weka-users

[Weka-users] New releases: WEKA 3.8.7 and WEKA 3.9.7

[Eibe F. <eib...@wa...>]

Hi everyone,

New releases of WEKA 3.8 and 3.9 are now available. The most noteworthy additions to the collection of downloadables at

  https://waikato.github.io/weka-wiki/downloading_weka/

are platform‑specific distributions for ARM Windows and ARM Linux. As usual, we also provide an ARM version for macOS, along with Intel versions for all three platforms. In addition, there is a platform‑independent ZIP file that can be used with any Java 8 or later JVM on your system (although some WEKA packages beyond core WEKA may require recent Java versions).

All platform‑specific distributions are bundled with a corresponding Java 25 JRE provided by BellSoft. We have switched to BellSoft because they make available JREs for all six supported platforms that include JavaFX, which is required by some of the visualisation tools in WEKA packages.

The primary purpose of this release was to make WEKA work cleanly with Java 25 and to reduce the number of --add-opens arguments required when running WEKA from an operating system’s command‑line interface. We have also fixed a few other command‑line handling bugs and addressed some additional issues. Please see the changelog for a complete list of changes.

As with all recent versions of WEKA, the 3.8 (“stable”) and 3.9 (“development”) releases have identical functionality. We continue to maintain the 3.9 release series in anticipation of potential major future changes, which would be introduced there only.

WEKA has not been modularised and continues to run on the classpath rather than the module path. When running WEKA from an operating system’s command‑line interface without using the provided weka.sh (macOS, Linux) or RunWeka.bat (Windows) scripts included in the platform‑specific distributions, one --add-opens option is still required when using Java 25. For example, to run J48 from the command line:

  java -cp weka.jar --add-opens=java.base/java.lang=ALL-UNNAMED weka.Run J48 -t ...

Depending on whether code involving JavaFX or native library access is executed, the following additional flags may be used to suppress warnings:

  --enable-native-access=ALL-UNNAMED --enable-native-access=javafx.graphics

All WEKA packages should work on all supported platforms, including wekaPython and RPlugin. We have also made a netlibNative package for Windows ARM, and the netlibNative package for Linux now includes native code for both ARM and Intel. This enables fast native matrix algebra for schemes such as LinearRegression, GaussianProcesses, and related learners.

Cheers,
Eibe

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Khaled EL-G. <kha...@gm...>]

Thanks for your reply.
I will try to modify the mentioned code.

Cheers,
Khaled

On Wed, Nov 12, 2025 at 1:22 AM Eibe Frank via Weka-users <
wek...@li...> wrote:

