using PredictionAppender for many files

Hi all,

When testing a model, I wanna save a file containing the predicted class for each instance. This can be done
using PredictionAppender (under Evaluation Tab in KnowledgeFlow) which is then attached to a DataSink
to save the file. But I wanna repeat this for a lot of files, So I'd like to find something like this in either
Experimenter or SimpleCLI.

Thank you,
Bahar

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

assigning new cases to previously developed clusters

Does anyone know how to do these? I am using Explorer in Weka 3.7.

I have a data set which I have divided up into a Training set and a Testing set.
I clustered the training set.

Now is there a way to assign each record from the Test set to the
cluster in the Training set to which it is closest?

Thank you.

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

RE: How to modify source code of Weka

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

csh script in CLI

Hi all,

There's a csh script example in Weka tutorial. From the tutorial I understand that the lines should be
executed in the SimpleCLI, but CLI doesn't accept them. (It says "Command must be one of: ..."). I'd be
thankful if you let me know the correct way of executing this example. 

Thank you,
Bahar

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Unit tests

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Re: buying a PC for data mining projects

I highly recommend a set up that will take a 64 bit OS.  Memory
space for java on a 4 G RAM Machine is only about 1.4 GB.  That
can be very frustrating.

--- On Sat, 1/30/10, wekalist-request <at> list.scms.waikato.ac.nz
<wekalist-request <at> list.scms.waikato.ac.nz> wrote:

