v1.1.1

MatrixLMnet v1.1.1

Diff since v1.1.0

• updated backtransform functions
• updated function to get one-standard-error lambda value
• updated tests
• updated ci.yml

Merged pull requests:

• v1.1.1 (#24) (@GregFa)
• v1.1.1 (#25) (@GregFa)

v1.1.0

MatrixLMnet v1.1.0

Diff since v1.0.2

• updated keyword syntax: hasXIntercept and hasZIntercept was renamed addXIntercept and addZIntercept in the mlmnet() function
• upgraded documentations
• added more testing functions

v1.0.2

Update dependencies versions in the Project.toml.

v1.0.1

Patch

• Fixing issue with mlmnet_cv_summary() function.
• Adding tests for cross validation summary functions
• Updating comments in the code.

v1.0.0

Major

• Elastic Net penalization has been implemented for the different methods: ista, fista, fista-backtracking, and admm.
• Package offers L1 and L2 penalized estimation of matrix linear models.
• Functions' arguments and keywords arguments have been upgraded.

Minor

• New tests for mlmnet( ) and mlmnet_cv( ) (i.e., for cross-validation) have been added.
• To stabilize positive definiteness in floating-point eigenvalues, singular value decomposition is now estimated.

Deprecations

• Keyword arguments isXIntercept and isZIntercept are deprecated, use hasXIntercept and hasZIntercept instead.
• Keyword argument isStandardize is deprecated, use toNormalize instead.
• Keyword arguments isXInterceptReg and isZInterceptReg are deprecated, use toXInterceptReg and toZInterceptReg instead.