[Biopython-dev] [Bug 2947] Bio.HMM calculates wrong viterbi path

bugzilla-daemon at portal.open-bio.org bugzilla-daemon at portal.open-bio.org
Fri Feb 4 10:05:18 UTC 2011


http://bugzilla.open-bio.org/show_bug.cgi?id=2947





------- Comment #6 from biopython-bugzilla at maubp.freeserve.co.uk  2011-02-04 05:05 EST -------
(In reply to comment #4)
> 
> Looks like Peter's fix treats all starting states as equally probable,
> there is no longer a special start state. That's reasonable, although the
> coding is a little confusing:
>

It was Phillip's fix.

(In reply to comment #5)
> FWIW, I think the right thing with respect to begin states is to require the
> user to explicitly specify an begin state in the state alphabet, e.g.:
> class coin:
>     def __init__(self):
>         self.begin_state_name = "begin"
>         self.letters = ["u", "f"]

If we go that route, we'll need to make very clear the differences between a
HMM Alphabet (of states) and a biological sequence alphabet (Bio.Alphabet). It
must make sense in many cases to use a biological sequence alphabet, but in
general adding HMM attributes to the class does not make sense.

We really need someone to volunteer to take over this code (and sort out the
overlap between Bio.MarkovModel and/or Bio.HMM.MarkovModel), write some
documentation for the tutorial, and sort out these remaining issues. Are either
of you interested?

> 
> I agree the variable names "main_state" and "cur_state" are confusing and
> should be changed.
> 

I'll happily merge/cherry-pick a simple diff to do that only if you do that on
github, or apply a patch if you upload it here.

Thanks,

Peter


-- 
Configure bugmail: http://bugzilla.open-bio.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.



More information about the Biopython-dev mailing list