Recent content by Arrebimbomalho

Thank you for the reference. I carefully went through it (I already knew a bit about regression and classification trees) and the RF is a natural extension. But only for RT and CTs, it seems. If you want to put together several sampled Markov chains, how would you do it? I was reading a bit on...

And the problem is in the optimization step...When it passes arguments to the msm.optim function, the entrance data is about 15 MB. But somehow it explodes in the function iteration and that´s when I get the dreaded RAM limitation message "Cannot Allocate Vector of size X GB"..

Well it´s panel data of about 6 million individuals observed over 20-25 time periods (it is unbalanced), so we´re talking dataframes with around 120 million rows

Ah ah, thanks :) I do realize its a bit long, but I thought it was important to provide context...I'll make a revised shorter version. As regards your last suggestion, that seems interesting. Its possible to develop a series of models for many samples, of course, but how would you aggregate the...

Hi everyone, My first post! :-) I apologize for how long it will be, but please bear with me. I am looking for help in a topic related to Markov chain modeling with covariates. I have a decent knowledge of the underlying theory, but not exactly how to implement some aspects of it in practical...