 This session will be an overview of the GIS application within DHIS2. We will start by demonstrating place search. This allows you to find various locations within the DHIS2 GIS application. We will then demonstrate the use of the facility layer, thematic layers, boundary layers, Google Earth Engine, and external map layers. These layers allow you to add various data types to the map, and we will describe these in much more detail as we go through the session. We will also discuss thematic layer options. This includes legends, methods, and classes. These allow you to define how your data will appear on the maps that you have created. We will also show the application of custom legend sets to the maps that you have created. Let's go ahead and get started with the session. In this session, we will go over using the GIS application. In order to access the GIS application, we will go to our apps and find the GIS application. After we enter the GIS application, we can see what the interface looks like. Let's describe the interface a little bit before we move on, as this looks quite a bit different when compared to the data visualizer and pivot table. Up top, we have the various layers that are available. The first layer that's available is referred to as the event layer. This is where we can add individual events to the map. So for example, if we are registering individual cases of malaria, we can map them using the event layer. The next layer is the facility layer. The facility layer allows you to map your facilities within the GIS application. After the facility layer, we have thematic layers 1 through 4. This allows us to map all of our health data in the form of data elements and indicators that are taken directly from DHIS2. After the systematic layer, we have what's referred to as the boundary layer. The boundary layer allows us to add administrative layers to our map. This includes areas such as regions, districts, provinces, states, etc. After the boundary layer, we have another layer called the Google Earth Engine. Have a look online to learn a bit more about how this works in practice. The Google Earth Engine allows us to add some datasets provided by Google satellites to our maps. We will show this feature in a little bit. Lastly, we have the external layer. The external layer allows us to add layers that are external to DHIS2. We will cover this as well in this tutorial. On the right hand side, we have the layer stack. Once we start adding actual layers to the map, we will define this a bit more. The plus and minus sign allow you to zoom in and out of the map. We have a magnifying glass button which allows you to search for locations on the map itself. And then we have a ruler which allows you to measure distances and areas. Let's go over a couple of these features now. Let's start by searching for a location. We can search for a location by using the magnifying glass. We type in the location we're looking for. And it will provide us with a list that matches the search term that we're entering. If we click on the location, it will take us to it. We can of course zoom in to the location to see a bit more detail. Let's zoom back out. The next thing we can do is add an external layer to this map. We click on external layer and then edit layer. It will give us a prompt to say select layer. There are a couple external layers already available for you to use. Have to add these external layers will be discussed in other academies. For now, we will just focus on using those layers that are already available. When we select the layer, we can then click on update. This will add that layer to the map. On the layer stack, we can quickly add and remove layers. For example, that external layer that I just added, we can remove it by clicking on the check mark next to the external layer option. You can see this quickly removes that layer from the map. We can also add it back to the same process. Let's add one layer from Google Earth Engine. We'll go to Google Earth Engine and click on edit layer. It then provides us with a prompt to select a data set. From the drop down, there are several data sets available. You can read more about how this data is collected online. After we select a data set, there are a couple different parameters that we need to define. Here we can define the minimum and maximum elevation, the color scale, and the steps. If we look at another data set, for example temperature, we actually have to select the period associated with the temperatures we want to display. We also define the min and max degrees, the color scale, and the steps. So each data set has different associated parameters that we need to define. Going back to the elevation data set, we'll just leave it as default and click on update. Now we can see that this elevation data has been applied to the map. Let's just remove this pin point for the location so we can see things a bit more clear. We do that by clicking on the X next to the search box, where we searched for our location initially. Let's clear this map and add in an example where we're working with some of the health data that we've been using in Trainingland. Now as we mentioned earlier, Trainingland is a fictional location. We're going to remove the initial background layer from this example because Trainingland is essentially sitting in the middle of the ocean. So I'm just going to deselect the OCM light layer from the layers tab. This gives us a map with no background. We discussed earlier the different layers that are available through the interface. Let's add in the boundary layer for Trainingland. In order to add the boundary layer, I go to boundary layer and click on edit layer. When we open up the boundary layer, it asks us to select the organization unit