Jeff Fry | 1 Jul 2004 19:06
Picon
Picon

Problem using WEKA v3.4 with MySQL/MyODBC

Dear Weka List,

I'm really looking forward to using WEKA in my research, but I have a
problem importing instances into WEKA Explorer from MySQL.

I would be very grateful if you could suggest anything I might try; these
are the symptoms...

I have WEKA v 3.4, Sun J2RE v1.4.2 running on an NT 4, SP 6 PC 
client talking to an MySQL database (v3.23.58) on a Linux server, via 
ODBC (MyODBC v 3.51.8).

I configured a User DSN called VOICESforWEKA (MyODBC confirms 
it can connect to the data source), and specified the DSN in the 
DatabaseUtils.props file, which lives in the same directory as weka.jar
(the working directory I hope???) and contains two lines...

jdbcDriver=sun.jdbc.odbc.JdbcOdbcDriver
jdbcURL=jdbc:odbc:VOICESforWEKA

When I click on the Open DB... button in WEKA Explorer, a dialogue 
box labelled weka.experiment.InstanceQuery appears. I type 
jdbc:odbc:VOICESforWEKA as the databaseURL, and Select * from 
Pair as the query (username & password are blank, sparseData is 
FALSE) and press OK.

No Instances are loaded, instead, an error dialogue box appears (red 
circle, white x icon), labelled Load Instances, with the message 
'Couldn't read from database: null'. 

(Continue reading)

AllysonBritoEstadao | 2 Jul 2004 08:09
Picon
Favicon

RES: Wekalist Digest, Vol 17, Issue 2

Hi Guys,

I'm a master degree student, from UFPA/Brazil and I'm studying Text Mining
for my thesis.

I'd like to use htdig (http://www.htdig.org) to crawl the websites I'd like
to mine to search for clusters, classification rules, or things like that.
It's easy to convert the binary format of database of htdig for a arff file.
But I never worked with this kind of mine. 

Could you help or send me any material about text mine, using weka or any
other tool or technique?

Tks

Allyson
alesulis@libero.it | 2 Jul 2004 16:24
Picon
Favicon

Re:Problem using WEKA v3.4 with MySQL/MyODBC

dear jeff, 
i have for you a question: do you have data types in your database? if yes, try
, if it's possible, to change the type from data to text, and see if you solve
the problem.
I had the same problem on an access database, and it was due to this!!!!

Regards
Alessandro


---------- Initial Header -----------

>From      : wekalist-bounces <at> list.scms.waikato.ac.nz
To          : "WEKA mailing list" wekalist <at> list.scms.waikato.ac.nz
Cc          : 
Date      : Thu, 01 Jul 2004 18:06:22 +0100
Subject : [Wekalist] Problem using WEKA v3.4 with MySQL/MyODBC

> Dear Weka List,
> 
> I'm really looking forward to using WEKA in my research, but I have a
> problem importing instances into WEKA Explorer from MySQL.
> 
> I would be very grateful if you could suggest anything I might try; these
> are the symptoms...
> 
> I have WEKA v 3.4, Sun J2RE v1.4.2 running on an NT 4, SP 6 PC 
> client talking to an MySQL database (v3.23.58) on a Linux server, via 
> ODBC (MyODBC v 3.51.8).
> 
(Continue reading)

Dr. Maicel E. Monzon Perez | 2 Jul 2004 16:48
Picon

hello!!

Hello list, Can you see me?

It is a prove

M <at> icel

----- Original Message -----
From: <wekalist-request <at> list.scms.waikato.ac.nz>
To: <wekalist <at> list.scms.waikato.ac.nz>
Sent: Thursday, July 01, 2004 9:07 PM
Subject: Wekalist Digest, Vol 17, Issue 2

> 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. Problem using WEKA v3.4 with MySQL/MyODBC (Jeff Fry)
(Continue reading)

Jeff Fry | 2 Jul 2004 20:17
Picon
Picon

Re:Problem using WEKA v3.4 with MySQL/MyODBC

Thanks Alessandro,

as you suggested, I created a new test table in MySQL thus...