> From: wekalist-request <at> list.scms.waikato.ac.nz <wekalist-request <at> list.scms.waikato.ac.nz>
> Subject: Wekalist Digest, Vol 83, Issue 80
> To: wekalist <at> list.scms.waikato.ac.nz
> Date: Saturday, January 30, 2010, 5:08 PM
> Send Wekalist mailing list
> submissions to
>     wekalist <at> list.scms.waikato.ac.nz
> 
> To subscribe or unsubscribe via the World Wide Web, visit
>     https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> or, via email, send a message with subject or body 'help'
> to
>     wekalist-request <at> list.scms.waikato.ac.nz
> 
> You can reach the person managing the list at
>     wekalist-owner <at> list.scms.waikato.ac.nz
> 
> When replying, please edit your Subject line so it is more
> specific
> than "Re: Contents of Wekalist digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: OSX + weka to remote MySQL Server
> (Nick Torenvliet)
>    2. Combining multiple classifiers using
> WEKA (Mohamad El Abed)
>    3. buying a PC for data mining projects
> (wessel van persie)
>    4. Re: buying a PC for data mining
> projects (Mary.Felkin <at> lri.fr)
>    5. Re: buying a PC for data mining
> projects (Winton Davies)
>    6. Re: buying a PC for data mining
> projects (Winton Davies)
>    7. RE: Combining multiple classifiers
> using WEKA
>       (=?gb2312?B?19658M+y?=)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Sat, 30 Jan 2010 10:33:39 -0500
> From: Nick Torenvliet <nick.torenvliet <at> gmail.com>
> Subject: Re: [Wekalist] OSX + weka to remote MySQL Server
> To: "Weka machine learning workbench list."
>     <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID:
>     <8531d53e1001300733j2088bb50m42707621775f83f8 <at> mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hey Mark, thanks again!  I'm really new to the Mac, so
> I looked around for
> how to set CLASSPATH and couldn't find a thing so I went
> with the ODBC
> Driver.
> 
> That's nifty, so each app has its own .profile it you
> will.  That's pretty
> brilliant really.
> 
> 
> 
> On Sat, Jan 30, 2010 at 2:46 AM, Mark Hall <mhall <at> pentaho.com>
> wrote:
> 
> > Nick Torenvliet wrote:
> >
> >> Hey all,
> >>
> >> After getting the weka / MySQL thing working on
> the localhost I thought
> >> I'd get my Mac Book up and running with it.
> >>
> >> I've got the ODBC connection working from the mac
> to my linux box in both
> >> the Mac ODBC admin utility and using MySQL
> Administrator.
> >>
> >> I've copied the odbc DatabaseUtils.props file into
> my home dir and weka
> >> picks it up.  Its got my working User DSN in
> it.
> >>
> >> When I go to connect I get Exception
> java.sql.SQLException: No suitable
> >> driver found for jdbc:odbc:markets ...
> >>
> >
> > Is there any reason why you are not using the MySQL
> JDBC driver on your
> > Mac?
> >
> >
> >
> >> weka.log complains as follows:
> >> ---Registering Weka Editors---
> >> Trying to add database driver (JDBC):
> org.gjt.mm.mysql.Driver - Error, not
> >> in CL
> >> ASSPATH?
> >>
> >>
> >> And DatabaseUtils.props contatins
> >>
> >>
> >> # JDBC driver (comma-separated list)
> >> jdbcDriver=sun.jdbc.odbc.JdbcOdbcDriver
> >>
> >> # database URL
> >> jdbcURL=jdbc:odbc:markets
> >>
> >> Which weka picks up.
> >>
> >> This markets is a working User DSN.
> >>
> >> What do I need to complete this picture?
> >>
> >> I launch weka from the icon that came with the
> install file not from the
> >> command line.
> >>
> >
> > If you launch by double clicking on the icon on the
> Mac then you'll need to
> > edit the Mac application's Info.plist file in order to
> add extra stuff to
> > the CLASSPATH for the application. Open a terminal and
> cd into the
> > weka.app/Contents directory and then load the
> Info.plist into a text eitor.
> > Look for the "ClassPath" tag and edit its value
> appropriately.
> >
> > Alternatively, just launch Weka from the command line
> with appropriate
> > CLASSPATH.
> >
> > Cheers,
> > Mark.
> >
> >
