the yellow, and areas higher than average trend toward darker blue.
In this example, all the values between 13 and 31 are given a color proportional to their value. Within those two values, this drawing style lets readers see even small variations in neighboring areas.
The values 13 and 31 are meaningful values to use as breaks in this drawing style. By default, smart mapping sets these breaks to one standard deviation above and below the mean. This ensures that up to 68% of the data is being shaded proportionally (if your data has a normal distribution).
The more extreme values are not the story this style emphasizes. Areas below one standard deviation from the mean (13% in this example) earn the solid yellow color in the color ramp. Areas above one standard deviation from the mean (31% in this example) earn the darkest blue color.
You can drag either handle in the histogram to any value to adjust the map's story. Equally important, so can anyone else who uses your map. For example, if someone wants to focus the map on the range 25% to 50%, they simply drag the handles to those values or double-click on their current values to type in the new breaks instead.
The Counts & Amounts (color) drawing style applies a color to each feature, based on the value of the attribute you choose to map.
There are six themes to choose from within this style. Each emphasizes different parts of the data.
You can explore these options so that you can choose which part of the data is interesting. Sometimes you want to emphasize the higher values in the data. Other times, you wish to emphasize the lower values in the data.
Usually, though, you simply want to explore your data to discover its story. These six themes help you do that. When all it takes is a single click to try something out, it's worth the effort.
Change the theme from "High to low" theme to the "Above and below" theme so that we can "center" the map's colors around a central value, like the average of the data (the default) or zero or a value you know is important, like a national average or a policy goal.
Instantly, the map changes and now tells the story a bit differently than before.
The "Above and below" theme puts equal emphasis on both ends of the data and its histogram. The map at right uses the exact same breaks (31% and 13%) as before, but instead of using a color ramp with a strong color and a weak color, the "Above and Below" theme uses two strong colors, with a weak or neutral color in between.
This is known as a diverging color ramp, because the two strong colors diverge from a common middle color.
As a result, readers of this map can easily see areas with a high, or low, percentage of Bachelor's Degree or Higher. Both stories are told in this map because both high and low percentages are emphasized.
The middle handle on this histogram controls the story. By default, smart mapping sets the middle handle at the mean of the data. This is always a reasonable place to start, because going into a map you may not know what is a significant value around which to center this diverging color ramp.
Another option is to set the middle handle based on a national average, a state/province average, or other meaningful number.
You can also apply a middle value based on a goal or target. If this data shows that the "average" county has 22% with Bachelor's Degree or Higher, and someone proposes a program to ensure every county reaches a goal of 20%, you have a simple adjustment to make: set the middle handle to 20%, and the map now tells that story.
Color ramps have a powerful effect on your map's story. At this stage, choose whatever color ramp you want. You are only trying to understand the data at this point. For example, switch to this purple and orange color ramp, which diverges from a neutral white color.
This closeup of the histogram next to the color ramp tells you what's happening on the map, with regard to color. Using the "Above and Below" theme, you have "anchored" the map around a central value of 22%, which is the mean of all counties in the U.S. All the variation in the color ramp is expended where it's needed most: on the features between 13% and 31%. Those features comprise about 68% of this particular data set. Extreme values lower than 13% have the same deep orange color, and values greater than 31% all have the same dark purple color.
There are other methods to apply color to the features on the map. Just below the histogram, turn on the "Classify data" option to explore a few other options. The map immediately responds, showing you a Natural Breaks classification using four classes.
Instead of applying your color ramp proportionally based on each feature's value, classifications group all features into a class, based on a method. Color is applied based on each feature's membership in a class. This map's story is visually simpler than an unclassed map, and there are times you will want to tell a simpler story. The tradeoff is that all the subtle differences are gone. See this blog for more discussion on the tradeoffs , including an example using the famous painting of the Mona Lisa.
The Natural Breaks classification method lets you choose how many classes you want, and does its best to find "natural" groupings given that number of classes.
From ArcGIS Online help :
"Natural breaks (also known as Jenks Optimal) classes are based on natural groupings inherent in the data. Class breaks that best group similar values and that maximize the differences between classes—for example, tree height in a national forest—are identified. The features are divided into classes whose boundaries 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."
With four classes, the diverging color ramp does not apply a "middle" color of white to the middle class.
Pro tip: you can edit the classes' labels to communicate the meaning of each class as well as its range of values. See below. This is a critical step in translating numbers into meaning.
Best practice: translate those numbers so that readers do not have to invent their own interpretations.
