https://github.com/st235/hse.modernoperationsresearchmethods.graphcoloring
https://github.com/st235/hse.modernoperationsresearchmethods.graphcoloring
Last synced: 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/st235/hse.modernoperationsresearchmethods.graphcoloring
- Owner: st235
- License: mit
- Created: 2023-01-14T12:34:30.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2023-02-26T22:22:46.000Z (over 3 years ago)
- Last Synced: 2025-03-04T20:43:58.354Z (over 1 year ago)
- Language: C++
- Size: 4.42 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Graph Coloring
This is an implementation of [graph coloring](https://en.wikipedia.org/wiki/Graph_coloring) problem.
| Graph | Colored |
| ----- | ------- |
|||
## Build
[`graph_colorization.cpp`](./graph_colorization.cpp) is the main file and contains the implementation. To build a CLI application you need to run the command below:
```bash
g++ -std=c++17 -O3 graph_colorization.cpp
```
Please, pay attention that app requires **C++ 17**. You should use one of the [compilers that support](https://en.cppreference.com/w/cpp/compiler_support/17) the standart.
## Algorithm
Let me breifly describe the approach used in this project.
I am using [DSatur algorithm](https://en.wikipedia.org/wiki/DSatur) by Daniel Brélaz to colorize the graph.
DSatur is a heuristic greedy graph colouring algorithm.
As any greedy approach it is considering the verticies
in a specific order assigning to them the smallest possible color.
To do so the algorithm is dynamically (i.e. at every step) choosing the next vertex by picking a vertex
the largest number of different colors among the vertices adjacent to the one under consideration.
Once the vertex has been coloured, the algorithm considers ramining uncolored vertices. If there is a tie
between a few vertices the next step is to consider their degree in uncolored subgraph.
## Output
The output of one program run can be found in **below** or in [`report.csv`](./report.csv).
**Please, do pay attention, that `report.csv` also contains the solution!**
```bash
Instance Colors Time, sec
myciel3.col 4 0
myciel7.col 8 0.001
latin_square_10.col 130 0.141
school1.col 17 0.007
school1_nsh.col 28 0.005
mulsol.i.1.col 49 0.001
inithx.i.1.col 54 0.007
anna.col 11 0
huck.col 11 0
jean.col 10 0
miles1000.col 42 0.001
miles1500.col 73 0.002
fpsol2.i.1.col 65 0.004
le450_5a.col 10 0.002
le450_15b.col 17 0.003
le450_25a.col 25 0.003
games120.col 9 0
queen11_11.col 14 0.001
queen5_5.col 5 0
```
If you're interested in visualising colors you can take a look
at [`colors_report.txt`](./colors_report.txt). This file contains all
colors that have been assigned to the vertices in order of their appearance in
a file containing the graph.
## Misc
### Changes to the template
The most significant change made to the original template is the `Check` method.
Instead of considering `0` as **not a color**, I am considering `kColorNoColor = -1`.
I am using colors as an index in other arrays, therefore indexing them from `0` seems
more convenient.
### Visualisation
Additionally, to the `Check` method I prepared a small Python script
to visualise graphs. You can find the script at [graph_visualisation.py](./graph_visualisation.py).
You need [streamlit](https://pypi.org/project/streamlit/) and [steamlit_agraph](https://pypi.org/project/streamlit-agraph/) installed. To run visualisation script use the command below:
```bash
streamlit run graph_visualisation.py
```
If you see the interface like the interface below the script is running just fine. Hooray 🎉
### Additional examples
| Graph | Colored |
| ----- | ------- |
|||
|||
|||
Thank you for reading!