> > _______________________________________________
> > Wekalist mailing list
> > Send posts to: Wekalist <at> list.scms.waikato.ac.nz
> > List info and subscription status:
> > https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> > List etiquette:
> > http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html<http://www.cs.waikato.ac.nz/%7Eml/weka/mailinglist_etiquette.html>
> >
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/20100130/96206fdf/attachment-0001.html
> 
> ------------------------------
> 
> Message: 2
> Date: Sat, 30 Jan 2010 16:34:24 +0100
> From: Mohamad El Abed <melabed <at> greyc.ensicaen.fr>
> Subject: [Wekalist] Combining multiple classifiers using
> WEKA
> To: "Weka machine learning workbench list."
>     <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID: <4B645180.9030903 <at> greyc.ensicaen.fr>
> Content-Type: text/plain; charset=ISO-8859-1;
> format=flowed
> 
> Dear all,
> 
> I have a question on combining multiple classifiers on a
> data set. For 
> example, i have build decision trees using two classifiers
> (C4.5 & 
> SimpleCart) by this java code:
> 
> J48 arb=new J48();
> arb.buildClassifier(dataset);
> System.out.println(arb);
> 
> SimpleCart arb=new SimpleCart();
> arb.buildClassifier(dataset);
> System.out.println(arb);
> 
> 
> My question is: What is the jave code to build the decision
> tree using 
> the combination of both algorithms?
> 
> 
> Thx in advance
> 
> 
> 
> ------------------------------
> 
> Message: 3
> Date: Sat, 30 Jan 2010 19:21:59 +0100
> From: wessel van persie <wesselvanpersie <at> googlemail.com>
> Subject: [Wekalist] buying a PC for data mining projects
> To: "Weka machine learning workbench list."
>     <Wekalist <at> list.scms.waikato.ac.nz>
> Message-ID:
>     <86d53a031001301021y3f971918udc77254da11e7be7 <at> mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
> 
> Dear Weka Users,
> 
> What would be a good pc for running WEKA experiments, with
> lets say a
> 1000 euro budget?
> A fast hard disk is needed for loading big datasets? -->
> Solid sate disk?
> 
> When its running for long hours energy efficient
> motherboard and CPU
> are needed?
> -- > Chipset / integrated video card with CPU?
> AMD: Gigabyte GA-MA78GM with AMD780G Chipset with AMD
> Phenom X4 9100e
> INTEL: ECS G33T with Intel G33 Chipset with Intel Core 2
> Duo E8400
> 
> What is the biggest bottleneck in data analysis?
> Memory? --> DDR3 / 1600MHz?
> 
> Operating system? Linux or windows 7?
> 
> Best regards,
> 
> Wessel
> 
> 
> 
> ------------------------------
> 
> Message: 4
> Date: Sat, 30 Jan 2010 19:54:16 +0100
> From: Mary.Felkin <at> lri.fr
> Subject: Re: [Wekalist] buying a PC for data mining
> projects
> To: "Weka machine learning workbench list."
>     <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID: <a223b33d0b37e1779805abd7689a9f5a.squirrel <at> www2.lri.fr>
> Content-Type: text/plain;charset=iso-8859-1
> 
> I would say as much RAM as possible and definitely Linux.
> 
> Mary
> 
> 
> 
> > Dear Weka Users,
> >
> > What would be a good pc for running WEKA experiments,
> with lets say a
> > 1000 euro budget?
> > A fast hard disk is needed for loading big datasets?
> --> Solid sate disk?
> >
> > When its running for long hours energy efficient
> motherboard and CPU
> > are needed?
> > -- > Chipset / integrated video card with CPU?
> > AMD: Gigabyte GA-MA78GM with AMD780G Chipset with AMD
> Phenom X4 9100e
> > INTEL: ECS G33T with Intel G33 Chipset with Intel Core
> 2 Duo E8400
> >
> > What is the biggest bottleneck in data analysis?
> > Memory? --> DDR3 / 1600MHz?
> >
> > Operating system? Linux or windows 7?
> >
> > Best regards,
> >
> > Wessel
> >
> > _______________________________________________
> > Wekalist mailing list
> > Send posts to: Wekalist <at> list.scms.waikato.ac.nz
> > List info and subscription status:
> > https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> > List etiquette:
> > http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