> The socket-based transfer of data between WEKA and Python implemented in
> the wekaPython package does not work in all cases currently: there can be
> issues with some unicode, missing values, and dates.
>
> The relevant code is here:
>
>
> https://git.cms.waikato.ac.nz/weka/weka/-/blob/stable-3-8/packages/internal/wekaPython/resources/py/pyServer.py
>
> https://git.cms.waikato.ac.nz/weka/weka/-/tree/stable-3-8/packages/internal/wekaPython/src/main/java/weka/python
>
> You could try modifying the Python code in the installed package to see
> whether you can fix the problem. It will be located in
>
> %HOME%/wekafiles/packages/wekaPython/resources/py/pyServer.py
>
> Alternatively, you could try simplifying the data.
>
> Cheers,
> Eibe
>
> *From: *Khaled EL-Gamily <kha...@gm...>
> *Date: *Friday, 7 November 2025 at 9:09 AM
> *To: *wek...@li... <wek...@li...>
> *Subject: *[Weka-users] transferring a dataset into Python as 'py_data'
> showing the dataset is null
>
> Some people who received this message don't often get email from
> kha...@gm.... Learn why this is important
> <https://aka.ms/LearnAboutSenderIdentification>
> Hi,
>
> When I open my data file (arff) in Weka Explorer, it opens correctly
> showing the right attributes and number of instances, etc.
>
> However, I came across an error in transferring a dataset into Python as
> 'py_data'
> showing the dataset is null.
>
> Please find below weka log info along with system info.
> using Weka 3.9.6
>
> Can you please help me in solving this error.
>
> Thanks,
>
> Khaled Elgamily
>
>
> 09:53:09: Weka Explorer
> 09:53:09: (c) 1999-2022 The University of Waikato, Hamilton, New Zealand
> 09:53:09: web: https://www.cs.waikato.ac.nz/~ml/weka/
> 09:53:09: Started on Wednesday, 5 November 2025
> 09:58:05: Base relation is now ti_signals_training_SUGR (4479 instances)
> 09:58:06: Transferred ti_signals_training_SUGR into R as "rdata"
> 09:58:08: Using default python in PATH
> 09:58:12: Transferred ti_signals_training_SUGR into Python as 'py_data'
> 09:58:12: java.lang.NullPointerException: Cannot invoke
> "weka.core.Instances.numInstances()" because "dataset" is null
> 09:58:49: Started Auto-WEKA for ti_signals_training_SUGR
> 10:06:32: Cannot invoke "java.lang.CharSequence.toString()" because
> "<parameter1>" is null
> 10:30:15: Started Auto-WEKA for ti_signals_training_SUGR
> 10:39:49: null
>
>
> >     java weka.core.SystemInfo
>
> WEKA_HOME: C:\Users\kjami\wekafiles
> file.encoding: Cp1252
> file.separator: \
> java.class.path: C:/Program Files/Weka-3-9-6/weka.jar;
> java.class.version: 61.0
> java.home: C:\Program
> Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64
> java.io.tmpdir: C:\Users\kjami\AppData\Local\Temp\
> java.library.path: C:\Program
> Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin;C:\WINDOWS\Sun\Java\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\Program
> Files\Common Files\Oracle\Java\javapath;C:\Program Files (x86)\Common
> Files\Oracle\Java\java8path;C:\Program Files (x86)\Common
> Files\Oracle\Java\javapath;C:\Program
> Files\Microsoft\jdk-11.0.16.101-hotspot\bin;C:\Program Files\NVIDIA GPU
> Computing Toolkit\CUDA\v11.7\bin;C:\Program Files\NVIDIA GPU Computing
> Toolkit\CUDA\v11.7\libnvvp;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0\;C:\Windows\System32\OpenSSH\;C:\Program
> Files (x86)\NVIDIA Corporation\PhysX\Common;C:\Program
> Files\dotnet\;C:\Program Files\Crucial\Crucial Storage Executive;C:\Program
> Files\NVIDIA Corporation\Nsight Compute 2022.2.0\;C:\Program Files\NVIDIA
> GPU Computing Toolkit\CUDA\v11.7\extras\CUPTI\lib64;C:\Program Files\NVIDIA
> GPU Computing
> Toolkit\CUDA\v11.7\include;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\bin;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\include;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\Program