levels. We've been discussing the use of organization unit levels in the data visualizer and the pivot table. The same principle applies now to these boundary layers. In this case, let's select an administrative boundary. We can select the districts for all of Trainingland. This will display all of the boundaries for the individual districts within Trainingland. There is the second tab referred to as options. Here we can add the labels. So in this case, if we say show labels, it will show the labels of the districts. Once I've selected the organization unit level, I click on update. You can see that it adds in the boundaries of the districts for Trainingland. If I hover over one of these districts, it will give me the name of that particular district. Note that in this example, I have Trainingland highlighted, and that's why it shows me the boundaries for all the districts in Trainingland. This is the same as our other organization unit selections in other examples. So if I only select animal region and say update, it will only show me the districts in animal region. So just make sure you have the proper parent organization unit selected when you're working through these examples. I'll switch back to showing the districts in the country. We'll close the boundary layer by clicking on the X, and now let's add a thematic layer. So now we want to add some actual data to this map. In order to do this, we can use the thematic layers 1 through 4. Let's add one layer first. So we'll click on the number 1, which represents our first thematic layer. We can then click on edit layer. We can see that the selection for data and periods is a bit different when compared to the pivot table and data visualizer tool. Here we select the data and the periods together in one tab. Previously, these were separated. In this first example, let's work with the data element. We will select data element from the value type. You can see the prompt changes from indicator group to data element group. We can work with the malaria data element group in this example. After we select the group, we can select our data element. We'll select rdt positives in this case. From the period type, we can select either a relative period or a predefined period. You can see these are now combined. Before you had a top and bottom area where you could select between relative and predefined periods. In the GIS application, you must define whether or not you're working with a relative or predefined period, and then subsequently select the period you want to work with. If I select relative period, you can see now that relative periods appear for me in the period selection. If I were to select a predefined period, such as yearly, then the periods available are associated with those predefined periods. Let's work with a relative period of this year. In a GIS map, you're only allowed to display data from one period at one time, so just keep that in mind when you're making these maps. After we've defined our data and our periods, we select our organization unit. So where do we want the data to come from? Typically we work with organization unit levels. This is the default selection method. If you click on the gear icon, however, you can change the selection mode to be more specific to select specific organization units. Let's change this to display the data by facility. From our dropdown, we'll deselect region and select facility. Our last tab is the options tab. We can define the legend type, the classes and the methods, the color scale, the low color size and the high color size, add the labels and define the aggregation type. Let's discuss this in a little bit of detail. First, we'll start with the legend type. There are two types, automatic and predefined. We will come back to the predefined legend set in another example. Next, we can define the classes and the methods. In order to explain this a bit more, we'll update our map and walk through this. The color scale allows us to choose which colors to apply to the legend of the map. There are a number of different color scales that are available. For the low color and the high color size, this applies only to point-based data. If I'm looking at facility-based data, we can change the low and high color size of these different points. We'll update the map and then see how we can change the low and high color size on the map itself. The aggregation type defines how I'm going to aggregate the data that I've chosen. In this case, we have selected the data element, malaria-RDT positives. By default, it is going to sum this data element. We can change how this is aggregated by selecting the aggregation type here. For now, we're just going to leave the aggregation type as the default. Let's click on Update in order to see the output of the map. After I click on Update, we can see that the different points appear on the map. This represents the number of malaria tests that are positive in each facility. If I hover over one of these points, I can see the value for the number of malaria tests that are positive within that particular facility. We can see that there is a legend on the right side. This gives us a breakdown of how the points represent the data. We can see that this lighter color here at the top represents the values starting from 34 all the way to 4700. We can see that this darker value at the bottom starts to represent those values that are greater than 18948. We can see also that those values that are lighter in color are smaller in size. Likewise, those values that are darker are larger in size. This is defined by the low and high color size in our options. If we change this and make it something different and then update our map, we can see that those points representing those colors that are higher in value