Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Speedup parameter planner parameters assignment #1001

Open
wants to merge 19 commits into
base: production
Choose a base branch
from

Conversation

kanbouchou
Copy link
Collaborator

@kanbouchou kanbouchou commented Jun 28, 2021

Summary

  • In production, assignment of planner parameters takes hundreds of microsec.
  • In feature, it is faster by 10x, tens of microsec.
  • For some application, this difference accumulates to 1 second or more of planning time.

Production behavior

  • assignment operator was loading members in derived class by serialize->deserialize with xml format.
  • assignment operator was doing more than assignment, namely, it was parsing _sPostProcessingPlanner and _sPostProcessingParameters from _sExtraParameters

Feature behavior

  • assignment operator(=) copies members by down casting, without going through xml string.
  • parsing of _sPostProcessingPlanner and _sPostProcessingParameters are done explicitly by LoadExtraParameters

Discussion points

  • _ptarget of GraspSetParameters is copied by environment body index as it is the behavior in production. Do we have a better identifier such as name or id?

Future TODO

  • start using json string instead of xml.

@kanbouchou kanbouchou changed the title wip: Speedup parameter planner parameters assignment Speedup parameter planner parameters assignment Jun 29, 2021
@tgn3000
Copy link
Contributor

tgn3000 commented Jun 29, 2021

Back then I did have a version that uses only json to assign planner parameters between derived classes.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants