[developers] Re: estimate and [incr tsdb()]

[Yi Zhang]

Hi Stephan,


> which version of estimate do you have installed (`see version.h').

I used the pre-compiled version of 'estimate' from
http://bulba.sdsu.edu/malouf/software/estimate.gz
Running 'estimate' without argument, I get the following message.

[yzhang at false delphin]$ estimate
--------------------------------------------------------------------------
Petsc Release Version 2.3.0, Patch 16, April, 26, 2005
See docs/changes/index.html for recent updates.
See docs/faq.html for hints about trouble shooting.
See docs/index.html for manual pages.
-----------------------------------------------------------------------
estimate on a linux-gnu named false by yzhang Wed Aug 31 10:04:34 2005
Libraries linked from /usr/local/src/petsc-2.3.0/lib/linux-gnu
Configure run at Mon Jun  6 12:51:24 2005
Configure options --with-default-optimization=O
--with-blas-lapack-dir=/usr/local/src/ATLAS/lib/Linux_PIIISSE1_2
--CXXFLAGS_O="-O3 -march=pentium3 -mfpmath=sse" -FFLAGS_O="-O3
-march=pentium3 -mfpmath=sse" --with-mpi=0 --with-dynamic=0
--with-clanguage=C++ --with-shared=0
-----------------------------------------------------------------------


The [incr tsdb()] I am using is the one packed in build 2005-08-02.
I am attaching a file that includes the output files of the training
session on ecoc data set (except the event input file, which is fairly
large and looks ok to me). I get the same output if the `estimate' is run
on terminal with the same event and variances files.


Thanks a lot,
Yi



On Wed, 31 Aug 2005, Stephan Oepen wrote:

> hi yi,
>
> > A closer look at Itsdb and MaxEnt shows that there's a problem with
> > the "-variances" argument. Since it's not documented in MaxEnt manual, I
> > have no clue what should I do with it. Is it safe if I just remove it?
>
> which version of estimate do you have installed (`see version.h').  the
> -variances option, for all i remember, specifies the prior, hence is a
> fairly important thing to get right.  we arrived at the current code by
> consultation with rob, although i would have to refresh my memory if i
> were to tell you more about the syntax :-).
>
> also, do you have a current [incr tsdb()] version installed?  what did
> you get in terms of status, warning, or error messages?  ideally, send
> me the full *common-lisp* buffer of a failed training session ...
>
>                                                            best  -  oe
>
> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> +++ Universitetet i Oslo (ILF); Boks 1102 Blindern; 0317 Oslo; (+47) 2285 7989
> +++     CSLI Stanford; Ventura Hall; Stanford, CA 94305; (+1 650) 723 0515
> +++       --- oe at csli.stanford.edu; oe at hf.uio.no; stephan at oepen.net ---
> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: maxent-estimate.tar.gz
Type: application/octet-stream
Size: 14887 bytes
Desc: 
URL: <http://lists.delph-in.net/archives/developers/attachments/20050831/12454632/attachment.obj>