memgraph / docs Goto Github PK

On the MERGE clause page, the link to the 3rd section is not working properly.

The following links at the beginning of the page are not leading me to those sections on the page:
Merging with ON CREATE and ON MATCH
3.1. Merging with ON CREATE
3.2. Merging with ON MATCH
3.3. Merging with ON CREATE and ON MATCH

https://memgraph.com/docs/cypher-manual/differences#unsupported-constructs

Under unsupported constructs there is mention of allShortestPaths, which we have now implemented, and is used the same way as wShortest:
https://memgraph.com/docs/memgraph/reference-guide/built-in-graph-algorithms#all-shortest-paths

Add RETURN after YIELD for each of the MAGE procedure code examples. Example:
Weakly connected components

CALL weakly_connected_components.get()
YIELD node, component_id;

An example case of when the above is not working (on Pandora dataset):

MATCH p=(e:Entity)-[:JURISDICTION_IN]->(:Country)
WHERE e.icij_id="7A15CC81D9DA787D79F94E5487CC47C9"
WITH project(p) as subgraph
CALL weakly_connected_components.get(subgraph)
YIELD node, component_id;

Error message: Query should either create or update something, or return results!

If the above is changed to:

MATCH p=(e:Entity)-[:JURISDICTION_IN]->(:Country)
WHERE e.icij_id="7A15CC81D9DA787D79F94E5487CC47C9"
WITH project(p) as subgraph
CALL weakly_connected_components.get(subgraph)
YIELD node, component_id
RETURN node, component_id;

then it works.

Hence, the code snippet in docs needs to change to:

CALL weakly_connected_components.get()
YIELD node, component_id
RETURN node, component_id;

On page https://memgraph.com/docs/mage/how-to-guides/run-a-query-module under tabs Docker hub and Docker build , the dev version of the image needs to be prefixed with a version, e.g.

docker run --rm -p 7687:7687 --name mage memgraph-mage:1.4-dev

At the moment, that exact version is added as an instruction, need to make a more general instruction so there is no need to update

Question:
On https://github.com/memgraph/docs/blob/master/cypher-manual/cypher-query-language.md, the query:

MATCH (p:Person)
WHERE EXISTS {
  MATCH (p)-->(c:Country)
  WHERE c.name = 'UK'
}
RETURN p;

throws an exception: Query failed: line 2:14 mismatched input '{' expecting {, ';'}
Also throws it for the second query, with a one-line subquery
Did we drop support for WHERE EXISTS?

Marko B explained that: we never supported that, evaluating a pattern/subclause as a function argument is not supported… the situation is a bit better now because you can do the same by executing custom function

Wrong directory name in --query-modules-directory flag value:
--query-modules-directory=/usr/lib/memgraph/query-modules
The correct directory is query_modules.
https://memgraph.com/docs/memgraph/reference-guide/configuration#query

Currently, only the BFS algorithm has been documented. Users are not aware of how to run DFS to traverse relationships.

The current NetworkX how-to guide is not a how-to guide and it should be created.

Docker whale is a bit lonely 😄 Can we add other icons/emojis as well?

While searching through the documentation, I expected to find some info about the BFS and DFS algorithms in the Cypher manual. It would be good to document it somewhere inside the Cypher manual.

Document:

manual transaction workflows using BEGIN, COMMIT, and ROLLBACK commands

We need to add a link to MAGE repo on the https://memgraph.com/docs/mage.

There is no date for Memgraph v2.4.0 changelog.

We are missing LIMIT clause in the Cypher manual

https://memgraph.com/docs/cypher-manual/clauses

find_edge_by_id:
it’s possible to store vertex ids on the app side, visualization already has them, elastic doesn’t but it should be possible to add

https://memgraph.slack.com/archives/CSFF16NCD/p1667392485720689

We have uniqueness and existence constraints in the how-to guides, but not in the reference guide.

https://github.com/memgraph/docs/blob/master/docs/reference-guide/graph-algorithms.md

Here, under Breadth First Search, under code:
MATCH (a {id: 723})-[edge_list:Type *bfs..10]-(b {id: 882}) RETURN *;
it says:
"The above query will find all paths of length up to 10 between nodes a and b"
and then goes on:
To find only the shortest path, simply append LIMIT 1 to the RETURN clause.
MATCH (a {id: 723})-[edge_list:Type *bfs..10]-(b {id: 882}) RETURN * LIMIT 1;

The Issue (I think) is:
BFS always returns only the shortest path, and LIMIT 1 has no effect there.

