Update CTLE function to CTLERob function. This new function have one more parameter ‘rlr_method’ which let user choose the robust regression method in ‘lmRob’,‘lmrob’,‘ltsReg’.
Update class definition of RobMixReg. The new class add one slot which return the posterior probability of the mixture regression.
Add method mixtureReg which let user run the mixture regression model flexibility by fixing the specific coefficient (slpoe value).
Add method Regression Based Suspace Learning (RBSL) that enable to fitting the mixture model for high dimension variable. The implementation is ‘CSMR’ function.
Add new wrapper function to deal with following four scenarios:
robust regression: one regression line and outliers.
flexible mixture regression: two regression lines without outliers. The flexible means that the coefficient of the regression line can be declare by the user.
robust mixture regression: two regression lines with outliers. The algorithm to solve this challenge is Component-wise Adaptive Trimming (CAT). the implementation is ‘CTLERob’ function. The details of the algorithm please refer our paper.
supervised subspace clustering: clustering heterogeneity objects into subgroup and selecting contributed attributes simultaneously. The algorithm to solve this challenge is RBSL. The details of the algorithm please refer our paper.
Add plot module for the above four scenarios.
Add two real dataset: Breast cancer multi-omic data and Two genes cytokine response data. The detail and example refer to package manual and vignette.