You will find loads in C++ and Java. JGap is good. Entirely depends on what is your requirement in GA implementation what level and type of mutation, crossover, selection, representation strategies need to be available, etc. just google for genetic algorithms alone you will find alot in your search. You might also need to be careful of the licenses some of them are only available for research purposes only and will not scale well in production environment. Your best bet is to implement your own - some of these libraries work like CMS specfic to specific contexts and requirements they are unable to cater to everyones needs.
GA is a global optimization strategy so they can be slow to work with - I wouldn't advise you to implement such algorithms in python unless you are using externally compiled libraries - a better language would be Java or C++.
The library supports several kinds of populations, generation strategies, selections, crossovers, mutations, reinsertions and terminations. There are a lot of these classic methods already implemented, like: roulette wheel selection, tournament selection, OX1 crossover, PMX crossover, RSM mutation, uniform mutation, elitist reinsertion, pure reinsertion, time evolving termination, fitness stagnation termination, etc.
You can extend the library for your need just implementing new classes of the available interfaces of classic GAs method.
DEAP is a python package that allows for all kinds of evolutionary algorithm including the genetic algorithm. It is built as more of a framework which allows you to easily plug in your own selection, crossover and mutation functions.
发布评论
评论（5）
Pyevlove
来自网站：
Pyevlove
From the website:
您会发现 C++ 和 Java 中的负载。 JGap不错。
完全取决于您在 GA 实现中的要求是什么，需要提供什么级别和类型的突变、交叉、选择、表示策略等。只需谷歌搜索遗传算法，您就会在搜索中找到很多内容。您可能还需要注意许可证，其中一些许可证仅可用于研究目的，并且在生产环境中无法很好地扩展。您最好的选择是实现您自己的 - 其中一些库像 CMS 一样工作，专门针对特定的上下文和要求，但它们无法满足每个人的需求。
GA 是一种全局优化策略，因此它们的工作速度可能很慢 - 我不建议您在 python 中实现此类算法，除非您使用外部编译的库 - 更好的语言是 Java 或 C++。
You will find loads in C++ and Java. JGap is good.
Entirely depends on what is your requirement in GA implementation what level and type of mutation, crossover, selection, representation strategies need to be available, etc. just google for genetic algorithms alone you will find alot in your search. You might also need to be careful of the licenses some of them are only available for research purposes only and will not scale well in production environment. Your best bet is to implement your own - some of these libraries work like CMS specfic to specific contexts and requirements they are unable to cater to everyones needs.
GA is a global optimization strategy so they can be slow to work with - I wouldn't advise you to implement such algorithms in python unless you are using externally compiled libraries - a better language would be Java or C++.
我使用 Pygene 这真的很好。
I use Pygene which is really good.
GeneticSharp 对于 C# 来说是一个不错的选择。
该库支持多种群体、生成策略、选择、交叉、突变、重新插入和终止。这些经典方法已经实现了很多，比如：轮盘选择、锦标赛选择、OX1交叉、PMX交叉、RSM突变、均匀突变、精英重插入、纯重插入、时间演化终止、适应度停滞终止等
。只需实现经典 GA 方法的可用接口的新类即可根据您的需要扩展库。
该库还支持 .NET 和 Mono。
GeneticSharp is a good option for C#.
The library supports several kinds of populations, generation strategies, selections, crossovers, mutations, reinsertions and terminations. There are a lot of these classic methods already implemented, like: roulette wheel selection, tournament selection, OX1 crossover, PMX crossover, RSM mutation, uniform mutation, elitist reinsertion, pure reinsertion, time evolving termination, fitness stagnation termination, etc.
You can extend the library for your need just implementing new classes of the available interfaces of classic GAs method.
The library also supports .NET and Mono.
DEAP 是一个 Python 包，允许使用包括遗传算法在内的各种进化算法。它更像是一个框架，允许您轻松插入自己的选择、交叉和变异功能。
https://github.com/deap/deap
DEAP is a python package that allows for all kinds of evolutionary algorithm including the genetic algorithm. It is built as more of a framework which allows you to easily plug in your own selection, crossover and mutation functions.
https://github.com/deap/deap