I know it’s just hot air until I actually do it but when/if I find the time I’d like to do a Knapsack example.
maz
seanmor5
Author of Genetic Algorithms in Elixir
Thanks for putting this in here! I’m trying to formulate how I can seamlessly integrate this into my library/if it’s even worth it.
Anybody have any thoughts?
0
seanmor5
Author of Genetic Algorithms in Elixir
Check out the Knapsack example I just pushed to master. I think it’s got some flaws. The optimal answer is 15 I believe. Let me know if you can improve it!
Also beware that the code might look slightly different as I’m working through some changes with the library right now.
Stay tuned!
1
dokie
I have submitted a small PR for you to examine the Knapsack problem. It is of course sensitive to the pseudo-random nature of the algorithm and for the test problems it sometimes finds the optimum and others times not.
3
seanmor5
Author of Genetic Algorithms in Elixir
A quick decision I would like some input on.
I have rewritten the way in which algorithms are configured in the library. The new way allows for configuration outside of the module which in turn supports the ability to configure the same problem with different parameters for easier comparison. The new style is also more “Elixir-y” in that it supports piping configuration options into the run function.
One thing I like about this is that it can help separate the algorithm from the problem definition. I’ve always viewed the implementation module as your definition and encoding of your problem and I believe this style does a better job separating the “how” from the “what.” That being said, this might not always make sense in cases where some parts of the algorithm are included in the implementation module (like a custom crossover method).
Basically, my question is: does it make sense to require configuration be done OUTSIDE of the implementation module rather than inside? Is there room for both configuration styles? Which makes more sense to you?
Old Style
defmodule MyGA do
use Genex, crossover_type: :two_point
...other opts...
New Style
defmodule MyGA do
use Genex
...Implementation...
end
[population_size: 1000]
|> use_crossover(:two_point)
|> use_selection(:random)
|> MyGA.run()
2
dokie
My preference after playing around with the code is the New Style. There would be a temptation to (falsely) tinker with the basic algorithm parts in order to “tune” a particular problem and tie its formulation too much to the solution approach. Other optimisation libraries keep the problem definition separate from the algorithm chosen to solve that problem.
1
seanmor5
Author of Genetic Algorithms in Elixir
Hey guys, sorry I haven’t been able to put much time into this library due to school; however, I have a quick announcement and update.
First, I added a new function interactive/2 up on the master branch for creating Interactive Genetic Algorithms. Basically, you call interactive/2 in your fitness_function/1 definition.
The function accepts chromosome and view which is a function for displaying the Chromosome struct. It defaults to IO.inspect.
This came from a paper I read on Web Design Optimization using GAs. Basically, each Chromosome is displayed to a user and THEY determine the fitness based on some arbitrary criteria. In the case of web pages, it was on how appealing the page looked.
This concept was pretty cool to me so I added the functionality to Genex. You can see the example in the examples directory.
v1.0 Update
I am pushing towards v1.0 slowly but surely. This will include major restructuring of the library and introduce a bunch of new and exciting concepts including Multi-Objective Optimization, Multiple Populations, Customizable Evolutions, different reinsertion strategies, etc. I’m also working on a Web Visualizer of GAs using Phoenix Live View.
I expect v1.0 to be pretty full featured and well-tested when it’s released, and because my time is limited with school, I don’t anticipate it will be out for another month or two.
I plan to use this library for some Academic Research (not just trivial examples like I have been), so expect some pretty cool things in the near future!
If anybody wants to help out, let me know!
6
seanmor5
Author of Genetic Algorithms in Elixir
Genex v1.0.0-beta Released
Hey everbody, I’m not sure if anybody actually uses this library, but I have a ton of fun developing it and I’ve used it for some research projects too.
Anyway, I’ve just released the beta version of v1.0.0. It’s beta because there’s not much documentation or testing done, but the design of the library will pretty much remain the same. v1.0 has some major breaking changes, but it’s super powerful and I’m really excited about what you can do with Genex right now.
What’s New
- Changed
encoding/0togenotype/0. Semantic change for correctness. - New Crossover, Selection, Mutation Operators.
- Support for Multi-Objective Optimization
- Ability to fully customize every step of an evolution.
- Hall of Fame to track best solutions.
- Ability to save and load evolutions.
- Ability to represent solutions as any Enum datatype.
- Addition of several convenient generation functions.
- Addition of several benchmark functions for GP, SOC, MOC, Binary GAs, etc.
- Addition of penalty functions for constraint problems.
- Addition of interactive genetic algorithms.
- Ability to create custom visualizations through visualization behaviour.
What’s Next
I have a lot of really cool things planned, but my time is pretty off and on with school and other obligations. I plan to start posting about some of the awesome things you can do with Genex soon, time willing.
If anybody wants to see anything in particular added in, or has any questions about the library, please feel free to reach out to me. I hope somebody out there finds some use out of this, if only to mess around and have fun!
3
maz
I probably already told you this buy I did a genetic algorithm in university for the backpack problem (in Java, don’t laugh that’s all I knew at the time) and it was a ton of fun. I truly think the potential of genetic algorithms has yet to be realized. Thanks for this.
1
seanmor5
Author of Genetic Algorithms in Elixir
I agree! I believe Genetic Algorithms will play a hand in some ground breaking developments in the future - they’re so simple, but they can do so much.
Most of this library is for my own research and discovery. I find things in papers, other libraries, etc. that I like and try to implement them. I’m happy other people get some enjoyment out of it as well.
0