cancel
Showing results for 
Search instead for 
Did you mean: 

Head's Up! These forums are read-only. All users and content have migrated. Please join us at community.neo4j.com.

Is gds.alpha.triangles not able to list all triangles in large graphs?

Hi there,

I am looking to list all of the triangles in some large graphs from stanford. (one example is the web-Google graph)

It seems that gds.alpha.triangles is exactly what I want, but when I run the algorithm, only about one tenth of the triangles are listed.

Does anyone know why this might be, and how I can list all of the triangles?

Bonus Questions:
-Why is this algorithm only in the alpha stage?
-Isn't triangle enumeration (i.e. listing) needed for calculating things such as the clustering coefficient, or does neo4j use approximation algorithms for calculating such values?

1 REPLY 1

What version of GDS are you running?

GDS.triangleCount moved the product tier in the 1.3 version of GDS. If you're looking to run local clustering coefficient, that is also in the product tier, and uses the output of triangle counts.

triangles remains in the alpha tier due to lack of demand, but it shouldn't have any bugs in it (and we fixed bugs in 1.2 and 1.3). If there are still problems with returning all triangles, please create a ticket on our github! https://github.com/neo4j/graph-data-science/issues