From aeebce96914f191a6c0c42e0679e72dde431bed2 Mon Sep 17 00:00:00 2001 From: Paula Date: Mon, 30 Jun 2025 15:51:49 +0200 Subject: [PATCH] nodes vs vertices --- site/content/3.13/concepts/data-models.md | 24 ++++++---- site/content/3.13/graphs/_index.md | 56 ++++++++++++----------- 2 files changed, 44 insertions(+), 36 deletions(-) diff --git a/site/content/3.13/concepts/data-models.md b/site/content/3.13/concepts/data-models.md index 32fe05de7b..6f40bdaa06 100644 --- a/site/content/3.13/concepts/data-models.md +++ b/site/content/3.13/concepts/data-models.md @@ -45,23 +45,27 @@ Graphs are comprised of **vertices** and **edges**. Both are documents in ArangoDB. Edges have two special attributes, `_from` and `_to`, that reference the source and target vertices by their document identifiers. -You can store vertices and edges with as many properties as you need, as both +{{< info >}} +In ArangoDB, **vertices** are also referred to as **nodes**. +{{< /info >}} + +You can store nodes and edges with as many properties as you need, as both are fully-fledged documents (JSON objects). -You can organize vertices and edges in sets using collections, with vertices in +You can organize nodes and edges in sets using collections, with nodes in **document collections** (also referred to as **vertex collections**) and edges in **edge collections**. This [graph](../graphs/_index.md) model makes ArangoDB classify as a **Labeled Property Graph** store. The design with edges stored in edge collections enables true graph scalability, while keeping the promise of performant graph queries regardless of the number -of vertices and edges. +of nodes and edges. Edges are always **directed** in ArangoDB, which means they point from one -vertex to another. They cannot point both ways. However, you can create multiple -edges between a pair of vertices in both directions. When you **traverse** a -graph - a basic graph query algorithm that starts at a given vertex and then -walks along the connected edges to discover neighboring vertices - you can +node to another. They cannot point both ways. However, you can create multiple +edges between a pair of nodes in both directions. When you **traverse** a +graph - a basic graph query algorithm that starts at a given node and then +walks along the connected edges to discover neighboring nodes - you can specify whether you want to follow edges in the direction they are defined in (**outbound**), the opposite direction (**inbound**), or regardless of the direction (**any**). This means that you do not need to create an opposing edge @@ -69,14 +73,14 @@ for every edge that you want to be able to follow in both directions. Aside from basic graph traversal, ArangoDB offers [graph algorithms](../graphs/_index.md#supported-graph-algorithms) to find one -or multiple shortest paths between two vertices and can return a specified amount -of paths between two vertices in order of increasing length. +or multiple shortest paths between two nodes and can return a specified amount +of paths between two nodes in order of increasing length. You can perform operations directly on the documents of graphs and run graph traversals using ad-hoc sets of vertex and edge collections. These are called **anonymous graphs**. However, no graph consistency is enforced. You can create **named graphs** and use the interfaces for named graphs, which ensure graph -consistency. For example, removing a vertex removes all connected edges, too. +consistency. For example, removing a node removes all connected edges, too. Low-level operations can still cause dangling edges, nonetheless.