> >
> 
> 
> 
> 
> 
> ------------------------------
> 
> Message: 5
> Date: Sat, 30 Jan 2010 11:14:47 -0800
> From: Winton Davies <wdavies <at> cs.stanford.edu>
> Subject: Re: [Wekalist] buying a PC for data mining
> projects
> To: "Weka machine learning workbench list."
>     <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID:
>     <87bf164a1001301114r6605a2bbo10f05c5ce1e0c009 <at> mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Consider using Amazon's ec2 cloud cluster. You can rent
> pretty huge clusters
> for around a dollar an hour. If what you are doing is
> easily map-reducable,
> then you can even divide up the tasks among multiple nodes.
> This clearly is
> not out of the box behaviour, but simple parallelization
> (for example,
> different parameter settings), is very easy to arrange.
> 
> W
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/20100130/89d3ac20/attachment-0001.html
> 
> ------------------------------
> 
> Message: 6
> Date: Sat, 30 Jan 2010 11:15:36 -0800
> From: Winton Davies <wdavies <at> cs.stanford.edu>
> Subject: Re: [Wekalist] buying a PC for data mining
> projects
> To: "Weka machine learning workbench list."
>     <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID:
>     <87bf164a1001301115x5f6fd815o3d85fbcfc486262a <at> mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> btw, Not sure if you can use X-windows from EC2, but Weka
> runs in a graphics
> free mode anyway.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/20100130/cacff04f/attachment-0001.html
> 
> ------------------------------
> 
> Message: 7
> Date: Sun, 31 Jan 2010 07:07:02 +0800
> From: =?gb2312?B?19658M+y?= <bison0927 <at> hotmail.com>
> Subject: RE: [Wekalist] Combining multiple classifiers
> using WEKA
> To: wekaList weka <wekalist <at> list.scms.waikato.ac.nz>
> Message-ID: <COL114-W11500AD298A01F26A6685ABA5A0 <at> phx.gbl>
> Content-Type: text/plain; charset="gb2312"
> 
> 
> Hi, I am not sure the meaning of combining here, I guess
> you may try do some voting here. if you can get the the
> classification result and its corresponding probability of
> each sample, you may simply choose the one with higher
> probability. 
> 
> so you loop through every sample, and then read the
> decision for each sample and compare the probability, choose
> the decision with higher probability.
> 
> > Date: Sat, 30 Jan 2010 16:34:24 +0100
> > From: melabed <at> greyc.ensicaen.fr
> > To: wekalist <at> list.scms.waikato.ac.nz
> > Subject: [Wekalist] Combining multiple classifiers
> using WEKA
> > 
> > Dear all,
> > 
> > I have a question on combining multiple classifiers on
> a data set. For 
> > example, i have build decision trees using two
> classifiers (C4.5 & 
> > SimpleCart) by this java code:
> > 
> > J48 arb=new J48();
> > arb.buildClassifier(dataset);
> > System.out.println(arb);
> > 
> > SimpleCart arb=new SimpleCart();
> > arb.buildClassifier(dataset);
> > System.out.println(arb);
> > 
> > 
> > My question is: What is the jave code to build the
> decision tree using 
> > the combination of both algorithms?
> > 
> > 
> > Thx in advance
> > 
> > _______________________________________________
> > Wekalist mailing list
> > Send posts to: Wekalist <at> list.scms.waikato.ac.nz
> > List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> > List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
>     
>         
>           
>   
> _________________________________________________________________
> Ô¼»á˵²»ÇåµØ·½£¿À´ÊÔÊÔ΢ÈíµØͼ×îÐÂmsn»¥¶¯¹¦ÄÜ£¡
> http://ditu.live.com/?form=TL&swm=1
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/20100131/7fb40227/attachment.html
> 
> ------------------------------
> 
> _______________________________________________
> Wekalist mailing list
> Wekalist <at> list.scms.waikato.ac.nz
> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> 
> 
> End of Wekalist Digest, Vol 83, Issue 80
> ****************************************
> 

