GEO Tutorial: Unlock a Hidden Potential of One-Line Expressions in QGIS

Unlock a Hidden Potential of One-Line Expressions in QGIS

Required Resource

• QGIS (tutorial developed using QGIS 3.28 LTR; any 3.x version is compatible)

Featured Data Source

Overview

This tutorial aims to teach you how to compare values from multiple columns in QGIS. The main idea is to illustrate the integration of a comparison statement into a Field Calculator expression. You’ll discover how to assess similarities between individual row values across several attribute columns. The presented workflow involves creating a new field and a comparison expression to populate the attribute cells with integer values representing true (1) or false (0) conditions. If the values across the evaluated columns are the same, the returned attribute value is 1; otherwise, it is 0. You don’t need any programming background to complete this tutorial, just a basic familiarity with QGIS.

To spark your interest, imagine yourself in the following situation: The Mississippi Wetland Association has contracted your GIS consulting company to map wetland changes in coastal Mississippi counties. You were given a point file representing Woody Wetlands, extracted from the 2001 National Land Cover Database (NLCD). Out of an extensive request list, you have decided to first tackle a change detection assessment. This tutorial will guide you through the required steps to get this task promptly off your list.

Step 1. Download, Open, and Review the Data

In this step, you will download the tutorial data, open the shapefile, and review its attribute table.

• Download the GEO_Tutorial data using the link listed in the Feature Data Source section. Save the zipped file to a location on your local drive that you can easily find and unzip. Remember, there should be no spaces or special characters in the name of the folder where you save the tutorial data.
• Create a new QGIS project and add your favorite basemap layer to provide a visual context. For example, you can use the Browser panel and select the OpenStreetMap available under the XYZ tiles.

• Insert the Woody_Wetlands_2001 shapefile from the GEO_Tutorial folder you unzipped in step 1A.
• Zoom in to the point layer and ensure its correct position. The locations shown on the map cover the extent of the coastal zone in Mississippi and represent the Woody Wetlands sites in 2001.
• Right-click on the Woody_Wetlands_2001 layer. Select the option to Open Attribute Table. Review the existing columns and check the number of features. Confirm that the table has three NLCD-related attribute fields and 2,000 records. Note that all attributes under the NLCD_2001 column represent a single land cover class: Woody Wetlands. Most attributes stored in the other two columns similarly represent the Woody Wetlands class. However, occasionally, they correspond to other land classes. For example, as shown in the figure below, some represent Emergent Herbaceous Wetlands.
• Scroll down the table to evaluate the attributes stored in the NLCD_2011 and NLCD_2021 columns. As you can see, a visual assessment of 2,000 records for multiple columns is ineffective, and you need to use a different method to accomplish this task.

• Take a minute and think how you would evaluate row values in the available NLCD-related columns to easily detect features that don’t represent Woody Wetlands at all times during the analyzed period.
• Don’t close the attribute table since you will continue using it in the next step. Save the Project to ensure you don’t lose your progress.

Step 2. Construct a Comparison Expression in the Field Calculator

Hoping to uncover additional ideas, you posted the task challenges on the QGIS online forum. You received numerous helpful suggestions on how to compare the available NCLD-related fields and determine how many points classified as Woody Wetlands in 2001 represent the same land cover according to 2011 and 2021 classifications. You opted to automate this process by building a comparison expression in the Field Calculator tool. Below, you will discover a detailed description of how to achieve this, so keep reading!

• To begin the process, start the editing mode for the Woody_Wetlands_2001 layer and open the Field Calculator . Fill out the required parameters, as described below. Ensure that no selection is present by verifying that the Only update 0 selected features option is grayed out.
• Since you do not want to change any of the existing fields, verify that the box next to the Create a new field option is checked.
• Name the output field “Change” and make sure to use the Integer(32 bit) as an Output field type. Note that the type of field you need to create relates directly to the type of attributes the calculation will return, and in this example, the attributes are an integer type. You can keep the default value for the Output field length option.
• Place the cursor inside the Expression window and type in the following line of code:

("NLCD_2001" = "NLCD_2011") AND ("NLCD_2011" = "NLCD_2021")

• Before you click the OK button, you need to be familiar with a couple more principles. The key idea behind the expression is to return a value of 1, in case all NCLD-related fields are identical, and otherwise a value of 0. This is commonly done with the conditional statement, but here, we will use a comparison statement instead. The comparison statements return Boolean values true (1) or false (0). More specifically, by incorporating the logical operator AND, you will get a return value of 1 for true conditions (the NLCD-related fields are the same) and 0 for any of the two false conditions (the NLCD-related fields are different).
• Note that you can type the entire expression, including the field names, into the Expression window manually, but to eliminate potential syntax errors, it is easier to build it by double-clicking on the names shown under the Fields and Values menu. This way, names of the attribute fields will be automatically placed between double straight quotes ("). If you use a single quote (') instead, this will indicate text, and the expression will not be able to reference attribute values and will ultimately fail. This is a commonly overlooked syntax rule, so don’t forget that in QGIS field calculations, single and double quotes can’t be used interchangeably.
• Note a series of buttons located directly under the Expression window. Test these buttons to find out how they can help you generate expressions.
• If needed, correct the syntax errors. This should be evident as you will see a red text warning indicating that your Expression is Invalid. Use the options described above to help you make the required corrections. When all the expression problems are resolved, a Preview of the expected calculation result will appear. This means that you are ready to run your expression.
• One last time, verify that your inputs correspond to the figure below, and click the OK button to run the calculation.
• When completed, scroll down through the attribute table rows, and visually evaluate the results. If done correctly, a new field named “Change” is added to the table and attributed with values of 0 and 1.

