Skip to content

Uncertainty class #6

@willat343

Description

@willat343

There is demand for an "uncertainty" or "noise characteristics" class that catpures sigma/sigmas/variance/variances/covariance/precision/stiffness. The values can be precomputed and requires certain criteria (e.g. for stiffness we require covariance to be positive-definite unless we compute a pseudo-inverse).

This class can contain the functionality that currently exists in the stiffness and covariance files.

Currently the create of stiffness/covariance matrices from json is managed in sensorbox. Either:

  • sensorbox should be able to create instances of this uncertainty class from json, or
  • mathbox should inherit this functionality (adding a json dependency)
  • or all the json functionality should move to another package like convert

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions