Code

Direct Hybrid Factor Specification#1805

Code

Merged

varunagrawal merged 36 commits intodevelopborglab/gtsam:developfrom direct-hybrid-fgborglab/gtsam:direct-hybrid-fgCopy head branch name to clipboardSep 20, 2024

Commits

Commits on Aug 22, 2024

• remove extra imports in DiscreteBayesNet.cpp

  varunagrawal

  committed
  75d4724

  Copy full SHA for 75d4724

• minor edits

  varunagrawal

  committed
  dce5641

  Copy full SHA for dce5641

• rename X1 to X0 and X2 to X1

  varunagrawal

  committed
  9e77eba

  Copy full SHA for 9e77eba

• provide logNormalizers directly to the augment method

  varunagrawal

  committed
  3fc1019

  Copy full SHA for 3fc1019

• update GaussianMixture::likelihood to compute the logNormalizers

  varunagrawal

  committed
  cfef6d3

  Copy full SHA for cfef6d3

• Tests which verify direct factor specification works well

  varunagrawal

  committed
  30bf261

  Copy full SHA for 30bf261

• remove extra prints

  varunagrawal

  committed
  bfaff50

  Copy full SHA for bfaff50

• add docstring and GTSAM_EXPORT for ComputeLogNormalizer

  varunagrawal

  committed
  665d755

  Copy full SHA for 665d755

Commits on Aug 23, 2024

• Merge branch 'working-hybrid' into direct-hybrid-fg

  varunagrawal

  committed
  03e61f4

  Copy full SHA for 03e61f4

• compute logNormalizers and pass to GaussianMixtureFactor

  varunagrawal

  committed
  07a0088

  Copy full SHA for 07a0088

Commits on Aug 25, 2024

• Merge branch 'working-hybrid' into direct-hybrid-fg

  varunagrawal

  committed
  62b32fa

  Copy full SHA for 62b32fa

Commits on Sep 5, 2024

• Merge branch 'develop' into direct-hybrid-fg

  varunagrawal

  committed
  fbffd79

  Copy full SHA for fbffd79

• better comments

  varunagrawal

  committed
  13193a1

  Copy full SHA for 13193a1

• Merge branch 'develop' into direct-hybrid-fg

  varunagrawal

  committed
  0bab8ec

  Copy full SHA for 0bab8ec

Commits on Sep 6, 2024

• improved comments

  varunagrawal

  committed
  51a2fd5

  Copy full SHA for 51a2fd5

• some more refactor and remove redundant test

  varunagrawal

  committed
  615c04a

  Copy full SHA for 615c04a

• fix test

  varunagrawal

  committed
  9dc29e0

  Copy full SHA for 9dc29e0

• Merge branch 'develop' into direct-hybrid-fg

  varunagrawal

  committed
  05a4b7a

  Copy full SHA for 05a4b7a

• replace emplace_back with emplace_shared

  varunagrawal

  committed
  24ec30e

  Copy full SHA for 24ec30e

Commits on Sep 15, 2024

• Merge branch 'hybrid-error-scalars' into direct-hybrid-fg

  varunagrawal

  committed
  506cda8

  Copy full SHA for 506cda8

• fixes

  varunagrawal

  committed
  336b494

  Copy full SHA for 336b494

• fix tests

  varunagrawal

  committed
  de68aec

  Copy full SHA for de68aec

Commits on Sep 18, 2024

• Merge branch 'develop' into direct-hybrid-fg

  varunagrawal

  committed
  b895e64

  Copy full SHA for b895e64

• undo accidental rename

  varunagrawal

  committed
  987ecd4

  Copy full SHA for 987ecd4

Commits on Sep 19, 2024

• remove duplicate test and focus only on direct specification

  varunagrawal

  committed
  80d9a5a

  Copy full SHA for 80d9a5a

• relinearization test

  varunagrawal

  committed
  717eb7e

  Copy full SHA for 717eb7e

• print nonlinear part of HybridValues

  varunagrawal

  committed
  f875b86

  Copy full SHA for f875b86

• Merge branch 'develop' into direct-hybrid-fg

  varunagrawal

  committed
  2937533

  Copy full SHA for 2937533

• add documentation for additive scalar in the error and remove the 0.5 since it gets cancelled out during normalization

  varunagrawal

  committed
  9b6facd

  Copy full SHA for 9b6facd

• rename ComputeLogNormalizer to ComputeLogNormalizerConstant

  varunagrawal

  committed
  244661a

  Copy full SHA for 244661a

• fix docstring for HybridGaussianFactor

  varunagrawal

  committed
  4f88829

  Copy full SHA for 4f88829

• logNormalizationConstant is now a method for Gaussian noise model

  varunagrawal

  committed
  d60a253

  Copy full SHA for d60a253

• update tests

  varunagrawal

  committed
  cea0dd5

  Copy full SHA for cea0dd5

Commits on Sep 20, 2024

• hide sqrt(2*value) so the user doesn't have to premultiply by 2

  varunagrawal

  committed
  1ab82f3

  Copy full SHA for 1ab82f3

• logDetR method which leverages noise model for efficiency. Build logDeterminant and logNormalizationConstant on top of it.

  varunagrawal

  committed
  364b4b4

  Copy full SHA for 364b4b4

• comprehensive unit testing

  varunagrawal

  committed
  67a8b8f

  Copy full SHA for 67a8b8f