[Spambayes] How many statistics are considered as good?

[Harry Sigerson]

Dear Spambayes, 
     I am using the March '05, v1.0.4
     'SpamBayes' has been working away so well, thank you. 
     I am using an email agent called 'Virtual Access', it is of the 
'normal' POP3/SMTP sort. 
     
     I think I should be doing something to help it. 
     I think it may be handling too much in the way of statistics or if 
you like more than it needs to handle. 
     There are 14,209 spam. 
     There are 8,623  good messages. 
     When I look in the C:\Program Files\SpamBayes\bin there are 10# 
folders; 3# .dll; 6# .exe; and one called   sb_tray.exe.log . I've had 
a look at that last file but I am none the wiser. 
     
     I believe that these both the ham and the spam or at least the 
spam(?) should be reduced in size; that the size of these files may be 
slowing things down. I have the impression that there is a hesitation 
at the start of SpamBayes working cycle. 
     This probably is done by deleting a file or files or an early part 
of a file(s). 
     
     Am I correct in this? If so which files should be deleted?
     
     I have read the section 4.6 "Do I need to keep the spam after it 
has been trained?" It explains what SpamBayes does and the minimum it 
needs to do it but does not tell me what to do to reduce things to that 
minimum. 
     
     I'll continue to look; the info is no doubt there in the web site 
somewhere. A quick clue would be appreciated. 
          
     Regards, 
     Harry Sigerson.