> Files\nodejs\;C;C:\Users\kjami\AppData\Local\Android\Sdk\platform-tools;C:\Program
> Files\Git\cmd;C:\Users\kjami\anaconda3\envs\vecgi\Scripts;C:\Program
> Files\NVIDIA Corporation\NVIDIA app\NvDLISR;C:\Program
> Files\Amazon\AWSCLIV2\;D:\vcpkg\;C:\Users\kjami\anaconda3\Scripts;C:\Users\kjami\anaconda3\Library\bin;C:\Users\kjami\AppData\Local\Python\bin;C:\Users\kjami\.cargo\bin;C:\Program
> Files\MySQL\MySQL Shell
> 8.0\bin\;C:\Users\kjami\AppData\Local\Microsoft\WindowsApps;C:\Users\kjami\AppData\Roaming\npm;C:\Users\kjami\.dotnet\tools;C:\Program
> Files\JetBrains\CLion
> 2025.1.1\bin;;D:\vcpkg;C:\Users\kjami\AppData\Local\Programs\Microsoft VS
> Code\bin;C:\Users\kjami\AppData\Local\Python\bin;.;C:\Users\kjami\wekafiles\native
> java.net.useSystemProxies: true
> java.runtime.name: OpenJDK Runtime Environment
> java.runtime.version: 17.0.2+8-LTS
> java.specification.name: Java Platform API Specification
> java.specification.vendor: Oracle Corporation
> java.specification.version: 17
> java.vendor: Azul Systems, Inc.
> java.vendor.url: http://www.azul.com/
> java.vendor.url.bug: http://www.azul.com/support/
> java.vendor.version: Zulu17.32+13-CA
> java.version: 17.0.2
> java.version.date: 2022-01-18
> java.vm.compressedOopsMode: Zero based
> java.vm.info: mixed mode, sharing
> java.vm.name: OpenJDK 64-Bit Server VM
> java.vm.specification.name: Java Virtual Machine Specification
> java.vm.specification.vendor: Oracle Corporation
> java.vm.specification.version: 17
> java.vm.vendor: Azul Systems, Inc.
> java.vm.version: 17.0.2+8-LTS
> javax.xml.parsers.DocumentBuilderFactory:
> com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl
> javax.xml.parsers.SAXParserFactory:
> com.sun.org.apache.xerces.internal.jaxp.SAXParserFactoryImpl
> jdk.debug: release
> jna.loaded: true
> jnidispatch.path:
> C:\Users\kjami\AppData\Local\Temp\jna-102071294\jna548758446729928430.dll
> jnlp.jogamp.tmp.cache.root: jln8395679852527007385
> jri.ignore.ule: yes
> line.separator: \r\n
> memory.initial: 512MB (536870912)
> memory.max: 22048MB (23119003648)
> native.encoding: Cp1252
> org.xml.sax.parser: com.sun.org.apache.xerces.internal.parsers.SAXParser
> os.arch: amd64
> os.name: Windows 11
> os.version: 10.0
> path.separator: ;
> r.home: C:\Program Files\R\R-4.5.1
> r.libs.user: C:\\Users\\kjami\\AppData\\Local/R/win-library/4.5
> sun.arch.data.model: 64
> sun.awt.enableExtraMouseButtons: true
> sun.boot.library.path: C:\Program
> Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin
> sun.cpu.endian: little
> sun.cpu.isalist: amd64
> sun.io.unicode.encoding: UnicodeLittle
> sun.java.command: weka.gui.GUIChooser
> sun.java.launcher: SUN_STANDARD
> sun.jnu.encoding: Cp1252
> sun.management.compiler: HotSpot 64-Bit Tiered Compilers
> sun.os.patch.level:
> ui.currentLookAndFeel: com.sun.java.swing.plaf.windows.WindowsLookAndFeel
> ui.installedLookAndFeels:
> javax.swing.plaf.metal.MetalLookAndFeel,javax.swing.plaf.nimbus.NimbusLookAndFeel,com.sun.java.swing.plaf.motif.MotifLookAndFeel,com.sun.java.swing.plaf.windows.WindowsLookAndFeel,com.sun.java.swing.plaf.windows.WindowsClassicLookAndFeel,com.formdev.flatlaf.FlatLightLaf,com.formdev.flatlaf.FlatDarkLaf,com.formdev.flatlaf.FlatIntelliJLaf,com.formdev.flatlaf.FlatDarculaLaf
> user.country: US
> user.dir: C:\Program Files\Weka-3-9-6
> user.home: C:\Users\kjami
> user.language: en
> user.name: kjami
> user.script:
> user.timezone: Africa/Cairo
> user.variant:
> weka.distributed.hadoop.hadoop2: false
> weka.started.via.GUIChooser: true
> weka.version: 3.9.6
>
> _______________________________________________
> Weka-users mailing list
> Wek...@li...
> https://lists.sourceforge.net/lists/listinfo/weka-users
>

