hidden_paths does not respect sequence length about seqhmm HOT 4 CLOSED

The hidden_paths function computes the most probable states also for the missing values, which is how unequal sequence lengths are handled internally. But indeed, it makes often sense to set the tails to voids for visualization etc. I'll add an an option for this.

from seqhmm.

In my case, and I suspect in many cases, they are not true "missing" observations. The sequences simply have different length. I will add a check in my code to reset these positions to void until you are able to add this option.

from seqhmm.

Yes this makes sense. I just pushed an update, the hidden_paths function now has an argument respect_void (default is TRUE) which replaces the hidden states with voids according to the observed sequences. Let me know if you encounter problems, I didn't fully test this yet.

from seqhmm.

Seems to be working fine. Thanks for the quick fix

from seqhmm.