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

RE: How to modify source code of Weka

Hi, 

Thank you very much for your help, I have followed the steps in the tutoriel
of the link page
"http://weka.wikispaces.com/Eclipse+3.4.x+%28Subversion%29". I succeed in
compiling the code source of Weka although  a number of errors is shown.
But, when I click on Run as a java application and I choose the
"GuiChooser", I have access to Weka screenshot as usual. 

Thank you, 

Best regards,

jose maria-4 wrote:
> 
> 
> I think that you extrated the source files inside eclipse workspace
> folder. Try moving the source code folder to another location outside the
> workspace.
> 
> Besides, have you tryed to make a project using subclipse and weka
> repository? I had no problems and all versions are there.
> 
> Regards Jose. 
>  
>> Date: Sat, 30 Jan 2010 15:21:35 -0800
>> From: dahmaniines <at> yahoo.fr
>> To: wekalist <at> list.scms.waikato.ac.nz
>> Subject: RE: [Wekalist] How to modify source code of Weka
>> 
>> 
>> Hi, 
>> 
>> Thank you for your help. I saw the two links. I applied the steps
>> described
>> in the page of the link " http://weka.wikispaces.com/Eclipse 3.4.x
>> (weka-src.jar) http://weka.wikispaces.com/Eclipse 3.4.x (weka-src.jar) ”:
>> Extracting the content of weka-src.jar and configuring Eclipse. 
>> I extract in a first step all the files from the weka-src.jar which are
>> then
>> stocked in a new directory called “ Weka”. 
>> 
>> Then, I open Eclipse and I create a new Project : a new java project
>> called
>> “Weka”. When I try to create a project from an existing source and I
>> click
>> on the “Borrow” Bottom in order to reach the folder “Weka” (in which we
>> save
>> files unzipped from weka-src.jar), the platform Eclipse displays
>> “Projects
>> located in the workspace folder must be direct sub folders of the
>> workspace
>> folder” as an error message. 
>> 
>> Even if I don’t choose this option, the problem persists. I complete the
>> remaining steps of configuration ("Exclude the src/test/java directory
>> from
>> the build path and make sure that the default output folder is the
>> following
>> directory Weka/build/classes") as they are mentioned in the link page.
>> But,
>> I can’t create the hierarchy in the “Project Explorer” on Eclipse knowing
>> that in the folder “Weka” we have the same hierarchy as it exists on
>> Weka-src.jar.
>> 
>> My target is to use some classifiers in order to modify them and to
>> extend
>> them. Should I export all Weka code or only the source code of these
>> classifiers? However, My problem still the same: How to configure Weka on
>> Eclipse 3.4.1 and to run after this project? 
>> As additional information, I use Windows XP and Eclipse 3.4.1.
>> Please, help me if you have any idea.
>> Thanks in advance.
>> Cordially,
>> 
>> 
>> Jose-72 wrote:
>> > 
>> > 
>> > Hi,
>> > 
>> > 
>> > 
>> > Have you seen the faq? Specifically those:
>> > 
>> > 
>> > 
>> > http://weka.wikispaces.com/Where+can+I+get+WEKA%27s+source+code%3F
>> > 
>> > http://weka.wikispaces.com/How+do+I+compile+WEKA%3F
>> > 
>> > 
>> > 
>> > The only "uncommon" thing you will have to do is to configure two
>> folders
>> > for source code. One for weka source code itself and other for unit
>> tests.
>> > It's explained in the wiki.
>> > 
>> > 
>> > 
>> > Regards, Jose.
>> > 
>> >> Date: Fri, 29 Jan 2010 03:28:52 -0800
>> >> From: dahmaniines <at> yahoo.fr
>> >> To: wekalist <at> list.scms.waikato.ac.nz
>> >> Subject: [Wekalist] How to modify source code of Weka
>> >> 
>> >> 
>> >> Hi, 
>> >> 
>> >> I used weka recently. I want to modify the code source of Weka. That's
>> >> why,
>> >> I extract all the files existing in the file .jar " weka-src.jar" and
>> I
>> >> create a new project on Eclipse which contains these files. But, this
>> >> project lead to a big number of errors.
>> >> 
>> >> I don't understand how to compile this project before adding or
>> modifying
>> >> the classes of weka-src.jar (I just copy the content of the file
>> >> weka-src.jar in the new project). 
>> >> Also, I don't know which main class I must use to compile the project. 
>> >> 
>> >> I need your help because I must in a firt step modify the weka source
>> >> code.
>> >> 
>> >> Thanks in advance.
>> >> 
>> >> Best regards,
>> >> 
>> >> 
>> >> -- 
>> >> View this message in context:
>> >>
>> http://old.nabble.com/How-to-modify-source-code-of-Weka-tp27364203p27364203.html
>> >> Sent from the WEKA mailing list archive at Nabble.com.
>> >> 
>> >> 
>> > 
>> > _________________________________________________________________
>> > ¡Seducción! 249 historias cada semana en el sitio nº1 para conseguir
>> una
>> > cita. ¡Regístrate!
>> > http://contactos.es.msn.com/?mtcmk=015352
>> > _______________________________________________
>> > Wekalist mailing list
>> > Send posts to: Wekalist <at> list.scms.waikato.ac.nz
>> > List info and subscription status:
>> > https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>> > List etiquette:
>> > http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
>> > 
>> > 
>> 
>> -- 
>> View this message in context:
>> http://old.nabble.com/How-to-modify-source-code-of-Weka-tp27364203p27388524.html
>> Sent from the WEKA mailing list archive at Nabble.com.
>> 
>> 
>  		 	   		  
> _________________________________________________________________
> ¿Aún no sabes qué móvil eres? ¡Descúbrelo aquí!
> http://www.quemovileres.com/
> _______________________________________________
> Wekalist mailing list
> Send posts to: Wekalist <at> list.scms.waikato.ac.nz
> List info and subscription status:
> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> List etiquette:
> http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
> 
> 