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Khaled EL-G. <kha...@gm...>]

Yes, the dataset is always transferred to R.

Thanks,
Khaled

On Wed, Nov 12, 2025 at 2:12 AM Michael Hall <mik...@gm...> wrote:

>
>
> > On Nov 6, 2025, at 4:22 AM, Khaled EL-Gamily <kha...@gm...>
> wrote:
> >
> > Hi,
> >
> > When I open my data file (arff) in Weka Explorer, it opens correctly
> > showing the right attributes and number of instances, etc.
> >
> > However, I came across an error in transferring a dataset into Python as
> 'py_data'
> > showing the dataset is null.
> >
> > Please find below weka log info along with system info.
> > using Weka 3.9.6
> >
> > Can you please help me in solving this error.
> >
> > Thanks,
> >
> > Khaled Elgamily
>
> Can you use the dataset without going to python?
>
> Default Java, or maybe R
> I noticed…
>
> 09:58:06: Transferred ti_signals_training_SUGR into R as “rdata"
>
> This looks familiar but I don’t remember,  is the dataset always
> transferred to R?
>
> _______________________________________________
> Weka-users mailing list
> Wek...@li...
> https://lists.sourceforge.net/lists/listinfo/weka-users
>

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Khaled EL-G. <kha...@gm...>]

Yes, the RPlugin package for WEKA is installed.

Thanks,
Khaled

On Wed, Nov 12, 2025 at 3:02 AM Eibe Frank via Weka-users <
wek...@li...> wrote:

> Yes, if the RPlugin package for WEKA has been installed.
>
> Cheers,
> Eibe
>
> *From: *Michael Hall <mik...@gm...>
> *Date: *Wednesday, 12 November 2025 at 1:12 PM
> *To: *wek...@li... <wek...@li...>
> *Subject: *Re: [Weka-users] transferring a dataset into Python as
> 'py_data' showing the dataset is null
>
>
>
> > On Nov 6, 2025, at 4:22 AM, Khaled EL-Gamily <kha...@gm...>
> wrote:
> >
> > Hi,
> >
> > When I open my data file (arff) in Weka Explorer, it opens correctly
> > showing the right attributes and number of instances, etc.
> >
> > However, I came across an error in transferring a dataset into Python as
> 'py_data'
> > showing the dataset is null.
> >
> > Please find below weka log info along with system info.
> > using Weka 3.9.6
> >
> > Can you please help me in solving this error.
> >
> > Thanks,
> >
> > Khaled Elgamily
>
> Can you use the dataset without going to python?
>
> Default Java, or maybe R
> I noticed…
>
> 09:58:06: Transferred ti_signals_training_SUGR into R as “rdata"
>
> This looks familiar but I don’t remember,  is the dataset always
> transferred to R?
>
> _______________________________________________
> Weka-users mailing list
> Wek...@li...
> https://lists.sourceforge.net/lists/listinfo/weka-users
> _______________________________________________
> Weka-users mailing list
> Wek...@li...
> https://lists.sourceforge.net/lists/listinfo/weka-users
>

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Eibe F. <eib...@wa...>]

Yes, if the RPlugin package for WEKA has been installed.

Cheers,
Eibe

From: Michael Hall <mik...@gm...>
Date: Wednesday, 12 November 2025 at 1:12 PM
To: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null



> On Nov 6, 2025, at 4:22 AM, Khaled EL-Gamily <kha...@gm...> wrote:
>
> Hi,
>
> When I open my data file (arff) in Weka Explorer, it opens correctly
> showing the right attributes and number of instances, etc.
>
> However, I came across an error in transferring a dataset into Python as 'py_data'
> showing the dataset is null.
>
> Please find below weka log info along with system info.
> using Weka 3.9.6
>
> Can you please help me in solving this error.
>
> Thanks,
>
> Khaled Elgamily

Can you use the dataset without going to python?

Default Java, or maybe R
I noticed…

09:58:06: Transferred ti_signals_training_SUGR into R as “rdata"

This looks familiar but I don’t remember,  is the dataset always transferred to R?