There is a broken link on the page https://memgraph.com/docs/memgraph/next/how-to-guides/encryption. In step 2 link Copy the SSL certificate inside of the Docker container doesn't work. The destination link should be https://memgraph.com/docs/memgraph/next/how-to-guides/work-with-docker#how-to-copy-files-from-and-to-a-docker-container.

In the reference guide https://memgraph.com/docs/memgraph/next/reference-guide/encryption the link move them into a Docker container has a wrong anchor, the correct one is how-to-copy-files-from-and-to-a-docker-container.

Also, should the sentence "If you are using Docker and want to use your own certificates, you need to move them into a Docker container in order to utilize them." If you are using Docker and want to use your own certificates, you need to copy them into a Docker container in order to utilize them."

It seems that EXPLAIN clause is missing in the Cypher manual. Should we add it?

The how-to guide for NetworkX in GQLAlchemy documentation should be written or deleted.

Add C++ in the first sentence and check if this needs to be updated somewhere else

https://memgraph.com/docs/memgraph/reference-guide/query-modules/implement-custom-query-modules/overview#python-api

In the case of a timeout, it will return an error to the user indicating that he should check the MAIN instance will return an error to the user proposing a check of REPLICAs' statuses as there might be network or hardware issues.

https://memgraph.com/docs/memgraph/reference-guide/replication

not he!

https://github.com/memgraph/docs/blob/master/docs/how-to-guides/replication.md

In "The cluster will consist of four nodes, one MAIN instance and two REPLICA instances.", should it be "consist of THREE nodes"?

One algorithm has multiple description points on MAGE docs. One is under ...docs/mage/query-modules/cpp/X and other part is on
.../docs/mage/algorithms/traditional-graph-analytics/X. Provide links from one to the other since no such link exists.

@vpavicic check if I described the issue correctly, and update if necessary

It would be great to add the reference to the orb's Github repository in the Lab/Playground documentation where graph visualization is mentioned.

In https://memgraph.com/docs/errors/memgraph/modules link mentioned in the Warning section is not in the Memgraph code. I've found the URL memgr.ph/modules.

In https://memgraph.com/docs/errors/memgraph/python-modules we have three errors, but two of them are the same.

The very first link in file CONTRIBUTING.md, about the Docusaurus documentation leads to "Page not found"

Make this stand out

https://memgraph.com/docs/memgraph/installation

Isolation levels are no longer a separate page; content related to isolation levels is now located on the Transactions page.

A redirect needs to be added from /reference-guide/isolation-levels to /reference-guide/transactions.

CSV load can go slow if you do not set up the index for MATCH/MERGE on nodes properties, we should consider adding hints about indexes in LOAD CSV docs. Consider Antonio recommendation on issue: memgraph/memgraph#684 (comment)
There were some similar issues on discord: https://discord.com/channels/842007348272169002/890969266600423504/1050748252078735410

Update the home page embed of the graph. It is showing too large nodes, probably due to the new GSS features.

https://memgraph.com/docs/memgraph/installation/direct-download-links
2.4.2 is updated, the previous versions need to checked and fixed

https://memgraph.com/docs/memgraph/connect-to-memgraph/drivers

Switch places to avaliable and quick start

Make languages be in columns or a table to avoid the vertical stretch

The algorithms overview page has a strange URL: https://memgraph.com/docs/mage/algorithms/algorithms.

At movie recommendation tutorial there are two mistakes:

1. Mistake:
This is not correct:
6. Average scores for first 10 movies:

MATCH (:User)-[:Rating]->(movie:Movie)
RETURN movie.title, avg(r.score) AS score
ORDER BY score DESC
LIMIT 10;

This would be correct (Missing variable name) :

MATCH (:User)-[r:Rating]->(movie:Movie)
RETURN movie.title, avg(r.score) AS score
ORDER BY score DESC
LIMIT 10;

2. Mistake:
This is not correct:
7. Create a new user and rate some movies:

MATCH (u:User {id:1000}), (m:Movie {title:'Mr. Holland's Opus'})

