To the list of courses || RASA2018 || To the theme || Estonian

Exercise 4040. Points 3, theme: Spatial Clustering

Open exercise
The attached file contains coordinates of farmsteads in Ridala parish from different maps (from the beginning of 20th century and from current topographical database). Code 1 marks presence and 0 absence. A village center is defined here by presence of at least 10 farmsteads within 500 m.
  1. How many farms have disappeared in comparison of these maps in this area?
  2. Which distance interval from the extinct farmsteads to a village center is most common?
  3. At which distance from the nearest village center is the number of extinct farmsteads the highest?

Use distance interval 1000 m. Coordinates are in meters.
Data: Ridala2019.txt


First, you have to copy the data to Excel and define space as column separator.
  • Open the SDC page Spatial clusters and have a look on the example data format there.
  • Prepare the input data (put a filter on column T_NA (current situation = 1) and then copy the filtered coordinates to the inpud cell.
  • Select DBSCAN clustering, press Boundaries from data, mark Minimum number of objects in cluster as 10 and Search radius as 500.
  • Press Calculate.
  • The result should be 14 clusters centers of which are the current village centers.
  • Save the results.
  • Have a look on the clusters as XY plot with the ortophoto or map background.

  • The next step is to set filter showing the coordinates of lost farmsteads using conditions: VERSTANE = 1 and T_NA = 0.
  • It is possible to get these coordinates using one-column filter and Copy-Paste. Although a more clever solution is to use advanced filter in Excel. Have you used it? If not – see: Instructions, Instructions, Instructions.

  • Now, open All Distances and copy the coordinates of village centers to the first input cell and the filtered coordinates of lost farmsteads to the other input cell.
  • Set 1000 as the Interval and 20 as the Number of histogram intervals.
  • Press Calculate and examine the results.
  • Calculate also the Nearest Neigbour Distances from the lost farmsteads to the village centers. Note that the NND function needs an ID column before X and Y coordinates.

Data and the excercise is prepared by Ott Koik.
Log in to send your results and to see the expected answer and responses from other students.