The Line Layer in Qlik Sense Maps
Until recently, the Qlik Sense map chart was an object that looked great on a dashboard without offering a great deal of analytic value. You could display a single dimension, such as Country, Province or City as an area or a point, and one or two measures using bubble size and colour. The heavy lifting of your geo data was left to Geo Analytics, the excellent, but expensive, add-on for Sense. That has changed with the last two releases of Qlik Sense, the April 2018 release bringing us area and point layers and the June 2018 release adding the line layer. It is the line layer that for me really adds value to the map chart, allowing us to analyse point-to-point data visually to highlight issues that would not be apparent using any other visualisation, particularly when looking at deliveries to customers, deliveries from suppliers or other logistics type data.
In order to analyse deliveries to customers, Qlik Sense's data profiling means you no longer need to gather GPS coordinates for all of your customers, Sense can use a field such as City in the customer address data to automatically geo-tag each customer's approximate location. Alternatively it's easy to find a list of the GPS coordinates of all South African post offices to link to the postal code in the address, or use some other method to get GPS coordinates, such as a reverse GPS lookup.
To analyse delivery data visually, you need a customer location, a delivery location, delivery time (days) and possibly sales value so that you can identify major customers. In this example I have two Distribution Centres (DC) and each Customer has a delivery location ([Place Name]). Delivery Days are calculated in the load script as [Delivery Date]-[Order Date] and each customer has a sales value.
The map object has an area layer for sales by Province as well as point layers for DC and Place Name. We're interested in the line layer though, which takes two dimensions, from and to. Both must be geo-tagged dimensions. DC is the from field and Place name is the to field. That is all you need to create map lines, but they can be made more useful by adding width and colour. In this example I have used count([Orders]) as width, to identify routes with a lot of deliveries and avg([Delivery Days]) as the colour measure, to identify problem routes, the higher the average deliver days, the darker red the route line. You can optionally add arrows to indicate the direction of movement of the line, as well as adjust the line curvature. The bubble size and colour of the Place Name bubbles is sum([Sales Value]), to identify high value customers.
Once we add these elements as layers on a map chart, we can immediately see that we have one problem route that stands out, a high value customer, with a dark red line, indicating a higher number of average delivery days. We can zoom in on this route and analyse why this particular route has a higher average days, perhaps it is related to the high number or orders, or some external factor like the shipping company or something else.
In a similar way you can analyse deliveries from suppliers, sales reps routes or any other point-to-point data.