ArcMap

Mapping as a reporting tool is exploding on websites. Jennifer LaFleur looks at geographical information systems that produce statistical maps and other visualisations. She explains how to use ArcView to analyse data geographically to dig deeper into your reporting.

The simplest map we’re used to seeing is a map telling a single story.

Such a map might point out where a small town is located along a state’s panhandle. It might trace a line where a proposed highway will run. Or it might show the states in which the presidential candidate won.

These are all examples of a single layer of information. A point for a town. A line for a road. An area for a vote winner. These may sound simple, but it is important to do a single map before doing more complex work with multiple layers of information.

When reporter, Ramona Smith, was covering the environment for the Philadelphia Daily News, she wanted to see how well the city’s recycling programme was doing. So she got the data. The city kept track of how much was being recycled, district by district. By mapping these districts she was able to clearly see which neighbourhoods were doing the most. And those that were doing the least.

When JJ Thompson was a graduate student at the University of North Carolina, she was curious: where are the biggest income gaps? She compressed census data into a handy index and the result was she could see where the widest margins existed between the rich and poor in the state.

When The Washington Post’s Dan Keating was exploring where high tech businesses were emerging, he plotted Dun and Bradstreet data around the nation’s capital, and was able to see where the corridor was developing.

In all three cases, a single map told the story. A reader could look at these maps and know whether they lived in a community that recycled the most, or had the biggest gap between rich and poor, or see whether new industry was moving into their area.

The following sections explain how how to put together a single layer map in ArcMap.

About the tools

A lot of what makes up a mapping program is familiar ground to those who’ve done basic computer-assisted reporting. Mapping programs still use data stored in tables. They may let you point-and-click your way through a question, and the language running behind those mouse moves is database query language or SQL. But mapping programs go several steps further and allow you to link data to maps, and to slice and dice data based on where it fits on the globe.

In addition to the data, you’re working with geography. Geography comes in several different forms. Outlined geographies such as counties or census tracts are known as polygons. Streets, rivers and railrways are lines. Cities or events may be symbolised by points. As you progress, you’ll learn how to work with each of these types of geography.

Bringing maps into ArcMap

ArcMap is the main component of the ArcGIS suite produced by ESRI, it is available at three licence levels: ArcEditor, ArcView, and ArcInfo. ArcMap is the application used to view, edit and query geospatial data, and create maps.

ArcMap lets you set aside a work area for what you want to do. This work space is called a map document. It’s similar to the way a database works in Access, and a project works in FoxPro. In map document you can store tables and a whole lot more. The project also serves as a traffic cop, keeping track of the different places in your computer where you store the various maps and databases within your project.

When you start ArcMap, it will ask you if you want to make a new map or use an existing map. Choose ‘a new empty map’.

You’ll get an empty window that is just waiting to be filled with a map.

To add a map, click the ‘add data’ button.

Go to the folder where your map has been saved as a .shp file (shp stands for shape file). ArcMap can read layer files as well as ArcView files called ‘shape’ (shp) files.

In the example, the map is set so that all the countries are the same colour. It also is in the ‘full extent’ zoom.

You can get a closer view by clicking on the zoom tool. With this tool, you can click and draw a box around the main land area.

On the left of your screen in ArcMap is the table of contents – that tells you what data you have open.

Along the bottom, a tab lets you switch between your display and source. The source tells you where you are getting the data for the map.

Behind the pretty pictures in any GIS program is a database. In this case it’s the data that comes with ArcMap and is already attached to the map. Let’s look at one record in the data behind this map.

Click the ‘identify’ icon and you’ll see the ‘identify results’ window pop up. As you click any county on the map, you’ll get a window that tells you all the fields in this file. In other words, you’re getting a peek at one record in the counties database.

Scroll through the window and look at all the data you have. You can leave this window open and click on another county and it will switch to that county’s record.

This is great, if you just want the information for one country, but what if you want to see the whole table? To see the whole table at once, close the identify window and right click on the ‘counties’ layer in the table of contents and choose open attribute table. You’ll see the sort of thing you’re used to if you use Access or Foxpro – rows and columns of data.

Looking through the data, we might want to create a thematic map based on some of the data.

Click on the far left of any record. Notice how it highlights the row. That’s one method of selecting information in your data. When you close the table, you’ll see that the county you selected is highlighted on the map. The map goes both ways.

Choose the select tool and click on a county.

Saving your map

Select ‘file’ then ‘save as’ and give your file a name. Keep in mind, ArcMap is really just saving an index of where all your information is, not all the information used in the map. So if you move stuff around on your computer, ArcMap will be used.

To do that in this example, you would need to double-click on the ‘Euopre’ layer in the table of contents. Making sure the ‘symbology’ tab at the top is selected.

The map in here is using ‘single symbol’ symbology. Another type of map we might want to choose is ‘categories’.

Using this symbology, every change in a value gets its only symbol, or colour. Let’s say that we want each country to be a different colour. In the box labelled ‘value field’, choose ‘state_name’. Then click ‘add all values’ at the bottom. Then click ‘apply’ then ‘ok’.

The legend type is ‘unique value’ that means that every change in the ‘values_field’ (in this case ‘state_name’) gets a unique value.

This map is interesting, but it really doesn’t give us any information. You may have noticed that the file contains demographic data. Go back to the layer properties window and choose ‘quantities’.

Choose ‘graduated colours’, then under ‘fields’ choose the value called “per_hh’, which is persons per household.

ArcMap automatically assigns ranges and colours. You can change the colour ‘ramp’to a variety of things.

Next, click ‘apply’ in the lower right corner then ‘ok’ and close the legend properties window. Your map shading will change. If there are too many colours go back to the legend editor and reduce the number of categories. To do that, under ‘classification’, choose ‘number’ and change it. Click ‘apply’ then ‘ok’.

If this makes the map look strange with breaks that are not easy for readers to understand you can create your own breaks. Go back to the layer properties window. To change the classification, click the ‘classify’ button. And although this window looks like the cockpit of a jet, it has some very helpful information. It graphs the data so we can see the distribution and it summarises our data.

Click on the down arrow next to ‘natural breaks’. You’ll see that there are several types of distributions we could use for our data.

Natural Breaks

This is based on analysis ArcMap does to try to figure out where the data naturally breaks – or where big jumps occur in the data.

Equal interval

Each break will be an equal distance apart (10 20 30…)

Quantile

This breaks the data so the same number of counties would be in each category.

Manual

This lets us pick the breaks. Most journalists use this because even categories are easier for readers to understand. Try a quantile map with four categories (quartiles).

Next, let’s do a manual map so we can make the categories easier to understand. To use this, you must first pick the number of classes you want. Then, you can set the breaks by either typing them in the ‘break values’ box or by moving the lines on the graph. Let’s type them in: 2.5 and 4.0. The top value will default to the highest value, in this case 8.7.

You can type in the values that you want for the range and edit the labels.

Using persons per household it was easy to just map the median value, but what if we wanted to map a particular race category. Would it make sense to just map the number of men or women? When you work with that data, you always need to base it on population. Thats where the ‘normalise’ function comes in. For your value field choose ‘males’, then we need to ‘normalise’ it to the population. In the ‘normalisation’ drop-down menu, choose ‘population’.