You can adjust the map to use five classes, or more (up to 10 classes). With an odd-number of classes, the diverging color ramp applies a "middle" color of white to the middle class.
Note how, in all these maps, the outlines of the polygons are very subtle. Strong, dark outlines destroy any visual patterns at work in your data. That is why smart mapping defaults to light, subtle outlines.
Look at the histogram. It tells you exactly how the data is grouped within each class. Each classification method applies a different set of "slices" to the histogram...and to your data. Natural Breaks analyzes the data itself to find "natural" breaks by which to slice the data.
Let's look at some additional examples using this same data.
Change the classification to "Equal Interval" and note how different the map looks. There are almost no dark purple areas on the map. How can the map tell such a different story from the same data?
The histogram tells why. The intervals in this classification method are determined by the difference between the minimum and maximum values in your data, divided by the number of classes you choose. This data is slightly skewed, so only a few counties get that dark purple color.
From ArcGIS Online help :
"Equal interval classification divides the range of attribute values into subranges of equal size. With this classification method, you specify the number of intervals (or subranges), and the data is divided automatically. For example, if you specify three classes for an attribute field whose values range from 0 to 300, three classes with ranges of 0–100, 101–200, and 201–300 are created.
The map looks so different than the previous map because equal interval method simply slices the map into the number of evenly sized classes you tell it to. It's like dividing a pizza into exact, evenly sized pieces.
You can see in the histogram that the intervals are all of equal size. You can also see from the histogram that the vast majority of features will be orange in color.
Next, change the map to the "Quantile" method. This method slices the data such that each class has the same number of features. From ArcGIS Online help :
"With quantile classification, each class contains an equal number of features—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 features or values in each class, use quantile classification.
Because features are grouped in equal numbers in each class, the resulting map can often be misleading. Similar features can be placed in adjacent classes, or features with widely different values can be put in the same class. You can minimize this distortion by increasing the number of classes."
Note also how the mean of the data may or may not end up inside the middle class.
The "Standard deviation" classification method centers the color ramp around the mean, and sets the width of each interval to be equivalent to one standard deviation. From ArcGIS Online help :
"Standard deviation classification shows you how much a feature's attribute value varies from the mean. By emphasizing values above the mean and below the mean, standard deviation classification helps show which features are above or below an average value. For greater detail in your map, you can change the class size from 1 standard deviation to .5 standard deviation."
Edit the legend in the right side pane if needed.
By now you are starting to think "which way is best?" when it comes to applying color to your thematic map data. "Best" depends on your purpose, so when applying color, it's important to see how powerful your choices are.
The statistics of the data are informative, and there are times you will need to present the data from a statistical perspective.
Usually, your map will become much more valuable as you incorporate real-world numbers, like a national average or a goal.
The one thing you never want to do is to accept the defaults and pass that off as a final product. Automated cartography is not a thing - maps need you in the loop, finding meaningful values to apply to the map's legend. This creates meaning, which you accentuate through your choice of color ramps. It's the intersection of the numbers' meaning and color that bring this style of map to life.
In the map at right, the breaks have been set manually in 11% intervals (11% is half of the mean) to see how the results look.
Switch "Classify data" off to return to an unclassed (or, proportional) method, where every value between the two handles receives a color in proportion to that value.
Zoom into a state to assess how the map looks at scales people are likely to view your map. In this case, county level data is often viewed by state, since people want to know how counties fare on the topic being mapped.
In the highlighted area, it is easy to see the slight variations among these purple counties' Percent with Bachelor's Degree or Higher. They are all above the 22% average.
When classified, those subtle differences disappear. A county with 30% is shaded white, while 31% nets a light purple.
The maximum values in this data set "pull" the purple color further away from the center of the data. In fact, a county with 22% (the average of the data) would be in the orange category.
Always assess how the breaks relate to the histogram, to see how color is being applied.
2) Refine Colors Last
Most people believe they have to get the colors right immediately, but as you have seen above, it's more important to work with the breaks and histogram first. The breaks you choose determine your story. Color can only accentuate, or hide, that story. Smart mapping helps you choose colors that suit the story you want to tell.
For most data and topics, "Above and Below" theme inspires the most useful conversations. It forces you to think critically about how to characterize the map's topic. "Bachelor's Degree or Higher is higher than average in the purple areas, and lower than average in the orange areas." That is the story this style helps you tell.
Change your layer's style to the "High to low" theme to see how color works when you want to emphasize just one part of the the data (the high, or the low) and tell that story with color.
