v0.9.0

LossFunctions v0.9.0

Diff since v0.8.1

Closed issues:

• No difference between WeightedMean and WeighedSum when normalize=false (#149)
• Revisiting MLJ <-> LossFunctions.jl integration (#157)

Merged pull requests:

• Remove ObsDim to generalize the code for iterables of observations (#158) (@juliohm)
• Revert order of arguments (#159) (@juliohm)

v0.8.1

LossFunctions v0.8.1

Diff since v0.8.0

Closed issues:

• Release tags on repo are not up to date with General.jl (#152)

Merged pull requests:

• remove unneeded generated function for OrdinalMarginLoss constructor (#153) (@vtjnash)
• Bump actions/cache from 1 to 3 (#154) (@dependabot[bot])
• Bump actions/checkout from 2 to 3 (#155) (@dependabot[bot])
• Bump codecov/codecov-action from 1 to 3 (#156) (@dependabot[bot])

v0.8.0

LossFunctions v0.8.0

Diff since v0.7.2

Closed issues:

• LogCosh Loss (#136)
• MethodError: objects of type L2DistLoss are not callable (#138)

v0.7.2

LossFunctions v0.7.2

Diff since v0.7.1

Closed issues:

• Cannot use LossFunctions in Julia 1.0 (#142)

Merged pull requests:

• Fix for Julia version < v1.3 (#143) (@juliohm)

v0.7.1

LossFunctions v0.7.1

Diff since v0.7.0

Closed issues:

• [docs] links dead (#141)

v0.7.0

LossFunctions v0.7.0

Diff since v0.6.2

Closed issues:

• Derivative of cross-entropy loss (#135)

Merged pull requests:

• Deprecate bang! variants towards a cleaner API (#133) (@juliohm)
• Deprecate MarginLoss + SparseArray in favour of CategoricalArray (#134) (@juliohm)
• Log cosh (#137) (@ven-k)

v0.6.2

LossFunctions v0.6.2

Diff since v0.6.1

v0.6.1

LossFunctions v0.6.1

Diff since v0.6.0

Closed issues:

• Support for categorical arrays (#117)

Merged pull requests:

• Reorganize codebase (#130) (@juliohm)
• Add support for CategoricalArrays.jl (#132) (@juliohm)

v0.6.0

LossFunctions v0.6.0

Diff since v0.5.1

Closed issues:

• Add ability to query domain of a loss function (#34)
• Add min and argmin to losses (#35)
• LowRankModels duplicate functionality (#86)
• Handling non-float targets (#87)
• Roadmap post 1.0 (#102)
• Probabilistic losses? (#108)
• Docs don't include all available loss functions (#111)
• Please don't export all of LearnBase (#112)
• logloss question (#114)
• Losses for multi class categorical variables (#118)
• Rename CrossentropyLoss to CrossEntropyLoss (#120)
• value versus evaluate (#121)

Merged pull requests:

• Remove space between docstring and thing (#106) (@JobJob)
• removes restriction on result to be an abstract array (#110) (@adam-r-kowalski)
• remove reexport (#113) (@joshday)
• Install TagBot as a GitHub Action (#116) (@JuliaTagBot)
• Add MisclassLoss as a generalization of ZeroOneLoss (#122) (@juliohm)
• Rename CrossentropyLoss to CrossEntropyLoss (#123) (@juliohm)
• Upgrade to Project.toml (#124) (@juliohm)
• Update docs (#125) (@juliohm)
• Stick to a single interface (#127) (@juliohm)
• Restart deprecation cycle (#128) (@juliohm)
• Update plot recipe (#129) (@juliohm)

v0.5.1: Merge pull request #106 from JobJob/space-out

Remove space between docstring and thing