v0.4.7

GPLikelihoods v0.4.7

Diff since v0.4.6

Merged pull requests:

• Remove outdated comment about Wikipedia (#102) (@simsurace)
• Add Aqua tests and clean up (#103) (@simsurace)

Closed issues:

• Release 0.4.5 has not been tagged (#94)

v0.4.6

GPLikelihoods v0.4.6

Diff since v0.4.5

Merged pull requests:

• update compat helper (#97) (@willtebbutt)
• Bump patch and compat (#101) (@willtebbutt)

v0.4.5

GPLikelihoods v0.4.5

Diff since v0.4.4

Merged pull requests:

• CompatHelper: bump compat for "Functors" to "0.3" (#91) (@github-actions[bot])

v0.4.4

GPLikelihoods v0.4.4

Diff since v0.4.3

Closed issues:

• Question on implementation and type stability (#77)

Merged pull requests:

• Some performance optimizations (#90) (@simsurace)

v0.4.3

GPLikelihoods v0.4.3

Diff since v0.4.2

Closed issues:

• NegativeBinomial likelihood enforces single parametrization (#69)

Merged pull requests:

• Fix different docs issues (#60) (@theogf)
• strict doc checks (#86) (@st--)
• CompatHelper: bump compat for "StatsFuns" to "1" (#87) (@github-actions[bot])

v0.4.2

GPLikelihoods v0.4.2

Diff since v0.4.1

Merged pull requests:

• Implement multiple parametrization for Negative Binomial (#80) (@theogf)

v0.4.1

GPLikelihoods v0.4.1

Diff since v0.4.0

Merged pull requests:

• CompatHelper: add new compat entry for "Zygote" at version "0.6" for package test (#74) (@github-actions[bot])
• export expected_loglikelihood (#75) (@st--)

v0.4.0

GPLikelihoods v0.4.0

Diff since v0.3.1

Closed issues:

• Move expected_loglik here & additional design discussion (#42)
• Categorical takes C-1 inputs for C classes (#55)
• Precompute the nodes and weights for GaussHermite (#71)

Merged pull requests:

• move expected_loglik here from ApproximateGPs.jl (#70) (@st--)
• less code duplication (#72) (@st--)

v0.3.1

GPLikelihoods v0.3.1

Diff since v0.3.0

Add the NegativeBinomialLikelihood

Merged pull requests:

• Add negative binomial likelihood (#63) (@simsurace)

v0.3.0

GPLikelihoods v0.3.0

Diff since v0.2.7

Breaking:

• The parametrization from GammaLikelihood went from {Talpha,Tlink} to {Tlink,Talpha}
• The default constructor for CategoricalLikelihood will return CategoricalLikelihood{BijectiveSimplexLink{SoftMaxLink}} instead of CategoricalLikelihood{SoftMaxLink}

Merged pull requests:

• BREAKING swap order of parametrization of GammaLikelihood (#41) (@st--)
• Rework categorical to allow multiple variants (#61) (@theogf)