become larger. We can also see that this legend is separated into five classes. This is determined based on the number of classes we've defined in our Thematic Layer 1 options. We've currently defined five classes. Let's increase this and see what effect this has. You can see when I increase the number of classes, the points on the legend increase before we had five points on the legend scale. Now we have seven points on the legend scale and the values that they represent are changed. Now the first color represents 34 to 3400+. This is different from our value before when we had five classes. We can also discuss the method a little bit more. Right now, the method is equal intervals. This means that the interval within the legend tries to be equal. We can see there's roughly a difference of 3500 between each of the points on the legend scale. For example, here we have 34 to 3400+, 3400+, to 6789, 6789 to 10000+. This is roughly a difference of 3400 between each point on this legend scale. It's not equal, of course, but it's trying to maintain some semblance of equality between each interval on the legend scale. If we change this to equal counts and update, you can see that what it tries to do now is have an equal number of values within each legend point. Before it separated the data based on an interval, regardless of how many values belonged to that legend point within our legend. Now it is trying to have an equal number of values belonging to each legend point. We can see here, for example, at the lowest end we have values of 34 to 427. This has 24 data values associated with that legend. The next value is 427 to 1077. The differences between these two are not exactly equivalent, but the number of values within the scale is trying to be equal between each interval. We can see, for example, at the top we have a scale of 6533 to 23677. Here we have a difference that is much greater than that small difference of 400 between the start and end point of these data values representing this point on the legend scale. So when we're defining the legends in our options, we should think about how this applies to our map a little bit more, in particular how this affects the data values that will be displayed on the map. We'll just reset this to 5 and we'll show you the effect it has. See that it's still trying to maintain an equal number of counts within each of these points on the legend scale. As we have this data by facility, it might be useful to actually add the facilities to the map. In order to do this, we'll close the thematic layer and we'll add in a facility layer. We'll click on facility layer followed by edit layer. In this case, we have the organization units grouped together. It will give us a prompt to select the organization unit group set. In this example, we've separated the organization units based on their ownership, their type, and their location, whether or not they're urban or rural. So if we select the type group set, this will separate the facilities based on their classification type on the map itself. We will discuss how to define these organization unit groups in other sessions. Once we've selected our organization unit group, we then select the level of the organization unit hierarchy. In this case, we selected type as our organization unit group. This only applies to our facilities. This is the only level in which we actually have facility information. In this case, we can't add facility layers to the country, region, or district because we do not have facilities at that level. Lastly, we have the options tab where we can add the names of the facilities and add a circular area for the radius. Before we do that, let's update the map to see what effect the facility layer has. You can see that when I update the map, the facilities are added to the map. These are separated by type based on the classifications that have been defined within DHIS2. As mentioned, how to define these classifications will be discussed in later sessions. We could go back and edit the layer to make the high and low color size appear a bit larger or we could remove the facilities from this particular map so we could see the data once again. If we zoom in, we can start to see the separation of this data a little bit better as well as the facilities and their type. If we hover over the facility, it will give us the name of the particular facility. Let's remove this layer and continue on. We can temporarily remove this layer by using the layer stack option. We'll just deselect the facility layer. Let's zoom back out. We can actually use this other magnifying glass when we're zoomed in to zoom back to our content. This will allow us to zoom back out to the full map very quickly. Let's add in a second thematic layer. If we go to thematic layer 2 and click on edit layer, we can now add in a second layer. Remember we discussed this concept of having different layers in a map being represented by different pieces of data. As we're adding more health data to this map, it's a good idea to add some type of data that is related to the initial data we mapped on the first layer. Our first layer contains information for the malaria-RDT positives. So we should try and add a second layer that relates to this first layer in some way. In this case, we can compare the malaria-RDT positives with the number of tests that have been performed. We'll work with data elements again and select malaria. Let's select the malaria-RDTs that have been done. We should select the same period as we did previously. We selected a relative period and we selected this year. As for the organization units, let's select the district level. This will allow us to have one layer of data that shows information by district and another layer of data that shows