https://github.com/bhekimpilondhlela/graph-layout

This project is on graph theory and basic Fraud Detection using graphs. A method called Barycenter is used to achieve the minimum edge crossings, vertices are also arranged accordingly and also edges are colored accordingly.
https://github.com/bhekimpilondhlela/graph-layout

barycentric-coordinates edge-coloring edge-crossings graph-algorithms graph-layouts java java-graph-ranking

Last synced: 3 months ago
JSON representation

This project is on graph theory and basic Fraud Detection using graphs. A method called Barycenter is used to achieve the minimum edge crossings, vertices are also arranged accordingly and also edges are colored accordingly.

• Host: GitHub
• URL: https://github.com/bhekimpilondhlela/graph-layout
• Owner: BhekimpiloNdhlela
• License: mit
• Created: 2017-05-08T20:21:43.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2020-10-20T17:50:45.000Z (over 4 years ago)
• Last Synced: 2023-08-22T09:37:45.635Z (almost 2 years ago)
• Topics: barycentric-coordinates, edge-coloring, edge-crossings, graph-algorithms, graph-layouts, java, java-graph-ranking
• Language: Java
• Homepage:
• Size: 795 KB
• Stars: 1
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# NETWORK/ GRAPH LAYOUT

*********************************************************************************************************************

*********************************************************************************************************************

### Author		                :Bhekimpilo Ndhlela

### Name of Project           :Network/ Graph Layout

*********************************************************************************************************************

*********************************************************************************************************************

## Graphical User Interface Example:

### Text File Format

     0 -> 1

 

     1 -> 2

 

     2 -> 3

 

     3 -> 4

 

     7 -> 1

 

     1 -> 5

 

     3 -> 6

 

     2 -> 8

 

     2 -> 8

 

     1 -> 9

 

     10 -> 1

 

     10 -> 1

 

     10 -> 1

  

     10 -> 1

     8 -> 6

     8 -> 6

     8 -> 6

     NOTE: Some of the edges are repeated because i wanted to illustrate the coloring of the edges that occur the most.

     This is specified or briefly explained at the "What Happens:" Section, the last paragraph.

	

### Output Image of the Above Graph:

 ![Alt text](example_image.png)

### Note: Cyclic Graphs

       Please note that cyclic graphs or networks are not used in this algorithm, The algorithm works for uncyclic graphs or networks. If a cycle is

       detected in the graph at hand then the program or application will terminate and indicate that a cycle has been detected.

## Dependancies	:

Graph/ Network Layout depends on classes written by me(Bhekimpilo Ndhlela) which include:

       Layout.java (main/client library)

       GraphDummyNode.java

       GraphInitializer.java

       GraphNodeOrdering.java

       GraphOutput.java

       GraphRanking.java

       

       LayoutUnitTester.java(for coverage testing purposes)

It also depends on the sorting algorithms libraries from Princeton University, I rewrote some of them and edited most of them these algorithms include:

       Merge.java(inplace sorting algorithm for the BaryCenter method)

       AcyclicLP.java

       Bag.java

       DepthFirstOrder.java

 

       Digraph.java

       DirectedCycle.java

       DirectedEdge.java

       Queue.java

       

       Stack.java

       StdDraw.java

       Topological.java

       

  

## How and What To Compile:

This is a maven project.

Compliation: Run "mvn package" to produce a runnable jar. 

             Run "mvn -Dtest=LayoutUnitTester test" to run unit test

Execution: cd target

           java -jar Graph-Layout-1.0-jar-with-dependencies.jar ../testCases/SMALL\ GRAPHS/SpecSample.txt 

       

    

## What Happens(Briefly):

     The Algorithm has four major components. After having read the input file and creating the nodes and edges, it first ranks the nodes into different

     levels (Smin, Smax, S0, S1, ..., Sk). Secondly, it adds dummy nodes to the graph where edges stretch over rank sets. Thirdly, it orders the nodes 

     inside each level to minimize edge crossings. The last step is to find coordinate positions for each node and spline control points to draw all the

     edges between the nodes.

     Consider each rank set Si. For every node v in Si, consider its adjacent nodes w (in other ranked sets). If v is more than one rank away from any w,

     create a virtual (dummy) node t in rank set Si+1, and add a chain of virtual edges and nodes v-t0, t0-t1,..., tk-w, stepping through rank sets that

     are one rank apart.

     Draw the edges as straight lines, with bends at the virtual nodes. Do not draw the virtual nodes. Finally, edges are to be colourcoded to indicate 

     that they occur often (or not, if they occur only a few times). That is, you have to keep count of how often every edge occurs in the graph, and then

     colour the edges that occur the most red, the edges that occur second-most orange, and the edges that occur third-most dark yellow (make sure that the

     yellow is dark enough to be clearly visible on the screen). All other edges are to be drawn in black. For example, suppose there are five parallel

     edges between nodes 1 and 2, four edges between nodes 2 and 3, three edges between 1 and 3, and all other edges occur only once. Then you have to draw

     a single edge between 1 and 2, but colour it red, one edge coloured orange between 2 and 3, and one edge coloured dark yellow between 1 and 3. All other

     edges will be black.