graph-algorithms | Алгоритмы из книги Алгоритмы на графах | Machine Learning library

I would like to use the neo4j algo.algorithms. I am running neo4j desktop verison on Windows.

I tried to install the graph algorithms extension following these instructions: https://neo4j.com/docs/graph-algorithms/current/introduction/

But I ran into a problem immediately since there is no install option:

How can I install Graph Algorithms?

I have three Person nodes with various relationships. Each node has a latitude and longitude.

First, I find the combinations of pairs of nodes:

I am running neo4j on an ec2 machine and I want to run some graph algorithms. The problem is that some of the algorithms I want to use are not installed properly and I am getting this error : There is no procedure with the name xxxx registered for this database instance.

By executing CALL apoc.help('algo') I only get 17 procedures and the procedures I want to use are not listed. Some of the procedures not listed are apoc.algo.betweenness, apoc.algo.closeness and algo.pagerank.stream

neo4j version : neo4j 3.5.14, apoc-version : 3.5.0.8, neo4j-graph-algorithms-version : 3.5.14

I have also added this to my neo4j config file

I work with a Neo4j database where I only have read permission.

I'm trying to run some of the algo algorithms. E.g. a community detection algo.scc.

According to the documentation algo.scc has a parameter write which

Specifies if the result should be written back as a node property.

However when I run it with write set to false

I wanted to test some graph algorithms, node similarity in particular. So I've setup a docker container using neo4j:3.5 (which is v3.5.14 at the moment) official docker image. Additionally I've downloaded APOC v3.5.0.7 and Graph Algorithms v3.5.4.0 and added this line to the config:

dbms.security.procedures.unrestricted=algo.*

Unfortunately function algo.nodeSimilarity seems not to be present in the registered functions:

I'm creating a CMake project whose two main files are:

I'm using Neo4j 3.5.9 from a Docker image, and I'm trying to use the Jaccard Similarity algorithm. However, when I run a query like this:

My graph contains the labels X and UnknownX which are linked by the relationship :MAYBE and I'd like to identify which is the proper X if match (u:UnknownX {name: 'John Doe'})-[r:MAYBE]-() return count(r) > 1. For this I figured Jaccard might be a decent initial approach.

The documentation of the Jaccard similarity implementation of Neo4j's algorithm plugin states that

When calling a procedure using an evaluated inner query (cypher projection)which would be creating an node or edge a forbidden error is thrown for the admin user. It is possible to run the query outside the procedure.

Maybe it is something of an configuration issue?

I am trying to run a query using the timetree procedure of GraphAware and the LabelPropagation of Neo4j. But I've been able to get the forbidden error using a much simpler query. I am using Neo4j Desktop. I have tried giving the admin all the possible roles.

For example