LLRBTrees

Translated from Robert Sedgewicks original

A Red-Black Tree structure is an object for accessing ordered-associative arrays in at most 2 log₂N steps. Where N is the number of elements in the array.

Experimentally, this particular design requires on average log₂N - 0.5 steps for a successful search.

Installation

To install run the code:

    Pkg.clone("github.com/epaaso/LLRBTrees.jl")

This package uses very old julia. But for the spanish-speakers checkout the LLRB Trees explanation in the docs flder.

Instantiation

A tree object can be instantiated specifying the types for key and value:

    tree = LLRBTree{Int, ASCIIString}()

Or by specifying the values of the root:

    tree = LLRBTree(1,"a")

The tree LLRBTree is built of TreeNodes that are built the same way.

Manipulation

For indexing, addition and deletion, the methods from Base have been adopted:

tree = LLRBTree(1,"a")
push!(tree, 2, "b")
push!(tree, 3, "c")

    b       
/       \  
a        c     


delete!(tree, 1)

    c       
/           
b_r  


tree[2]

Nullable("b")

tree[2]="now"

    c       
/            
now_r

In addition to that, there is a method to return an ordered list from the elements in the tree:

ordereredpairs(tree)

4-element Array{Any,1}:
 [5,5]  
 [6,6]  
 [10,10]
 [15,15]

Visualization

There is also a separate module, that uses a slight modification of GraphViz.jl to make a graph plot of the trees. It can be found in https://github.com/epaaso/LLRBVisualize.jl.

Acknowledgements

Many thanks to David P. Sanders, researcher at UNAM, for tutoring this project. Also to Erik Jimenez for also translating the code.

Bibliography

Sedgewick, R. (2008, April). Left-leaning red-black trees. In Dagstuhl Workshop on Data Structures (p. 17).
Robert Sedgewick; Kevin Wayne (21 February 2011). Algorithms. Addison-Wesley Professional. ISBN 978-0-13-276256-4.

epaaso / LLRBTrees.jl