From mboxrd@z Thu Jan  1 00:00:00 1970
X-Msuck: nntp://news.gmane.io/gmane.emacs.gnus.user/2009
Path: news.gmane.org!not-for-mail
From: Hubert Canon <hcanon@free.fr>
Newsgroups: gmane.emacs.gnus.user
Subject: better gnus-thread-score-function
Date: Thu, 30 Jan 2003 15:24:41 +0000
Organization: Inforama, Sophia Antipolis
Message-ID: <85bs1yy81y.fsf.Hubert.Canon@free.fr>
NNTP-Posting-Host: main.gmane.org
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-15
Content-Transfer-Encoding: 8bit
X-Trace: sea.gmane.org 1138668593 13514 80.91.229.2 (31 Jan 2006 00:49:53 GMT)
X-Complaints-To: usenet@sea.gmane.org
NNTP-Posting-Date: Tue, 31 Jan 2006 00:49:53 +0000 (UTC)
Original-X-From: nobody Tue Jan 17 17:30:02 2006
Original-Path: quimby.gnus.org!newsfeed1.e.nsc.no!nsc.no!nextra.com!news.tele.dk!news.tele.dk!small.news.tele.dk!fr.usenet-edu.net!usenet-edu.net!proxad.net!feeder2-1.proxad.net!news3-1.free.fr!not-for-mail
Original-Newsgroups: gnu.emacs.gnus
X-Accept-Language: fr,fr-BE,en,en-UK,en-US,de,ar,tr,it
X-Troll: ya des chances...
X-Comment: membre du CAS, du GASDPQL, du MACQPAD et du MMQLPI
User-Agent: Gnus/5.090007 (Oort Gnus v0.07) Emacs/21.2
 (sparc-sun-solaris2.7)
X-Face: +6ROl)bD3}7=jo{T&\>oU+kz,Rj"Tpo5Sc\BG1$HVG!_Z05$,c"!Wf+f1U/FZ)fZdAA9<hQ
 j_{.57`j._')@f9cic{VsJzeyCYSY`?n~YG](J#<SGKD$QhEZErAnR5o1T_8f_SNl!t_c9F[a7=09!
 hrXwi0\1$H
Cancel-Lock: sha1:pn4WiyCwDO86lDIdTWmk2OPDhjg=
Original-NNTP-Posting-Date: 30 Jan 2003 15:25:31 MET
Original-NNTP-Posting-Host: 62.161.245.216
Original-X-Trace: 1043936731 news3-1.free.fr 1321 62.161.245.216:65063
Original-X-Complaints-To: abuse@proxad.net
Original-Xref: bridgekeeper.physik.uni-ulm.de gnus-emacs-gnus:2149
Original-Lines: 38
X-Gnus-Article-Number: 2149   Tue Jan 17 17:30:02 2006
Xref: news.gmane.org gmane.emacs.gnus.user:2009
Archived-At: <http://permalink.gmane.org/gmane.emacs.gnus.user/2009>


Hi all !

[please excuse me english, it is not my mother tongue]

I use adaptive scoring and I sort by total thread score. I didn't find
a suitable gnus-thread-score-function. I used to use '+ but it gives
too much importance to big threads, and 'max is not so good because
one good author in a bad thread lift it up too much. So I tried to do
my own : I wanted to give the thread the mean score.

Here is my try :

(defvar my-gnus-thread-score-function-divisor
;;; we hope that there will not be more than 100000 articles in a thread
  100000)
(defun my-gnus-thread-score-function-nb-articles (x)
  (floor (+ 0.5 (* (- x (floor x))
                   my-gnus-thread-score-function-divisor))))
(defun my-gnus-thread-score-function (article-score &rest subthread-scores)
;;; the result is the mean with the number of articles in the
;;; thread in the fractionnal part.
  (if subthread-scores
      (let ((total-article-number
             (apply '+ 1.0 (mapcar 'my-gnus-thread-score-function-nb-articles
                                   subthread-scores))))
        (+ (ceiling
            (apply '+ article-score subthread-scores)
            total-article-number)
           (/ total-article-number
              my-gnus-thread-score-function-divisor)))
    (+ (float article-score) (/ 1.0 my-gnus-thread-score-function-divisor))))
(setq gnus-thread-score-function 'my-gnus-thread-score-function)


I tested a bit, and it seems to give the expected value.

What do you think ?