Skip to content

GitLab

  • Menu
Projects Groups Snippets
    • Loading...
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
  • Sign in / Register
  • F FEDOT
  • Project information
    • Project information
    • Activity
    • Labels
    • Members
  • Repository
    • Repository
    • Files
    • Commits
    • Branches
    • Tags
    • Contributors
    • Graph
    • Compare
  • Issues 87
    • Issues 87
    • List
    • Boards
    • Service Desk
    • Milestones
  • Merge requests 1
    • Merge requests 1
  • CI/CD
    • CI/CD
    • Pipelines
    • Jobs
    • Schedules
  • Deployments
    • Deployments
    • Environments
    • Releases
  • Monitor
    • Monitor
    • Incidents
  • Packages & Registries
    • Packages & Registries
    • Package Registry
    • Container Registry
    • Infrastructure Registry
  • Analytics
    • Analytics
    • Value stream
    • CI/CD
    • Repository
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Activity
  • Graph
  • Create a new issue
  • Jobs
  • Commits
  • Issue Boards
Collapse sidebar
  • ITMO-NSS-team
  • FEDOT
  • Issues
  • #737

Closed
Open
Created Jun 21, 2022 by Elizaveta Lutsenko@LizLutsenkoOwner

Add configurable InitialPopulation factory

Created by: gkirgizov

It will generalise the way of getting initial population for EvoGraphOptimiser.

Thoughts on this:

  • Currently optimiser accepts optional list of fixed initial individuals. Instead it would be more flexible to provide some kind of factory of initial population.
  • It's good to hide from GraphOptimiser details of how initial population is generated. This should hide self.initial_individuals and details in _init_population from the optimiser. Why so? Because for better separation of responsibilities the configuration of optimiser must happen outside of optimiser. It can be, for example, done in Composer.
  • We need to support ways to configure initial graph generation function OR initial graphs The class InitialPopulationBuilder is a step in this direction, and it already encapsulates much of functionality related to initial graph generation.
  • Drop InitialPopulationBuilder.with_mutation-- only Optimiser should really know about mutations or other evo operators
Assignee
Assign to
Time tracking