Path: csiph.com!usenet.pasdenom.info!news.albasani.net!rt.uk.eu.org!newsfeed.xs4all.nl!newsfeed3a.news.xs4all.nl!xs4all!post.news.xs4all.nl!not-for-mail
Date: Sat, 10 May 2014 21:44:59 +0100
From: MRAB <python@mrabarnett.plus.com>
User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:24.0) Gecko/20100101 Thunderbird/24.5.0
MIME-Version: 1.0
To: python-list@python.org
Subject: Re: Question on Debugging a code line
References: <283e285b-4ab3-4ec7-a8be-4f1e047d9645@googlegroups.com>
In-Reply-To: <283e285b-4ab3-4ec7-a8be-4f1e047d9645@googlegroups.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.9866.1399754708.18130.python-list@python.org>
Lines: 70
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:71274

On 2014-05-10 20:27, subhabangalore@gmail.com wrote:
> Dear Room,
>
> I was trying to go through a code given in http://en.wikipedia.org/wiki/Forward%E2%80%93backward_algorithm[ Forward Backward is an algorithm of Machine Learning-I am not talking on that
> I am just trying to figure out a query on its Python coding.]
>
> I came across the following codes.
>
>>>> states = ('Healthy', 'Fever')
>>>> end_state = 'E'
>>>> observations = ('normal', 'cold', 'dizzy')
>>>> start_probability = {'Healthy': 0.6, 'Fever': 0.4}
>>>> transition_probability = {
>     'Healthy' : {'Healthy': 0.69, 'Fever': 0.3, 'E': 0.01},
>     'Fever' : {'Healthy': 0.4, 'Fever': 0.59, 'E': 0.01},
>     }
>>>> emission_probability = {
>     'Healthy' : {'normal': 0.5, 'cold': 0.4, 'dizzy': 0.1},
>     'Fever' : {'normal': 0.1, 'cold': 0.3, 'dizzy': 0.6},
>     }
>
> def fwd_bkw(x, states, a_0, a, e, end_st):
>      L = len(x)
>      fwd = []
>      f_prev = {} #THE PROBLEM
>      # forward part of the algorithm
>      for i, x_i in enumerate(x):
>          f_curr = {}
>          for st in states:
>              if i == 0:
>                  # base case for the forward part
>                  prev_f_sum = a_0[st]
>              else:
>                  prev_f_sum = sum(f_prev[k]*a[k][st] for k in states) ##
>
>              f_curr[st] = e[st][x_i] * prev_f_sum
>
>          fwd.append(f_curr)
>          f_prev = f_curr
>
>      p_fwd = sum(f_curr[k]*a[k][end_st] for k in states)
>
> As this value was being called in prev_f_sum = sum(f_prev[k]*a[k][st] for k in states marked ##
> I wanted to know what values it is generating.
> So, I had made the following experiment, after
> for i, x_i in enumerate(x):
> I had put print f_prev
> but I am not getting how f_prev is getting the values.
>
> Here,
> x=observations,
> states= states,
> a_0=start_probability,
> a= transition_probability,
> e=emission_probability,
> end_st= end_state
>
> Am I missing any minor aspect?
> Code is running fine.
>
> If any one of the esteemed members may kindly guide me.
>
The values calculated in the inner loop are being put into the dict 
'f_curr'
and then, when that loop has completed, 'f_prev' is being bound to that
dict.

'f_curr' is bound to a new dict just before the inner loop, ready for
the new values.