[
https://issues.apache.org/jira/browse/MATH-178?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Paul Cowan updated MATH-178:
----------------------------
Attachment: ZipfDistribution2.patch
Attached is a version which addresses the test failure (note to self: rerun the
tests after that 'one last tidy-up before submitting the patch, that's what
they're FOR!)
1) No real opinion as to whether this is 1.2-worthy or should wait for 2.0. We
have no pressing need for it to be included; if it's ready + everyone's happy
with it in time for 1.2, well and good, if not, then 2.0 is fine.
2) I will move the harmonic stuff into its own class. Should this just live in
o.a.commons.math.special, or is there a more suitable place?
3) I'll admit I don't know a better algorithm. I'm working on it, but the math
is WAY over my head. Wikipedia gives a generating function
http://upload.wikimedia.org/math/4/2/b/42bbc9c32faf3beee2f6920de77fb770.png at
("where Lim(z) is the polylogarithm, and | z | < 1. ") but my math skills are
so atrophied that working out whether that can be converted into an easy
generating algorithm (and if so, how) is a little beyond me. There's some
interesting info at http://www.research.att.com/~njas/sequences/A001008 -- it
looks like Knuth covers harmonic numbers in SOME way in TAOCP, volume 1, but I
don't have a copy to hand.
I'll keep looking at this; if no such number can be found precomputing a lookup
table may be the best bet (perhaps multiple implementations of ZipfDistribution
could exist; one which does it 'naively' but slowly, one which caches an array
of size [numberOfElements] on first access (speed at the cost of memory)
> Implement Zipf distribution
> ---------------------------
>
> Key: MATH-178
> URL: https://issues.apache.org/jira/browse/MATH-178
> Project: Commons Math
> Issue Type: New Feature
> Reporter: Paul Cowan
> Priority: Minor
> Attachments: ZipfDistribution.patch, ZipfDistribution2.patch
>
>
> Per my post to commons-dev --
> http://mail-archives.apache.org/mod_mbox/commons-dev/200801.mbox/[EMAIL
> PROTECTED] -- I'd like to propose adding a Zipf distribution to commons-math.
> I have a patch, but it's incomplete and somewhat inefficient; I'd like to
> throw it up for discussion though. To follow.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
