This week's article is a republish of the great article by @Tony Maber about the Spider Graph tool written back in 2019. I referred to this article in MapInfo Monday: Connecting Objects in Two Tables with a Line.
Happy #MapInfoMonday!
Spider Graph
Spider Graph is used to map relationships between tables based on a designated field in MapInfo Pro. By joining the tables based on the designated field, MapInfo Pro then creates a new table of lines that connect the objects from the original tables based on this join – effectively "connecting the dots".
The Spider Graph tool provides visual and analytic insight into your data. The table, developed using the Spider Graph tool, contains the lines linking locations, the chosen attributes from the input table(s), and a column that stores the length of each line.
Examples of applying the Spider Graph tool could include creating links between:
- customers and stores they visit
- warehouses and delivery locations
- mobile phones and telecom antennas
- crime scenes and possible offenders
- postcode centroids (as part of sales territories) to branch offices
- where students are going to school, or patients are going to hospitals
- and vehicle theft and recovery locations.
Working with the Spider Graph Tool
To demonstrate the power of the Spider Graph tool, we will look at the business problem of customers and stores. In this case, there are 1,514 customers from a membership database whose closest location to shop is one of the 6 stores shown. These are presented in the map below – customers as red dots and stores as blue diamonds.
Our customer data includes attribute fields for purchase history, CAMEO-based geo-demographic segmentation, the closest store, the distance to the closest store, and the customer's preferred store. The closest store analysis was completed using the Distance Calculator tool in MapInfo Pro.
Using the "PreferredStore" attribute, we will be able to determine travel distance comparisons and analyze other attributes for the customer group that does not shop at their closest store using the Spider Graph tool.
The first step is to load the Spider Graph tool using the Tool Manager. Next, you access the Spider Graph dialog as follows:
The Spider Graph dialog will then appear. For this example, the customer table is chosen as the origin, and the stores table is chosen as the destination. The attribute field in both tables that contains the data on which to join the tables is selected in each case. The attribute fields available from both tables are shown on the R/H side of the dialog and allow users to determine which attributes they would like returned with the line linking the two objects.
Selecting key attributes here allows for further analysis of records. For example, what CAMEO Groups are most represented in those records where customers choose not to shop at their closest store, or what is the average distance customers travel to their preferred store compared to the average distance to their closest store?
In the example above, I have chosen not to color-code the resulting lines linking the points (this can be done later on the lines using any of the attributes returned). I have chosen the distance unit (kilometers) and ticked the option to add a field for the line distance.
Once satisfied with the selections in the Spider Graph dialog, click on the Create Lines button to generate the outcome. If a map window is active, the results will be displayed in that window; otherwise, a new map window will be presented containing the Spider Graph results.
In our example, the map and browser generated by running the Spider Graph tool are shown below.
Analysing Results
A quick review of these Spider Graph results revealed some actionable insights about the customer base involved in the study. Some of these insights include:
- 345 (22.8%) of customers shop at stores that are not their closest store.
- 8% of these customers are in the top 2 CAMEO Groups (which is significantly higher than the sample from all 1,514 customers used in the study at 67.3%) – this may indicate a propensity for more affluent customers to be more selective about where they shop or have reason to shop at stores other than for convenience (eg. closer to work, major shopping mall vs neighbourhood strip shops, etc)
- The 345 customers who did not shop at their closest store chose to travel an extra 57.3% further to shop at their preferred store
The final map image below uses green dots to represent customers who shop at their closest store and red dots to represent those customers who do not. This visual insight allows users to identify trends or factors (geographic barriers, infrastructure influences, movement patterns toward the city, etc.) that may be influencing customer behavior.
So, in a very short time, MapInfo Pro and the Spider Graph tool have delivered visual and analytical insights into the relationship between two datasets.
We would love to hear how MapInfo Pro users have applied the Spider Graph tool in solving problems. Or please feel free to let us know which MapInfo Pro tool/add-in is your favorite.
------------------------------
Peter Horsbøll Møller
Principal Presales Consultant | Distinguished Engineer
Precisely | Trust in Data
------------------------------