LeftChildRightSiblingTrees.jl

Memory-efficient representation of a tree with arbitrary number of children/node
Author JuliaCollections
Popularity
9 Stars
Updated Last
9 Months Ago
Started In
January 2020

LeftChildRightSiblingTrees

A left child, right sibling tree (frequently abbreviated as "LCRS") is a rooted tree data structure that allows a parent node to have multiple child nodes. Rather than maintain a list of children (which requires one array per node), instead it is represented as a binary tree, where the "left" branch is the first child, whose "right" branch points to its first sibling.

Concretely, suppose a particular node, A, has 3 children, a, b, and c. Then:

  • a, b, and c link to A as their parent.
  • A links a as its child (via A's left branch); a links b as its sibling (via a's right branch), and b links c as its sibling (via b's right branch).
  • A's right branch would link to any of its siblings (e.g., B), if they exist.
  • Any missing links (e.g., c does not have a sibling) link back to itself (c.sibling == c).

Tradeoffs

An LCRS tree is typically more memory efficient than an equivalent multi-way tree representation that uses an array to store the children of each node. However, for certain tasks it can be less performant, because some operations that modify the tree structure require iterating over all the children of a node.

Demo

Creating a Tree

Can addchild or addsibling.

julia> using LeftChildRightSiblingTrees

julia> mum = Node("Mum");

julia> me = addchild(mum, "Me");

julia> son = addchild(me, "Son");

julia> daughter = addchild(me, "Daughter");

julia> brother = addsibling(me, "Brother");  # equivalent: to `addchild(mum, "Brother")`

Querying about nodes:

julia> lastsibling(me)
Node(Brother)

julia> isroot(mum)
true

julia> isleaf(me)
false

julia> isleaf(daughter)
true

Iterating the Tree/Nodes

Iteration goes through all (direct) children. The .data field holds the information put in the tree. we can use this to draw a simple visualization of the tree via recursion.

julia> for child in mum
           println(child)
       end
Node(Me)
Node(Brother)

julia> function showtree(node, indent=0)
           println("\t"^indent, node.data)
           for child in node
               showtree(child, indent + 1)
           end
       end
showtree (generic function with 2 methods)

julia> showtree(mum)
Mum
        Me
                Son
                Daughter
        Brother

LeftChildRightSiblingTrees also has a built in function for showing this kind of info:

julia> LeftChildRightSiblingTrees.showedges(mum)
Mum has the following children: Me    Brother
Me has the following children: Son    Daughter
Son has no children
Daughter has no children
Brother has no children

Manipulating the tree

See the docstrings for graftchildren! and prunebranch!.

Credits

This existed as an internal component of ProfileView since its inception until it was split out as an independent package.