Hello, One possibility is to do a nearest neighbor analysis with a distance matrix. To do so, import your shape files first. Then go to Vector -> Analysis Tools -> Distance matrix. The input point layer should be your 500 points layer. The target point layer should be your 3000 points layer. Next define unique ID fields. Check the "Use only the nearest(k) target points". In your case you have to enter the value 1, because you are looking for the nearest point. Select the output path and click "ok". Afterwards you have to load the created matrix (as text layer). When you are loading this data make sure that you select no geometry, because this file is just text columns. I think this should work für your case.

Best regards Paul.

In the same spirit, you could get x/y/z coordinates of the two layers and save the attribute table as csv table.

Then you can use a knn algorithm with R (kNN package) or Python (package Scipy : spatial.kdTree or spatial.ckdtree) to associate each one of your 500 points to the nearest one (k=1) in the 3000 points dataset.

After that rebuild the table and import it into Qgis and it is done.

This is maybe too much manipulations in your case which also demands some programming skills.

However, it is a very efficient way when your start using bigger datasets. Moreover if you know how to manage geodatas in R or Python, the full process can be done without opening Qgis in a very short time.

Best regards

Hi, in the toolbox, use distance to nearest hub and select your point layers (maybe save the csv to shape first). Select line and you will get distance in meter.

Cheers,

Michael

Nearest Neighbor Analysis

Paul Griesberger

Thank you so much for your help and reply. May I ask what is the purpose of loading the created matrix as text layer?

Also, I did everything you mentioned and it worked, but I got 3000 distances whereas I expected to have 500 values. I checked the "Use only the nearest target point".

Thank you!

Michael Kempf

Thank you so much for your answer.

I am receiving the following error.

Hello Termeh Hesam Mahmoudinejad,

Sorry, i mixed up the layers. The input point layer has to be your 3000 points layer. The target point layer has to be your 500 points layer. 

Adding layers as text layers is one possibility to load .csv files in QGIS.

Best regards

Paul

Thanks Paul Griesberger .

I got the following error this time.

Hello Termeh Hesam Mahmoudinejad . It is just a guess, but maybe your data has missing geometries. Check all rows of your 500 points layer for missing geometry values.

Best regards Paul

Hmm, maybe wrong geometry? Do the files contain the same coordinate system?

if you want, send me the files:

michael.kempf(a)archaeologie.uni-freiburg.de

Best

Michael