Probability Convolutions 
Author Message
 Probability Convolutions

I am writing a program to calculate the nth convolution for a p.d.f.
My question actually addresses looping (simple enough) and
how to put all this stuff in one function.

If I define f1, f2, X, and P as :

  f1 =. ' +//. */~x.':11
  f2 =. ' +//. +/~x.':11
  Y =. 1 5 7
  P =. 0.1 0.2 0.7

Then each time the following three statements
are typed I get a convolution for the p.d.f

  index =. /: Y =. f2 Y
  Y =. index { Y
  P =. index { f1 P

This function should execute these three comands n times.

Any suggestions on how I should write it?

-----
(Actually there is no need to sort at each interation
but later I plan to add some code to remove duplicate
values in the Y vector.)
en



Mon, 14 Nov 1994 08:14:43 GMT  
 Probability Convolutions
Emmett McLean:
   I am writing a program to calculate the nth convolution for a
   p.d.f.

whatever that is...

   My question actually addresses looping (simple enough) and how to
   put all this stuff in one function.

   If I define f1, f2, X, and P as :
     f1 =. ' +//. */~x.':11
     f2 =. ' +//. +/~x.':11
     Y =. 1 5 7
     P =. 0.1 0.2 0.7

   Then each time the following three statements are typed I get a
   convolution for the p.d.f
     index =. /: Y =. f2 Y
     Y =. index { Y
     P =. index { f1 P

   This function should execute these three comands n times.
   Any suggestions on how I should write it?

Well, there's a couple ways that I can think of.  The first would be
to use an explicit definition:
t =. 0 0 $''    [.    l =. '0# t=:t,y.':''

l 'f1 =.   '' +//. */~x. '':11'
l 'f2 =.   '' +//. +/~x. '':11'
l 'Y  =.   1 5 7'
l 'P  =.   0.1 0.2 0.7'
l '$. =. ,3 # ,: $.'
l 'index =. /: Y =. Y =. f2 Y'
l 'Y =. index { Y'
l 'P =. index { f1 P'
l 'Y ,: P'
doit =. t : ''
doit ''

But this involves building up that explicit definition, which is kind
of akward to work with.

A second way of doing this would be to decide that Y and P will be
represented by the same piece of data.  For example, 'Y ,: P' which I
used in the explicit definition to be able to return both results.  If
I do this, I can define functions which operate on this data structure
to give Y and P:

 f1 =. ' +//. */~y.':11
 f2 =. ' +//. +/~y.':11
 Y  =. ' 0 { y.':11
 P  =. ' 1 { y.':11
 Y2 =. ' f2 Y y.':11
 index =. '/: Y2 y.':11
 Y3 =. ' (index y.) { Y2 y.':11
 P3 =. ' (index y.) { f1 P y.':11
 next =. Y3,:P3

Then you can do
   next ^: 3   (1 5 7 ,: 0.1 0.2 0.7)

And, yes, this involves some redundant calculations.  In principle,
this function (next) can be rearranged automatically to minimize
redundancy.  However, since no one has implemented that yet, we can do
it by hand.  To see how, let's expand 'next' out a few levels:
 next
 Y3,:P3





hook:


It might be nice to get rid of the dual grade-ups, but it may be more
trouble than it's worth.  Let's see... consider this as three
transformations, first extract Y and P and apply f1 and f2
respectively.  Second, duplicate that information and append the
gradeup of the f2 result.  Third, build the Y and P structure [for the
next pass, or for final consumption:



 pass3 =. (2&{  {  0&{) ,:  (2&{  {  1&{)

then, do
      next^:(3) 1 5 7 ,: 0.1 0.2 0.7

Presumably, for large enough arguments, this will be more efficient
than the simpler form,

Whether it's worth the trouble is, of course, a matter of opinion.

--

The U.S. government went another thousand million dollars into debt today.



Mon, 14 Nov 1994 14:18:10 GMT  
 
 [ 2 post ] 

 Relevant Pages 

1. Want convolution idioms

2. Convolution Solution

3. convolution

4. NumPy signaltools 0.20 - N-D convolution and N-D order filtering

5. convolution code

6. Convolution Subroutine

7. ? Fast algorithm for convolution integral

8. 2D convolution

9. N-D convolution and N-D order filtering with Numeric python

10. Looking for 3 dimensional FFT and/or convolution routines

11. Image convolution in TCL?

12. APL probability workspace wanted

 

 
Powered by phpBB® Forum Software