Recently,  I've used the R package RcppHMM to test Continuous HMM/GMM classification. For now, I faced a limitation, the possibility to use our personal dataset to train the models.

My questions :

- Is it possible to replace the "set.seed" instruction by a personal dataset with matrix length?

- Are there examples of the implementation of continuous HMM/GMM multi classifiers with alternative specified dataset?

- Is it possible to have the entire source code of the library?

An assistance can be useful.