Several days ago, I asked about Calculating average data of neighbour grids based on location and returning data to grid in QGIS, in order to understand the trend of distribution of some attributes across the space.

In the operation, I realized a problem:

This alone would naturally offer values distorted (reduced) on the border of the area (lik the edge of town) where lack of polygons, given the presence of so-called ”false zeros” outside the squares/grids containing polygons(figur 1-A).These squares have zero values a priori – containing no polygon they have no chance to possess any attribute and therefore play a reasonable part in the analysis.

I would like to know if it is possible to estimate the degree of distortion mathematically and it is easy to work out some kind of correction?

Values of grids with polygons located close to the borders are increased proportionally to the size of “false zero effect”.

It could be like this (figure 1):

enter image description here

A – raw uncorrected result, B – corrected numbers

  • Are they 0 (zero) or NULL/NODATA? Jul 29, 2021 at 21:10
  • Thanks for your asking! In two of the grid layers they are null, the other one is zero.
    – TigerLily
    Jul 30, 2021 at 2:16


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.