--

-- 
View this message in context: http://old.nabble.com/How-to-modify-source-code-of-Weka-tp27364203p27410546.html
Sent from the WEKA mailing list archive at Nabble.com.

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Re: buying a PC for data mining projects

Has no-one considered hooking up Weka to Terracotta + Ehcache.  
Terracotta to distribute the processing load and Ehcache to handle huge  
datasets.

Many Thanks
Dr Nicholas Walton

My folks were English ... We were too poor to be Britsih. There were so  
many of us in my family, I was eight years old before it was my turn in  
the bathroom. Four of us slept in the same bed. When we got cold,  
mother threw on another brother.

Bob Hope, My Life in Jokes (2003)
On 1 Feb 2010, at 19:40, Emory Creel wrote:

> I highly recommend a set up that will take a 64 bit OS.  Memory
> space for java on a 4 G RAM Machine is only about 1.4 GB.  That
> can be very frustrating.
>
> --- On Sat, 1/30/10, wekalist-request <at> list.scms.waikato.ac.nz  
> <wekalist-request <at> list.scms.waikato.ac.nz> wrote:
>
>> From: wekalist-request <at> list.scms.waikato.ac.nz  
>> <wekalist-request <at> list.scms.waikato.ac.nz>
>> Subject: Wekalist Digest, Vol 83, Issue 80
>> To: wekalist <at> list.scms.waikato.ac.nz
>> Date: Saturday, January 30, 2010, 5:08 PM
>> Send Wekalist mailing list
>> submissions to
>>     wekalist <at> list.scms.waikato.ac.nz
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>>     https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>> or, via email, send a message with subject or body 'help'
>> to
>>     wekalist-request <at> list.scms.waikato.ac.nz
>>
>> You can reach the person managing the list at
>>     wekalist-owner <at> list.scms.waikato.ac.nz
>>
>> When replying, please edit your Subject line so it is more
>> specific
>> than "Re: Contents of Wekalist digest..."
>>
>>
>> Today's Topics:
>>
>>    1. Re: OSX + weka to remote MySQL Server
>> (Nick Torenvliet)
>>    2. Combining multiple classifiers using
>> WEKA (Mohamad El Abed)
>>    3. buying a PC for data mining projects
>> (wessel van persie)
>>    4. Re: buying a PC for data mining
>> projects (Mary.Felkin <at> lri.fr)
>>    5. Re: buying a PC for data mining
>> projects (Winton Davies)
>>    6. Re: buying a PC for data mining
>> projects (Winton Davies)
>>    7. RE: Combining multiple classifiers
>> using WEKA
>>       (=?gb2312?B?19658M+y?=)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Sat, 30 Jan 2010 10:33:39 -0500
>> From: Nick Torenvliet <nick.torenvliet <at> gmail.com>
>> Subject: Re: [Wekalist] OSX + weka to remote MySQL Server
>> To: "Weka machine learning workbench list."
>>     <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID:
>>     <8531d53e1001300733j2088bb50m42707621775f83f8 <at> mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hey Mark, thanks again!  I'm really new to the Mac, so
>> I looked around for
>> how to set CLASSPATH and couldn't find a thing so I went
>> with the ODBC
>> Driver.
>>
>> That's nifty, so each app has its own .profile it you
>> will.  That's pretty
>> brilliant really.
>>
>>
>>
>> On Sat, Jan 30, 2010 at 2:46 AM, Mark Hall <mhall <at> pentaho.com>
>> wrote:
>>
>>> Nick Torenvliet wrote:
>>>
>>>> Hey all,
>>>>
>>>> After getting the weka / MySQL thing working on
>> the localhost I thought
>>>> I'd get my Mac Book up and running with it.
>>>>
>>>> I've got the ODBC connection working from the mac
>> to my linux box in both
>>>> the Mac ODBC admin utility and using MySQL
>> Administrator.
>>>>
>>>> I've copied the odbc DatabaseUtils.props file into
>> my home dir and weka
>>>> picks it up.  Its got my working User DSN in
>> it.
>>>>
>>>> When I go to connect I get Exception
>> java.sql.SQLException: No suitable
>>>> driver found for jdbc:odbc:markets ...
>>>>
>>>
>>> Is there any reason why you are not using the MySQL
>> JDBC driver on your
>>> Mac?
>>>
>>>
>>>
>>>> weka.log complains as follows:
>>>> ---Registering Weka Editors---
>>>> Trying to add database driver (JDBC):
>> org.gjt.mm.mysql.Driver - Error, not
>>>> in CL
>>>> ASSPATH?
>>>>
>>>>
>>>> And DatabaseUtils.props contatins
>>>>
>>>>
>>>> # JDBC driver (comma-separated list)
>>>> jdbcDriver=sun.jdbc.odbc.JdbcOdbcDriver
>>>>
>>>> # database URL
>>>> jdbcURL=jdbc:odbc:markets
>>>>
>>>> Which weka picks up.
>>>>
>>>> This markets is a working User DSN.
>>>>
>>>> What do I need to complete this picture?
>>>>
>>>> I launch weka from the icon that came with the
>> install file not from the
>>>> command line.
>>>>
>>>
>>> If you launch by double clicking on the icon on the
>> Mac then you'll need to
>>> edit the Mac application's Info.plist file in order to
>> add extra stuff to