In this example, you see that the "High to low" theme is applied, and a simple color ramp is in use. The color ramp contains shades of one color: blue. The colors range from a very light blue to a saturated, dark blue. Areas with darker color get the attention, and areas with the lighter color are de-emphasized.
You can say or write that "Darker blue areas have more college graduates than average, while light blue areas have fewer." The entire map is blue, so you have to qualify which blue you are referring to, because to a reader, everything is blue. A monochromatic color ramp emphasizes only one end of the histogram.
You owe it to yourself and your map's future readers to try out different color ramps. Click on the current color ramp to browse these categories, such as "Reds and yellows" and "Colorblind friendly." See this blog for full details on color ramps available.
Many color ramps have a two-color sequence. In this example, you see that the ramp goes from yellow to dark green. The colors range from a light yellow to a light green and on to a saturated, dark green. Areas with darker color get the attention but areas with the yellow color are more visible than in the previous map.
You can say or write that "Green areas have more college graduates than average. Dark green areas have the highest concentrations in the country, while yellow areas have the fewest." This kind of color ramp still emphasizes the higher values in the data, but the low end "stands up" from the neutral basemap a bit more. There are two colors visible, but one of them is dominant.
The two-color color ramp in a Natural Breaks classification.
The two-color color ramp in an Equal Interval classification.
The two-color color ramp in a Quantile classification.
The two-color color ramp in a Standard Deviation classification.
The two-color color ramp in a Manual Breaks classification where each interval is set to 11% wide.
And, back to the smart mapping default: unclassed (or, proportional) color.
"Above and Below" theme, for comparison with the previous "High to low" theme.
You have seen a lot about the two most useful themes. Each of these other themes are worth trying out. Some will show immediate promise with your data, while others may not. That's OK - it takes only one click to get them started and get acquainted with these themes.
The "Above" theme does what its name implies: it focuses the map on values above a threshold. It defaults to the average of the data, shown in the histogram here as 22% with a Bachelor's Degree or Higher. All values below that threshold are shown in the same pale blue color - they are deliberately de-emphasized because you want the reader to focus on just one part of the data.
You can move that handle anywhere you need to make this map tell the story based on a different threshold.
The "Below" theme does what its name implies: it focuses the map on values below a threshold. It defaults to the average of the data, shown in the histogram here as 22% with a Bachelor's Degree or Higher. All values above that threshold are shown in the same pale blue color - they are deliberately de-emphasized because you want the reader to focus on just one part of the data.
You can move that handle anywhere you need to make this map tell the story based on a different threshold.
The "Centered on" theme focuses the map on values near a target value or a central value. By default, it applies full color to any values near the average of the data, shown in the histogram here as 22% with a Bachelor's Degree or Higher. As values get further away, transparency increases until reaching 1 standard deviation above or below the mean, beyond which all values are 100% transparent. They are deliberately de-emphasized because you want the reader to focus on just one part of the data: values at or near your central threshold.
The map at right is showing counties which are pretty close to a central value of 22%.
The "Extremes" theme focuses the map on extreme values in your data. By default, it applies full color to any values beyond 1 standard deviation above or below the mean. Values nearer to the central value (here, 21.9%) are 100% transparent. They are deliberately de-emphasized because you want the reader to focus on just one part of the data: extreme values beyond one standard deviation (or whatever thresholds you apply).
What Maps Will You Make?
You have seen how smart mapping in Map Viewer works alongside you, helping you focus on your topic and ask questions of your data. Choose an attribute, explore the drawing styles and themes within each style, adjust your breaks on the color ramp to use meaningful numbers, and lastly choose a color ramps that accentuates the data's story you wish to tell. Don't know which story to tell yet? Choose the "Above and Below" theme to prompt some thinking about what a meaningful "normal" value, or a goal, might be.
In just a few steps, you applied color to your layer's numerical data using the Counts & Amounts (color) drawing style. Coloring a polygon, line or point is appropriate when you choose to map a normalized number such as an average, mean, median, rate, ratio, percent, or index. If you want to map a total or count, use Counts & Amounts (size) drawing style, or Color and Size drawing style.
You saw many ways to explore your data using themes like "High to Low" and "Above and Below", as well as classification methods like "Natural Breaks" and "Equal Interval." Each has the potential to reveal a story in your data, within a few clicks.
With smart mapping doing the heavy lifting by taking the guesswork out of setting your map properties, you can freely explore new patterns and stories you did not know were present in the data.
Try it out today, and when ready share your maps with the hashtag #smartmapping and #ArcGISOnline.