_______________________________________________
Weka-users mailing list
Wek...@li...
https://lists.sourceforge.net/lists/listinfo/weka-users

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Michael H. <mik...@gm...>]

> On Nov 6, 2025, at 4:22 AM, Khaled EL-Gamily <kha...@gm...> wrote:
> 
> Hi,
> 
> When I open my data file (arff) in Weka Explorer, it opens correctly
> showing the right attributes and number of instances, etc.
> 
> However, I came across an error in transferring a dataset into Python as 'py_data'
> showing the dataset is null.
> 
> Please find below weka log info along with system info.
> using Weka 3.9.6
> 
> Can you please help me in solving this error.
> 
> Thanks,
> 
> Khaled Elgamily

Can you use the dataset without going to python?

Default Java, or maybe R 
I noticed…

09:58:06: Transferred ti_signals_training_SUGR into R as “rdata"

This looks familiar but I don’t remember,  is the dataset always transferred to R?

Re: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Eibe F. <eib...@wa...>]

The socket-based transfer of data between WEKA and Python implemented in the wekaPython package does not work in all cases currently: there can be issues with some unicode, missing values, and dates.

The relevant code is here:

https://git.cms.waikato.ac.nz/weka/weka/-/blob/stable-3-8/packages/internal/wekaPython/resources/py/pyServer.py
https://git.cms.waikato.ac.nz/weka/weka/-/tree/stable-3-8/packages/internal/wekaPython/src/main/java/weka/python

You could try modifying the Python code in the installed package to see whether you can fix the problem. It will be located in

%HOME%/wekafiles/packages/wekaPython/resources/py/pyServer.py

Alternatively, you could try simplifying the data.

Cheers,
Eibe

From: Khaled EL-Gamily <kha...@gm...>
Date: Friday, 7 November 2025 at 9:09 AM
To: wek...@li... <wek...@li...>
Subject: [Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

Some people who received this message don't often get email from kha...@gm.... Learn why this is important<https://aka.ms/LearnAboutSenderIdentification>
Hi,

When I open my data file (arff) in Weka Explorer, it opens correctly
showing the right attributes and number of instances, etc.

However, I came across an error in transferring a dataset into Python as 'py_data'
showing the dataset is null.

Please find below weka log info along with system info.
using Weka 3.9.6

Can you please help me in solving this error.

Thanks,

Khaled Elgamily


09:53:09: Weka Explorer
09:53:09: (c) 1999-2022 The University of Waikato, Hamilton, New Zealand
09:53:09: web: https://www.cs.waikato.ac.nz/~ml/weka/
09:53:09: Started on Wednesday, 5 November 2025
09:58:05: Base relation is now ti_signals_training_SUGR (4479 instances)
09:58:06: Transferred ti_signals_training_SUGR into R as "rdata"
09:58:08: Using default python in PATH
09:58:12: Transferred ti_signals_training_SUGR into Python as 'py_data'
09:58:12: java.lang.NullPointerException: Cannot invoke "weka.core.Instances.numInstances()" because "dataset" is null
09:58:49: Started Auto-WEKA for ti_signals_training_SUGR
10:06:32: Cannot invoke "java.lang.CharSequence.toString()" because "<parameter1>" is null
10:30:15: Started Auto-WEKA for ti_signals_training_SUGR
10:39:49: null