This would be correct (Double instead of single quotes):
MATCH (u:User {id:1000}), (m:Movie {title:"Mr. Holland's Opus"})

These two pages are Not Found:

1. https://memgraph.com/docs/memgraph/reference-guide/graph-algorithms
2. https://memgraph.com/docs/memgraph/under-the-hood/graph-algorithms

I get to 1. when I google "memgraph dfs" or "memgraph bfs" or "memgraph wshortest", it is the first result.
I believe it should be https://memgraph.com/docs/memgraph/reference-guide/built-in-graph-algorithms

The 2. is when I use our docs search bar for "graph algorithms", this is also the first result.
I see that when I click the link or go through google it redirects to https://memgraph.com/docs/memgraph/reference-guide/built-in-graph-algorithms, and that works, but the same redirection does not happend when using our search bar.

https://memgraph.com/docs/memgraph/reference-guide/query-modules/implement-custom-query-modules/custom-query-module-example#readable-procedures

extern "C" int mgp_init_module(struct mgp_module *module, struct mgp_memory *memory) {
  try {
    mgp::memory = memory;

    AddProcedure(RandomWalk, "get", mgp::ProdecureType::Read,
                 {mgp::Parameter("start", mgp::Type::Node), mgp::Parameter("length", mgp::Type::Int)},
                 {mgp::Return("random_walk", mgp::Type::Path)}, module, memory);
  } catch (const std::exception &e) {
    return 1;
  }

 return 0;
}

return 0; was missing

Add some information as to when and why to use --replication-replica-check-delay-sec configuration flag to the reference guide and how-to guide...
Buda has the most info

There is a mage-cugraph Docker image but we don't mention it in our documentation. We should add it. This image is referenced in https://developer.nvidia.com/blog/running-large-scale-graph-analytics-with-memgraph-and-nvidia-cugraph-algorithms/.

There should be a page explaining the newly implemented FOREACH clause here:
https://memgraph.com/docs/cypher-manual/clauses

We need to update the README file so that it shows that the minimal Node.js version is 16.

When creating docker volumes the first argument of the -v flag is the absolute path to the directory.
In the following link: https://memgraph.com/docs/memgraph/import-data/csv-import-tool#how-to-use-the-csv-import-tool we don't even explain what mg_import directory is, or where should it be created, or how a user should use it. mg_import and mg_lab should be absolute paths on the host machine, e.g. /home/medo/repos/memgraph/debug/import_directory. I think we should replace those instances with a clear documentation. mg_import would become <absolute-path-to-import-directory-on-host-machine>.

Link is missing at the end of https://memgraph.com/docs/memgraph/reference-guide/security
last sentence in streams category. Can be found in the

The way to solve it:
https://memgraph.com/docs/memgraph/reference-guide/streams#create-a-stream ere , right above the "Start a stream" header

Discord user is reporting outdated proces for building memgraph mage from source: https://memgraph.com/docs/mage/installation/source
https://discord.com/channels/842007348272169002/890968933040009276/1040888626315276309

The function Format() is not needed in this example:

@NodeStyle HasLabel?(node, "Country") {
  label: Format(Property(node, "name"))
}

Debt to images below and dependencies regarding documentation related to PR -> #536

Memgraph 2.4 is out so we need to update links on this page - https://memgraph.com/docs/memgraph/installation/direct-download-links.

https://memgraph.com/docs/memgraph/#power-up-with-mage

While working on GQLAlchemy, Boris noticed that not all MERGE clause options are documented.

Merging with ON CREATE SET and ON MATCH SET with the same properties set to the same value is actually the same as one SET clause:

Example - the following query:

MERGE (p:Person {name: 'Angela'})
ON CREATE SET p.notFound = TRUE
ON MATCH SET p.notFound = TRUE;

is the same as the below query:

MERGE (p:Person {name: 'Angela'})
SET p.notFound = TRUE;

Hence, by following the Cypher grammar, with MERGE you can have ON CREATE SET, ON MATCH SET, ON CREATE SET and ON MATCH SET, and just SET.

This is also possible:

MERGE (p:Person {name: 'Angela'})
    ON CREATE SET p.notFound = TRUE
    ON MATCH SET p.notFound = TRUE
SET p.notFound = FALSE

We should add a Disconr icon in SVG format next to the GitHub icon in our menu bar. The link should go to https://discord.gg/memgraph.