On Thursday, June 19, 2014 12:30:12 PM UTC+5:30, Ian wrote:
> On Wed, Jun 18, 2014 at 11:50 PM,   wrote:
> > Thank you for the reply. But as I checked it again I found,
> > f_prev[k] is giving values of f_curr[st] = e[st][x_i] * prev_f_sum
> > which is calculated later and again uses prev_f_sum.
> f_prev is the f_curr that was calculated on the previous iteration of
> the loop.  At each iteration after the first, the script calculates
> f_curr based on the value of f_prev, that is, the old value of f_curr.
> Then it reassigns the newly computed f_curr to f_prev, making it now
> the previous, and on the next iteration it creates a new dict to store
> the next f_curr.  Does that make sense?

Dear Group,

The logic seems going fine. I am just trying to cross check things once more,
so trying to generate the values and see on myself. 

I am trying to see this line,
prev_f_sum = sum(f_prev[k]*a[k][st] for k in states)

a[k][st], and f_prev[k] I could take out and understood.
Now as it is doing sum() so it must be over a list,
I am trying to understand the number of entities in the list, thinking whether 
to put len(), and see for which entities it is doing the sum.

Experimenting, if any one feels may kindly send any idea.

Subhabrata Banerjee. 

Reply via email to