>     java weka.core.SystemInfo

WEKA_HOME: C:\Users\kjami\wekafiles
file.encoding: Cp1252
file.separator: \
java.class.path: C:/Program Files/Weka-3-9-6/weka.jar;
java.class.version: 61.0
java.home: C:\Program Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64
java.io.tmpdir: C:\Users\kjami\AppData\Local\Temp\
java.library.path: C:\Program Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin;C:\WINDOWS\Sun\Java\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\Program Files\Common Files\Oracle\Java\javapath;C:\Program Files (x86)\Common Files\Oracle\Java\java8path;C:\Program Files (x86)\Common Files\Oracle\Java\javapath;C:\Program Files\Microsoft\jdk-11.0.16.101-hotspot\bin;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\bin;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\libnvvp;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0\;C:\Windows\System32\OpenSSH\;C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common;C:\Program Files\dotnet\;C:\Program Files\Crucial\Crucial Storage Executive;C:\Program Files\NVIDIA Corporation\Nsight Compute 2022.2.0\;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\extras\CUPTI\lib64;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.7\include;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\bin;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\include;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\Program Files\nodejs\;C;C:\Users\kjami\AppData\Local\Android\Sdk\platform-tools;C:\Program Files\Git\cmd;C:\Users\kjami\anaconda3\envs\vecgi\Scripts;C:\Program Files\NVIDIA Corporation\NVIDIA app\NvDLISR;C:\Program Files\Amazon\AWSCLIV2\;D:\vcpkg\;C:\Users\kjami\anaconda3\Scripts;C:\Users\kjami\anaconda3\Library\bin;C:\Users\kjami\AppData\Local\Python\bin;C:\Users\kjami\.cargo\bin;C:\Program Files\MySQL\MySQL Shell 8.0\bin\;C:\Users\kjami\AppData\Local\Microsoft\WindowsApps;C:\Users\kjami\AppData\Roaming\npm;C:\Users\kjami\.dotnet\tools;C:\Program Files\JetBrains\CLion 2025.1.1\bin;;D:\vcpkg;C:\Users\kjami\AppData\Local\Programs\Microsoft VS Code\bin;C:\Users\kjami\AppData\Local\Python\bin;.;C:\Users\kjami\wekafiles\native
java.net.useSystemProxies: true
java.runtime.name<http://java.runtime.name>: OpenJDK Runtime Environment
java.runtime.version: 17.0.2+8-LTS
java.specification.name<http://java.specification.name>: Java Platform API Specification
java.specification.vendor: Oracle Corporation
java.specification.version: 17
java.vendor: Azul Systems, Inc.
java.vendor.url: http://www.azul.com/
java.vendor.url.bug: http://www.azul.com/support/
java.vendor.version: Zulu17.32+13-CA
java.version: 17.0.2
java.version.date: 2022-01-18
java.vm.compressedOopsMode: Zero based
java.vm.info<http://java.vm.info>: mixed mode, sharing
java.vm.name<http://java.vm.name>: OpenJDK 64-Bit Server VM
java.vm.specification.name<http://java.vm.specification.name>: Java Virtual Machine Specification
java.vm.specification.vendor: Oracle Corporation
java.vm.specification.version: 17
java.vm.vendor: Azul Systems, Inc.
java.vm.version: 17.0.2+8-LTS
javax.xml.parsers.DocumentBuilderFactory: com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl
javax.xml.parsers.SAXParserFactory: com.sun.org.apache.xerces.internal.jaxp.SAXParserFactoryImpl
jdk.debug: release
jna.loaded: true
jnidispatch.path: C:\Users\kjami\AppData\Local\Temp\jna-102071294\jna548758446729928430.dll
jnlp.jogamp.tmp.cache.root: jln8395679852527007385
jri.ignore.ule: yes
line.separator: \r\n
memory.initial: 512MB (536870912)
memory.max: 22048MB (23119003648)
native.encoding: Cp1252
org.xml.sax.parser: com.sun.org.apache.xerces.internal.parsers.SAXParser
os.arch: amd64
os.name<http://os.name>: Windows 11
os.version: 10.0
path.separator: ;
r.home: C:\Program Files\R\R-4.5.1
r.libs.user: C:\\Users\\kjami\\AppData\\Local/R/win-library/4.5
sun.arch.data.model: 64
sun.awt.enableExtraMouseButtons: true
sun.boot.library.path: C:\Program Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin
sun.cpu.endian: little
sun.cpu.isalist: amd64
sun.io.unicode.encoding: UnicodeLittle
sun.java.command: weka.gui.GUIChooser
sun.java.launcher: SUN_STANDARD
sun.jnu.encoding: Cp1252
sun.management.compiler: HotSpot 64-Bit Tiered Compilers
sun.os.patch.level:
ui.currentLookAndFeel: com.sun.java.swing.plaf.windows.WindowsLookAndFeel
ui.installedLookAndFeels: javax.swing.plaf.metal.MetalLookAndFeel,javax.swing.plaf.nimbus.NimbusLookAndFeel,com.sun.java.swing.plaf.motif.MotifLookAndFeel,com.sun.java.swing.plaf.windows.WindowsLookAndFeel,com.sun.java.swing.plaf.windows.WindowsClassicLookAndFeel,com.formdev.flatlaf.FlatLightLaf,com.formdev.flatlaf.FlatDarkLaf,com.formdev.flatlaf.FlatIntelliJLaf,com.formdev.flatlaf.FlatDarculaLaf
user.country: US
user.dir: C:\Program Files\Weka-3-9-6
user.home: C:\Users\kjami
user.language: en
user.name<http://user.name>: kjami
user.script:
user.timezone: Africa/Cairo
user.variant:
weka.distributed.hadoop.hadoop2: false
weka.started.via.GUIChooser: true
weka.version: 3.9.6