information by facility. For our options, we are just going to leave them default for now. As these two layers will be stacked upon one another, we should make sure that the colors complement each other. In this case, it will work out well if we just leave the color scale to be the same as the color scale we used previously. Let's update the map. We can see what effect this has on the map. We'll just decrease this facility layer legend. So now we can see the second thematic layer. Here we see the information associated with that second layer we've added to the map. We might want to change our method. In the first layer, we've used equal counts. Let's also use equal counts for the second thematic layer. In this map, we can look at the correlation between the RDT positive tests at the facility level and the total number of RDT tests done within that particular district. As a simplification, lighter facilities should match up with lighter districts. However, this might not always be the case in practice. We can save these maps as favorites just like we did with charts and tables. The favorites interface is a little bit different when compared to the pivot table and data visualizer applications. If we click on favorite, a prompt will come up that allows us to manage our favorites. Here we can search for favorites and also add new favorites. If I want to add a new favorite, click on add new, and then I'll provide the favorite with a name. We'll still follow the same naming convention that we've been using this entire time. Once we've provided the favorite with a name, we click on create. In order to pull the favorite back up, we just go to favorites, we search for the favorite, and then we can pull our favorite back up. So far, we've been using examples using an automatic legend. I mentioned earlier that we would come back to the concept of a predefined legend. Let's clear this map and go over this concept in a little bit more detail. In this example, we'll apply a predefined legend set to our map. How to create these predefined legends will be discussed in other academies. But for now, we are going to show you how to use the predefined legends that are available. If you remember, we used a predefined legend in our pivot table example previously to highlight different colors on the pivot table. The same concept applies to the map. I'm just going to remove the background layer quickly, and then I'm going to add in a thematic layer. We go to thematic layer one and then edit layer. This is following the same steps as previously. We'll work with an indicator this time, and we'll work with the immunization coverage indicator. Let's select BCG coverage. This is one of the indicators we worked with previously in that pivot table example. We can select yearly as our period type, and let's look at data from 2015. After we've selected our data and periods, we select our organization units. In this case, I'm going to look at this data by district. I will deselect region and then select district as the organization unit level. For my options, now I want to change the legend type from automatic to predefined. I will click on the dropdown and change the legend type to predefined. We then have another prompt in which we can select the legend set. We'll work with the EPI coverage, which is the one we used previously in the pivot table example. I'm going to leave everything else as its default values. Now we can update the map. We can see now that a customized legend set has been applied to this particular indicator. Remember, how to create these legends will be discussed in other academies, but it's possible to create and define your own legend types within DHIS2 and apply them to your data. Alright, let us quickly review some of the concepts we've covered in this GIS tutorial. We discussed the interface first. On the top left hand side, we have our various layers that we can add to the map. We went through every layer, but the event layer. We can add facilities, add health data, and add other supplemental information to the map. On the right hand side, we can then add and remove layers as required. We can also zoom in to the map and search different locations on the map. If I want to search, I'll click on the magnifying glass and look for my location. We can add in additional supplementary layers by using the Google Earth Engine or the external layer. Let's add an external layer as an example. We can also add in boundary layers, which allow us to add in the administrative boundaries of a particular location that we're interested in. We'll just remove the background layers from the map, and we'll add in the boundary layer. We can also add in thematic layers, which include information about our health data. Note that the selection of data and periods is a little bit different when compared to pivot tables and the data visualizer. First, we select our data and periods together, followed by our organization unit. Typically, we work with the select levels method of selecting organization units in the GIS application. If I go to options, then we can define our different options. We have the legend type, the classes and method, the color scale, and the low and high color size of different points as applied to the map. We can also add in labels and change the aggregation type. You can see when I update the map, the legend is then applied to the map based on the individual options we've selected for that particular semantic layer. Depending on how our data looks, we can always change how the data is displayed, depending on the options that we select. These are important considerations when making maps within DHIS2. Alright, this ends the GIS session. Please give the exercises a try and let us know if you have any additional questions.