Re: Why is my pR always the same?

On Tue, Oct 07, 2008 at 10:18:21AM -0700, Robert Swirsky-Warner wrote:

>    1. Mac OS 10.5
>    swirsky-tppllc:~ swirsky$ uname -a
>    Darwin swirsky-tppllc.local 9.5.0 Darwin Kernel Version 9.5.0: Wed
>    Sep  3 11:29:43 PDT 2008; root:xnu-1228.7.58~1/RELEASE_I386 i386

sounds like a 32bit - can we exclude it's 64bit?

>    2.swirsky-tppllc:~ swirsky$ crm -v
>     This is CRM114, version 20070810-BlameTheSegfault (0.6.8)

looks like you've an aged 0.6.8 TRE regex engine - can you upgrade to
0.7.5?

>    (It seems that the Mac came with one pre-installed, but I downloaded
>    and build what I presume to be the latest version)

latest tarball, not latest code, that's (w)get-able from crm114.sf.net/src/

>    I took the same "good" and "bad" sets that I trained the Paul
>    Graham/Baysean system with and trained CRM 114, like this:
>    {
>            learn <osb unique microgroom> (:*:_arg2:)
>    }
>    (with arguments like)
>    crm learn.crm BAD < bad.txt

does it even learn? try with clean BAD cssfile, learn a bad.txt, run 
~$ cssutil -r -b BAD
what do you get? same for GOOD.

>    Best match to file #1 (BAD) prob: 0.7139  pR: 0.3972
>    Total features in input file: 704
>    #0 (GOOD): features: 438256, hits: 1498, prob: 2.86e-01, pR:  -0.40
>    #1 (BAD): features: 446250, hits: 1687, prob: 7.14e-01, pR:   0.40
>    Note that the "Best match" line has a different pR. From TFM, it seems

well, only different printout and rounding.

>    #0 (GOOD): features: 438256, hits: 1498, prob: 2.86e-01, pR:  -0.40
>    #1 (BAD): features: 446250, hits: 1687, prob: 7.14e-01, pR:   0.40
>    What's the real story here?

dunno ... might be perfectly right, depends on training sets and text you
have feeded it here. We can only see it's a short one (704 features).
Actual examples would be useful (though likely not avail for a public
ML). But for testing any good/bad short texts collection would be fine.


-- 
paolo

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/