[Weka-users] transferring a dataset into Python as 'py_data' showing the dataset is null

[Khaled EL-G. <kha...@gm...>]

Hi,

When I open my data file (arff) in Weka Explorer, it opens correctly
showing the right attributes and number of instances, etc.

However, I came across an error in transferring a dataset into Python as
'py_data'
showing the dataset is null.

Please find below weka log info along with system info.
using Weka 3.9.6

Can you please help me in solving this error.

Thanks,

Khaled Elgamily


09:53:09: Weka Explorer
09:53:09: (c) 1999-2022 The University of Waikato, Hamilton, New Zealand
09:53:09: web: https://www.cs.waikato.ac.nz/~ml/weka/
09:53:09: Started on Wednesday, 5 November 2025
09:58:05: Base relation is now ti_signals_training_SUGR (4479 instances)
09:58:06: Transferred ti_signals_training_SUGR into R as "rdata"
09:58:08: Using default python in PATH
09:58:12: Transferred ti_signals_training_SUGR into Python as 'py_data'
09:58:12: java.lang.NullPointerException: Cannot invoke
"weka.core.Instances.numInstances()" because "dataset" is null
09:58:49: Started Auto-WEKA for ti_signals_training_SUGR
10:06:32: Cannot invoke "java.lang.CharSequence.toString()" because
"<parameter1>" is null
10:30:15: Started Auto-WEKA for ti_signals_training_SUGR
10:39:49: null


