v0.10.0

MLJLinearModels v0.10.0

Diff since v0.9.2

Merged pull requests:

• first pass at Gramian training for OLS (#146) (@adienes)
• Minor release with Gramian training for OLS (#151) (@tlienart)

Closed issues:

• should scale_penalty_with_samples = true be defualt? (#149)

v0.9.2

MLJLinearModels v0.9.2

Diff since v0.9.1

Closed issues:

• Improve documentation around available solvers (#78)
• Testing quantreg vs MLJLinearModels (#147)

Merged pull requests:

• Fix a doc typo (#144) (@ablaom)
• Fixes for quantile regression (#148) (@tlienart)

v0.9.1

MLJLinearModels v0.9.1

Diff since v0.9.0

Closed issues:

• Help with documentation review (#136)

Merged pull requests:

• Create MLJ compliant doc strings (#138) (@ablaom)
• Fix MLJ docstrings (#141) (@ablaom)

v0.9.0

MLJLinearModels v0.9.0

Diff since v0.8.1

Merged pull requests:

• Optim defaults doc (#139) (@olivierlabayle)

v0.8.1

MLJLinearModels v0.8.1

Diff since v0.8.0

Merged pull requests:

• more flexible optim options (#133) (@jbrea)
• docs for new options in Optim solvers (#134) (@jbrea)
• Some minor doc improvements (#137) (@ablaom)

v0.8.0

MLJLinearModels v0.8.0

Diff since v0.7.1

• (breaking) Change the default value of the regularisation parameter lambda in LogisticClassifier from 1.0 to eps() (#131)

Closed issues:

• Unexpected behaviour with LogisticClassifier (#131)

Merged pull requests:

• For a 0.8.0 release (#132) (@ablaom)

v0.7.1

MLJLinearModels v0.7.1

Diff since v0.7.0

Closed issues:

• predict throws an error for MultinomialClassifier on crabs dataset (#129)

Merged pull requests:

• closes #129 by fixing the interface after #123 (#130) (@tlienart)

v0.7.0

MLJLinearModels v0.7.0

Diff since v0.6.5

Closed issues:

• TagBot not working (#126)

v0.6.5

MLJLinearModels v0.6.5

Diff since v0.6.2

Closed issues:

• Multinomial regressor requires integer categories, and greater than zero (#58)
• Bug: LassoRegressor from MLJLinearModels returns null vector but the one from ScikitLearn does not (#81)
• Documentation generation is failing (#116)
• LogisticClassifier lambda and gamma are easy to confuse (#122)
• Problems when only one class manifest in the training target (#123)
• What is the new scale_penalty_with_samples=true doing? (#124)

Merged pull requests:

• Extend compat for DocStringExtensions and relax test (#119) (@ablaom)
• For a 0.6.2 release (#120) (@ablaom)
• Clarify LogisticClassifier.lambda docstring (#121) (@rikhuijzer)
• closes #123 (#125) (@tlienart)

v0.6.4

MLJLinearModels v0.6.4

Diff since v0.6.2

Closed issues:

• Multinomial regressor requires integer categories, and greater than zero (#58)
• Bug: LassoRegressor from MLJLinearModels returns null vector but the one from ScikitLearn does not (#81)
• Documentation generation is failing (#116)
• LogisticClassifier lambda and gamma are easy to confuse (#122)

Merged pull requests:

• Extend compat for DocStringExtensions and relax test (#119) (@ablaom)
• For a 0.6.2 release (#120) (@ablaom)
• Clarify LogisticClassifier.lambda docstring (#121) (@rikhuijzer)
• closes #123 (#125) (@tlienart)