>>> the CLASSPATH for the application. Open a terminal and
>> cd into the
>>> weka.app/Contents directory and then load the
>> Info.plist into a text eitor.
>>> Look for the "ClassPath" tag and edit its value
>> appropriately.
>>>
>>> Alternatively, just launch Weka from the command line
>> with appropriate
>>> CLASSPATH.
>>>
>>> Cheers,
>>> Mark.
>>>
>>>
>>> _______________________________________________
>>> Wekalist mailing list
>>> Send posts to: Wekalist <at> list.scms.waikato.ac.nz
>>> List info and subscription status:
>>> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>>> List etiquette:
>>> http://www.cs.waikato.ac.nz/~ml/weka/ 
>>> mailinglist_etiquette.html<http://www.cs.waikato.ac.nz/%7Eml/weka/ 
>>> mailinglist_etiquette.html>
>>>
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:  
>> https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/ 
>> 20100130/96206fdf/attachment-0001.html
>>
>> ------------------------------
>>
>> Message: 2
>> Date: Sat, 30 Jan 2010 16:34:24 +0100
>> From: Mohamad El Abed <melabed <at> greyc.ensicaen.fr>
>> Subject: [Wekalist] Combining multiple classifiers using
>> WEKA
>> To: "Weka machine learning workbench list."
>>     <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID: <4B645180.9030903 <at> greyc.ensicaen.fr>
>> Content-Type: text/plain; charset=ISO-8859-1;
>> format=flowed
>>
>> Dear all,
>>
>> I have a question on combining multiple classifiers on a
>> data set. For
>> example, i have build decision trees using two classifiers
>> (C4.5 &
>> SimpleCart) by this java code:
>>
>> J48 arb=new J48();
>> arb.buildClassifier(dataset);
>> System.out.println(arb);
>>
>> SimpleCart arb=new SimpleCart();
>> arb.buildClassifier(dataset);
>> System.out.println(arb);
>>
>>
>> My question is: What is the jave code to build the decision
>> tree using
>> the combination of both algorithms?
>>
>>
>> Thx in advance
>>
>>
>>
>> ------------------------------
>>
>> Message: 3
>> Date: Sat, 30 Jan 2010 19:21:59 +0100
>> From: wessel van persie <wesselvanpersie <at> googlemail.com>
>> Subject: [Wekalist] buying a PC for data mining projects
>> To: "Weka machine learning workbench list."
>>     <Wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID:
>>     <86d53a031001301021y3f971918udc77254da11e7be7 <at> mail.gmail.com>
>> Content-Type: text/plain; charset=ISO-8859-1
>>
>> Dear Weka Users,
>>
>> What would be a good pc for running WEKA experiments, with
>> lets say a
>> 1000 euro budget?
>> A fast hard disk is needed for loading big datasets? -->
>> Solid sate disk?
>>
>> When its running for long hours energy efficient
>> motherboard and CPU
>> are needed?
>> -- > Chipset / integrated video card with CPU?
>> AMD: Gigabyte GA-MA78GM with AMD780G Chipset with AMD
>> Phenom X4 9100e
>> INTEL: ECS G33T with Intel G33 Chipset with Intel Core 2
>> Duo E8400
>>
>> What is the biggest bottleneck in data analysis?
>> Memory? --> DDR3 / 1600MHz?
>>
>> Operating system? Linux or windows 7?
>>
>> Best regards,
>>
>> Wessel
>>
>>
>>
>> ------------------------------
>>
>> Message: 4
>> Date: Sat, 30 Jan 2010 19:54:16 +0100
>> From: Mary.Felkin <at> lri.fr
>> Subject: Re: [Wekalist] buying a PC for data mining
>> projects
>> To: "Weka machine learning workbench list."
>>     <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID: <a223b33d0b37e1779805abd7689a9f5a.squirrel <at> www2.lri.fr>
>> Content-Type: text/plain;charset=iso-8859-1
>>
>> I would say as much RAM as possible and definitely Linux.
>>
>> Mary
>>
>>
>>
>>> Dear Weka Users,
>>>
>>> What would be a good pc for running WEKA experiments,
>> with lets say a
>>> 1000 euro budget?
>>> A fast hard disk is needed for loading big datasets?
>> --> Solid sate disk?
>>>
>>> When its running for long hours energy efficient
>> motherboard and CPU
>>> are needed?
>>> -- > Chipset / integrated video card with CPU?
>>> AMD: Gigabyte GA-MA78GM with AMD780G Chipset with AMD
>> Phenom X4 9100e
>>> INTEL: ECS G33T with Intel G33 Chipset with Intel Core
>> 2 Duo E8400
>>>
>>> What is the biggest bottleneck in data analysis?
>>> Memory? --> DDR3 / 1600MHz?
>>>
>>> Operating system? Linux or windows 7?
>>>
>>> Best regards,
>>>
>>> Wessel
>>>
>>> _______________________________________________
>>> Wekalist mailing list
>>> Send posts to: Wekalist <at> list.scms.waikato.ac.nz
>>> List info and subscription status:
>>> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>>> List etiquette:
>>> http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
>>>
>>
>>
>>
>>
>>
>> ------------------------------
>>
>> Message: 5
>> Date: Sat, 30 Jan 2010 11:14:47 -0800
>> From: Winton Davies <wdavies <at> cs.stanford.edu>
>> Subject: Re: [Wekalist] buying a PC for data mining