• Assuming the operation was successful, it is important to save the edits . At this point, it might be a good idea to save the project as well. You may have noticed that the save icons are very similar and can be easily mixed up. The one small difference is a tiny red pencil at the corner of the save edits icon .
• Select features using an expression where the “Change” column values are set to the attribute indicating false conditions (0). Their spatial spread should be identical to the distribution of yellow points shown below. It is clear that the majority of the inspected points (1,849) remained unchanged. Only 151 points have changed the land cover class during the analyzed period. Now, you are ready to design the final map and share the results with your clients from the Mississippi Wetland Association.

​

Step 3. Alternative Solutions

Although this is an optional step, it might be well worth your time. As you may expect, there are multiple correct ways to write an expression that would achieve identical results to the ones described above. For example, you could employ the LIKE operator to search for a specific substring or use the conditional statement. Listed below are a few examples of these alternative solutions you can test. To make things easier, there is no need to create a new calculation field. You can run these expressions on the already-existing “Change” column. Simply check the option next to Update existing field and select the appropriate field name from the drop-down menu. While testing, fill out the remaining parameters as described in step 2. Each time you run a new expression, make sure to verify your results and clear any potential selections you make.

• The first solution relies on the LIKE operator and the % wildcard symbol representing any or no characters. The combination of the two is commonly used to look for a subset of text in a longer string. Note that the main structure of the expression is the same as the comparison statement described in step 2. Also, as mentioned earlier, double quotes (") indicate names of the attributes, while single quotes (') indicate text. In the example below, we are searching for the text ‘Woody’ in a longer string listed in the attribute “NLCD_2011”. The LIKE operator is case-sensitive, which means that searching for ‘woody’ versus ‘Woody’ will not produce the same results. Additionally, our expression is designed to look for the text ‘Woody’ with any characters before and after. Hypothetically, this means that evaluated strings such as ‘This Woody Park’, ‘Woody Park’, or ‘Woody’ are seen the same. Keep these concepts in mind while testing this expression.

("NLCD_2011" LIKE '%Woody%') AND ("NLCD_2021" LIKE '%Woody%')

• The second solution starts with the if() operator, analogous to the if() function you may have seen in Microsoft Excel. The basic form is as follows: if(condition, value if true, value if false). Initially, we define the logical condition, then we provide a value returned when the condition is true (1), followed by a value when the condition is false (0). All the structural expression elements must be separated by a comma and enclosed in parentheses.

if("NLCD_2011" = "NLCD_2001" AND "NLCD_2021" = "NLCD_2001", 1, 0)

• You can easily convert the expression shown above into a more elaborate solution. Multiple conditions can be combined into a longer series. For example, you can add a second condition accounting for a change detected in 2011 but no change in 2021 (restored to the original Woody Wetlands class). In that case, you would like to return a value of 2. This means some rows that would originally return 0 would now be marked as 2.

if("NLCD_2011" = "NLCD_2001" AND "NLCD_2021" = "NLCD_2001", 1, if("NLCD_2001" = "NLCD_2021", 2, 0))

Once you view the results, you may wonder why some cases of 1 were not changed to 2 while running the second condition. That is because our expression is considered from left to right, until the first true outcome occurs. More specifically, this means that when the comparing algorithm already produced a value of 1 for some row, it will skip the second condition entirely. This approach allows you to share more specific information with your client, especially if they would be interested in learning whether the change in land cover was permanent.

Conclusion

We have reached the end of our GEO Tutorial on incorporating a comparison statement into the QGIS field calculator. You can use concepts presented here to explore similarities and differences among values in any multiple columns. Minor adjustments allow for the use of similar expressions when analyzing strings, dates, integers, and other numeric values. In addition to detecting land cover change, these types of calculations could be helpful when evaluating the results of various classification methods, assessing datasets from various sources, or identifying typos in revised text attributes.

You have successfully finished this tutorial, but hopefully, you are up for a challenge. For example, it might be a good exercise to create a similar expression using just two of the available NLCD-related fields or add an extra fictitious field and expand the expression to more than three fields. Use this tutorial as your guide and have fun unlocking the hidden potential of field calculator expressions on your own!

Geosystems Research Institute

The Geospatial Education and Outreach Project (GEO Project) is a collaborative effort among the Geosystems Research Institute (GRI), the Northern Gulf Institute (a National Oceanic and Atmospheric Administration Cooperative Institute), and the Mississippi State University Extension Service. The purpose of the project is to serve as the primary source for geospatial education and technical information for Mississippi.

The GEO Project provides training and technical assistance in the use, application, and implementation of geographic information systems (GIS), remote sensing, and global positioning systems for the geospatial community of Mississippi. The purpose of the GEO Tutorial series is to support educational project activities and enhance geospatial workshops offered by the GEO Project. Each tutorial provides practical solutions and instructions to solve a particular GIS challenge.

This work was supported through funding by the National Oceanic and Atmospheric Administration Regional Geospatial Modeling Grant, Award #NA19NOS4730207.

The information given here is for educational purposes only. References to commercial products, trade names, or suppliers are made with the understanding that no endorsement is implied and that no discrimination against other products or suppliers is intended.

Publication 4032 (POD-08-24)

By Kate Grala, Research Associate III, Krzysztof Raczynski, PhD, Assistant Research Professor, and John Cartwright, PhD, Assistant Extension Professor, Geosystems Research Institute.