DEAP 0.8.2 documentation

Welcome! This is the documentation for DEAP 0.8.2, last updated Feb 06, 2013.

DEAP (Distributed Evolutionary Algorithms in Python) is a novel evolutionary computation framework for rapid prototyping and testing of ideas. Its design departs from most other existing frameworks in that it seeks to make algorithms explicit and data structures transparent, as opposed to the more common black box type of frameworks. It also incorporates easy parallelism where users need not concern themselves with gory implementation details like synchronization and load balancing, only functional decomposition.

The five founding hypotheses of DEAP are:

  1. The user knows best. Users should be able to understand the internal mechanisms of the framework so that they can extend them easily to better suit their specific needs.
  2. User needs in terms of algorithms and operators are so vast that it would be unrealistic to think of implementing them all in a single framework. However, it should be possible to build basic tools and generic mechanisms that enable easy user implementation of most any EA variant.
  3. Speedy prototyping of ideas is often more precious than speedy execution of programs. Moreover, code compactness and clarity is also very precious.
  4. Even though interpreted, Python is fast enough to execute EAs. Whenever execution time becomes critical, compute intensive components can always be recoded in C. Many efficient numerical libraries are already available through Python APIs.
  5. Easy parallelism can alleviate slow execution.

And these hypotheses lead to the following objectives:

Rapid prototyping
Provide an environment allowing users to quickly implement their own algorithms without compromise.
Parallelization made easy
Allow for straightforward parallelization; users should not be forced to specify more than the granularity level of their functional decomposition.
Preach by examples
Although the aim of the framework is not to provide ready made solutions, it should nevertheless come with a substantial set of real-world examples to guide the apprenticeship of users.

Parts of the documentation:

Indices and tables:

Meta information:

Docs for other versions

Other resources