CREATE TABLE `WEKA` (
  `field1` text NOT NULL,
  `field2` text
) TYPE=MyISAM

and populated it with 2 rows of text...

field1		field2
hello		world
good		night

but I still get exactly the same problem I described previously.

Kind regards :>)

On 2 Jul 2004 at 16:24, alesulis <at> libero.it wrote:

> dear jeff, 
> i have for you a question: do you have data types in your database? if
> yes, try , if it's possible, to change the type from data to text, and
> see if you solve the problem. I had the same problem on an access
> database, and it was due to this!!!!
> 
> Regards
> Alessandro
> 
(Continue reading)

Toni | 3 Jul 2004 09:32

Problem using WEKA v3.4 with SQLServer/ORACLE

Dear weka list,

I also have the same problem with Jeff Fry. I tried to connect weka with
SQLServer / Oracle by JdbcOdbcDriver. I configured user DSN named sql to
connect with SQLServer Database server. I tested that connection and
success. I specified the DSN in the DatabaseUtils.props file.

jdbcDriver=sun.jdbc.odbc.JdbcOdbcDriver
jdbcURL=jdbc:odbc:sql

But when I runned weka, I had a problem like this:

Loaded driver: sun.jdbc.odbc.JdbcOdbcDriver
Connecting to jdbc:odbc:sql
GenericObjectEditor: Problem making backup object
java.io.NotSerializableException: sun.jdbc.odbc.JdbcOdbcConnection
STATUS: Reading from database...
Executing query: SELECT * from anggota
Connecting to jdbc:odbc:sql
Getting metadata...
Completed getting metadata...
STATUS: Problem executing DB query SELECT * FROM ?

and there are any message box with message "Couldn't read from database:
null". This error also happened when i try with Oracle.

Can anyone give a suggestion with this problem ?

Thanks.
(Continue reading)

Pino Rino | 5 Jul 2004 12:20
Picon
Favicon

Database connection problems

Hi

I solved the DB connection problem as described in the
attached mail

Cheers 
Pino Rino

Note: forwarded message attached.

		
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail
Picon Favicon
From: Pino Rino <pino_rino <at> yahoo.com>
Subject: Suggested update in DatabaseUtils.java
Date: 2004-06-21 13:41:38 GMT
Dear WEKAers

I realized that when I connect my DB2 database to the
Explorer, there are some exeptions because the Data
types are written in low-case and up-case (e.g.
(Continue reading)

J.A.I.L. | 5 Jul 2004 21:12

Frequently used datasets on machine learning

Reading the WEKA book (Data Mining Practical Machine Learning Tools ...), 
on page 81 I found "... the performance of the 1R procedure was reported on 
sixteen datasets frequently used by machine learning researchers...".
Are those 16 datasets really commonly used? If so, are they freely 
downloadable somewhere?

T.I.A.
Jose Antonio I. López
P A Atherton | 6 Jul 2004 11:56
Picon
Favicon

Standard scripts

Dear list,

I was wondering if anyone has a collection of standard java code for doing
common tasks in Weka - e.g. applying filters, running an algorithm, etc -
that they have made available to others?

Up till now, I have been running Weka via the GUI, but was thinking that
it would be much easier to have some standard scripts to tweak rather than
setting things up over and over again...

TIA

Paul
Tom Fawcett | 6 Jul 2004 12:41
Picon

Re: Frequently used datasets on machine learning

On Monday 05 July 2004 12:12 pm, J.A.I.L. wrote:
> Reading the WEKA book (Data Mining Practical Machine Learning Tools ...), 
> on page 81 I found "... the performance of the 1R procedure was reported on 
> sixteen datasets frequently used by machine learning researchers...".
> Are those 16 datasets really commonly used? If so, are they freely 
> downloadable somewhere?

See:

ftp://ftp.ics.uci.edu/pub/ml-repos/machine-learning-databases/

Regards,
-Tom

Gmane