>> projects
>> To: "Weka machine learning workbench list."
>>     <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID:
>>     <87bf164a1001301114r6605a2bbo10f05c5ce1e0c009 <at> mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Consider using Amazon's ec2 cloud cluster. You can rent
>> pretty huge clusters
>> for around a dollar an hour. If what you are doing is
>> easily map-reducable,
>> then you can even divide up the tasks among multiple nodes.
>> This clearly is
>> not out of the box behaviour, but simple parallelization
>> (for example,
>> different parameter settings), is very easy to arrange.
>>
>> W
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:  
>> https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/ 
>> 20100130/89d3ac20/attachment-0001.html
>>
>> ------------------------------
>>
>> Message: 6
>> Date: Sat, 30 Jan 2010 11:15:36 -0800
>> From: Winton Davies <wdavies <at> cs.stanford.edu>
>> Subject: Re: [Wekalist] buying a PC for data mining
>> projects
>> To: "Weka machine learning workbench list."
>>     <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID:
>>     <87bf164a1001301115x5f6fd815o3d85fbcfc486262a <at> mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> btw, Not sure if you can use X-windows from EC2, but Weka
>> runs in a graphics
>> free mode anyway.
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:  
>> https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/ 
>> 20100130/cacff04f/attachment-0001.html
>>
>> ------------------------------
>>
>> Message: 7
>> Date: Sun, 31 Jan 2010 07:07:02 +0800
>> From: =?gb2312?B?19658M+y?= <bison0927 <at> hotmail.com>
>> Subject: RE: [Wekalist] Combining multiple classifiers
>> using WEKA
>> To: wekaList weka <wekalist <at> list.scms.waikato.ac.nz>
>> Message-ID: <COL114-W11500AD298A01F26A6685ABA5A0 <at> phx.gbl>
>> Content-Type: text/plain; charset="gb2312"
>>
>>
>> Hi, I am not sure the meaning of combining here, I guess
>> you may try do some voting here. if you can get the the
>> classification result and its corresponding probability of
>> each sample, you may simply choose the one with higher
>> probability.
>>
>> so you loop through every sample, and then read the
>> decision for each sample and compare the probability, choose
>> the decision with higher probability.
>>
>>> Date: Sat, 30 Jan 2010 16:34:24 +0100
>>> From: melabed <at> greyc.ensicaen.fr
>>> To: wekalist <at> list.scms.waikato.ac.nz
>>> Subject: [Wekalist] Combining multiple classifiers
>> using WEKA
>>>
>>> Dear all,
>>>
>>> I have a question on combining multiple classifiers on
>> a data set. For
>>> example, i have build decision trees using two
>> classifiers (C4.5 &
>>> SimpleCart) by this java code:
>>>
>>> J48 arb=new J48();
>>> arb.buildClassifier(dataset);
>>> System.out.println(arb);
>>>
>>> SimpleCart arb=new SimpleCart();
>>> arb.buildClassifier(dataset);
>>> System.out.println(arb);
>>>
>>>
>>> My question is: What is the jave code to build the
>> decision tree using
>>> the combination of both algorithms?
>>>
>>>
>>> Thx in advance
>>>
>>> _______________________________________________
>>> Wekalist mailing list
>>> Send posts to: Wekalist <at> list.scms.waikato.ac.nz
>>> List info and subscription status:  
>>> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>>> List etiquette:  
>>> http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
>>     
>>         
>>           
>>  
>> _________________________________________________________________
>> Ô¼»á˵²»ÇåµØ·½£¿À´ÊÔÊÔ΢ÈíµØͼ×îÐÂmsn»¥¶¯¹¦ÄÜ£¡
>> http://ditu.live.com/?form=TL&swm=1
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:  
>> https://list.scms.waikato.ac.nz/pipermail/wekalist/attachments/ 
>> 20100131/7fb40227/attachment.html
>>
>> ------------------------------
>>
>> _______________________________________________
>> Wekalist mailing list
>> Wekalist <at> list.scms.waikato.ac.nz
>> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
>>
>>
>> End of Wekalist Digest, Vol 83, Issue 80
>> ****************************************
>>
>
> _______________________________________________
> Wekalist mailing list
> Send posts to: Wekalist <at> list.scms.waikato.ac.nz
> List info and subscription status:  
> https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
> List etiquette:  
> http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

RE: How to modify source code of Weka

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

Inputting text file in Weka

_______________________________________________
Wekalist mailing list
Send posts to: Wekalist <at> list.scms.waikato.ac.nz
List info and subscription status: https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html