Re: Training frustration

Anne Wilson <[email protected]>
Newsgroups gmane.mail.bogofilter.general
Message-ID <[email protected]>
On Monday 11 February 2008 11:32, Pavel Kankovsky wrote:
> On Mon, 11 Feb 2008, Anne Wilson wrote:
> > Total   messages: 164
> >
> > Total        ham: 3
> > Misdetected  ham: 0
> >     retrain fail: 0
> >
> > Total       spam: 19
> > Misdetected spam: 92
> >     retrain fail: 92
>
> You've got 19 messages in the spam directory. This number is ok, according
> to the output of ls. The number of messages in the ham directory, i.e. 3,
> is probably ok as well. The other numbers are pretty suspicious. 92 failed
> messages out of 19? 164 total messages?
>
> It seems you removed old messages from the spam folder having had trained
>
> them:
> > I have made sure that the folders were compacted after deleting old,
> > trained, messages.
>
> but did not told trainbogo.sh to clean it stats dir ("stats.tmp" in the
> current working directory, probably your home dir, by default) and it
> still remembers them and makes hopeless attempts to recheck them every
> time you run it. Try running trainbogo.sh -c (if you really want to delete
> messages that have already been used for training).
>
> > ls -l /home/anne/Maildir/.INBOX.bogotrain_spam/cur/
> > total 160
> >
> > Total 160?
>
> 160 blocks occupied by files in the directory. It is probably ok.
>
That makes sense, and fits with what I suspected, but didn't know how to 
handle.

You seem surprised that I was cleaning out the 'trained' messages.  I thought 
it was a bad idea to keep running the same messages through the training.  Am 
I wrong?

I'll keep the -c parameter in the command in future.  What is the reason for 
it not being the default?

Anne
_______________________________________________
Bogofilter mailing list
[email protected]
http://www.bogofilter.org/mailman/listinfo/bogofilter
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.