Graphs vs Arrays, Maps, Tables, and Trees

A graph is not better than an array, map, table, or tree.

It answers a different kind of question.

That distinction matters because beginners often meet graphs as “the hard data structure” after learning the simpler ones.

That is the wrong framing.

Graphs are not the advanced version of arrays.

Graphs are what you use when the shape of the problem is relationships.

Array

An array says position matters.

items[0]
items[1]
items[2]

Use arrays for:

lists
sequences
queues
sorted data
ordered results

If your main question is “what comes next by position?”, an array is probably enough.

Map

A map says lookup matters.

const user = usersById.get("42");

Use maps for:

indexes
caches
direct access
deduplication
grouping

If your main question is “give me the thing with this key”, a map is probably enough.

Table

A table says records matter.

id | title | status | created_at

Use tables for:

records
filtering
aggregation
reporting
joins

Tables can store relationships too, but the raw table is not the mental model.

When you start asking how records connect over many hops, graph thinking starts showing up.

Tree

A tree says hierarchy matters.

root
  child
    grandchild

Use trees for:

folders
DOM nodes
ASTs
menus
parent-child ownership

A tree is actually a special kind of graph.

It is connected, has no cycles, and each child has one parent.

That is why trees feel simpler. They have strict rules:

one root
one parent per child
no cycles
exactly one path from root to any node

Many real systems refuse to stay that tidy.

A package can have many dependencies.

A user can have many friends.

A city can have many roads.

A web page can link to many pages and be linked from many pages.

When the world stops being a clean parent-child hierarchy, you are in graph country.

Graph

A graph says arbitrary relationships matter.

Use graphs for:

dependencies
routes
networks
flows
references
permissions
recommendations
state spaces

The graph question is usually not:

Where is this item stored?

The graph question is:

What is connected to this?
What can this reach?
What depends on this?
What is the shortest route?
Is there a loop?
What order is allowed?

That is the practical difference.

Other structures help you hold data.

Graphs help you follow relationships.