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

Solver tolerances do *not* scale with scalar precision #278

Closed
SeanCurtis-TRI opened this issue Apr 12, 2018 · 1 comment
Closed

Solver tolerances do *not* scale with scalar precision #278

SeanCurtis-TRI opened this issue Apr 12, 2018 · 1 comment

Comments

@SeanCurtis-TRI
Copy link
Contributor

Problem

The solvers GJKSolver_libccd and GJKSolver_indep use tolerances for the GJK and EPA algorithms to determine when the iterative algorithm should stop. These tolerances are hard-coded to a simple 1e-6 (see links above). This is an artifact from the former 32-bit floating point days.

The default values should depend on the underlying scalar type.

Proposal

Extend fcl::constants to provide some convenient, precision-dependent tolerance values and use those scalar-dependent functions to default initialize the solver values.

@sherm1
Copy link
Member

sherm1 commented Apr 12, 2018

👍

Quick survey: are people still using FCL in float precision?

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

No branches or pull requests

2 participants