>     java weka.core.SystemInfo

WEKA_HOME: C:\Users\kjami\wekafiles
file.encoding: Cp1252
file.separator: \
java.class.path: C:/Program Files/Weka-3-9-6/weka.jar;
java.class.version: 61.0
java.home: C:\Program
Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64
java.io.tmpdir: C:\Users\kjami\AppData\Local\Temp\
java.library.path: C:\Program
Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin;C:\WINDOWS\Sun\Java\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\Program
Files\Common Files\Oracle\Java\javapath;C:\Program Files (x86)\Common
Files\Oracle\Java\java8path;C:\Program Files (x86)\Common
Files\Oracle\Java\javapath;C:\Program
Files\Microsoft\jdk-11.0.16.101-hotspot\bin;C:\Program Files\NVIDIA GPU
Computing Toolkit\CUDA\v11.7\bin;C:\Program Files\NVIDIA GPU Computing
Toolkit\CUDA\v11.7\libnvvp;C:\Windows\system32;C:\Windows;C:\Windows\System32\Wbem;C:\Windows\System32\WindowsPowerShell\v1.0\;C:\Windows\System32\OpenSSH\;C:\Program
Files (x86)\NVIDIA Corporation\PhysX\Common;C:\Program
Files\dotnet\;C:\Program Files\Crucial\Crucial Storage Executive;C:\Program
Files\NVIDIA Corporation\Nsight Compute 2022.2.0\;C:\Program Files\NVIDIA
GPU Computing Toolkit\CUDA\v11.7\extras\CUPTI\lib64;C:\Program Files\NVIDIA
GPU Computing
Toolkit\CUDA\v11.7\include;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\bin;C:\tools\cudnn-windows-x86_64-8.4.1.50_cuda11.6-archive\include;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\Program
Files\nodejs\;C;C:\Users\kjami\AppData\Local\Android\Sdk\platform-tools;C:\Program
Files\Git\cmd;C:\Users\kjami\anaconda3\envs\vecgi\Scripts;C:\Program
Files\NVIDIA Corporation\NVIDIA app\NvDLISR;C:\Program
Files\Amazon\AWSCLIV2\;D:\vcpkg\;C:\Users\kjami\anaconda3\Scripts;C:\Users\kjami\anaconda3\Library\bin;C:\Users\kjami\AppData\Local\Python\bin;C:\Users\kjami\.cargo\bin;C:\Program
Files\MySQL\MySQL Shell
8.0\bin\;C:\Users\kjami\AppData\Local\Microsoft\WindowsApps;C:\Users\kjami\AppData\Roaming\npm;C:\Users\kjami\.dotnet\tools;C:\Program
Files\JetBrains\CLion
2025.1.1\bin;;D:\vcpkg;C:\Users\kjami\AppData\Local\Programs\Microsoft VS
Code\bin;C:\Users\kjami\AppData\Local\Python\bin;.;C:\Users\kjami\wekafiles\native
java.net.useSystemProxies: true
java.runtime.name: OpenJDK Runtime Environment
java.runtime.version: 17.0.2+8-LTS
java.specification.name: Java Platform API Specification
java.specification.vendor: Oracle Corporation
java.specification.version: 17
java.vendor: Azul Systems, Inc.
java.vendor.url: http://www.azul.com/
java.vendor.url.bug: http://www.azul.com/support/
java.vendor.version: Zulu17.32+13-CA
java.version: 17.0.2
java.version.date: 2022-01-18
java.vm.compressedOopsMode: Zero based
java.vm.info: mixed mode, sharing
java.vm.name: OpenJDK 64-Bit Server VM
java.vm.specification.name: Java Virtual Machine Specification
java.vm.specification.vendor: Oracle Corporation
java.vm.specification.version: 17
java.vm.vendor: Azul Systems, Inc.
java.vm.version: 17.0.2+8-LTS
javax.xml.parsers.DocumentBuilderFactory:
com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl
javax.xml.parsers.SAXParserFactory:
com.sun.org.apache.xerces.internal.jaxp.SAXParserFactoryImpl
jdk.debug: release
jna.loaded: true
jnidispatch.path:
C:\Users\kjami\AppData\Local\Temp\jna-102071294\jna548758446729928430.dll
jnlp.jogamp.tmp.cache.root: jln8395679852527007385
jri.ignore.ule: yes
line.separator: \r\n
memory.initial: 512MB (536870912)
memory.max: 22048MB (23119003648)
native.encoding: Cp1252
org.xml.sax.parser: com.sun.org.apache.xerces.internal.parsers.SAXParser
os.arch: amd64
os.name: Windows 11
os.version: 10.0
path.separator: ;
r.home: C:\Program Files\R\R-4.5.1
r.libs.user: C:\\Users\\kjami\\AppData\\Local/R/win-library/4.5
sun.arch.data.model: 64
sun.awt.enableExtraMouseButtons: true
sun.boot.library.path: C:\Program
Files\Weka-3-9-6\jre\zulu17.32.13-ca-fx-jre17.0.2-win_x64\bin
sun.cpu.endian: little
sun.cpu.isalist: amd64
sun.io.unicode.encoding: UnicodeLittle
sun.java.command: weka.gui.GUIChooser
sun.java.launcher: SUN_STANDARD
sun.jnu.encoding: Cp1252
sun.management.compiler: HotSpot 64-Bit Tiered Compilers
sun.os.patch.level:
ui.currentLookAndFeel: com.sun.java.swing.plaf.windows.WindowsLookAndFeel
ui.installedLookAndFeels:
javax.swing.plaf.metal.MetalLookAndFeel,javax.swing.plaf.nimbus.NimbusLookAndFeel,com.sun.java.swing.plaf.motif.MotifLookAndFeel,com.sun.java.swing.plaf.windows.WindowsLookAndFeel,com.sun.java.swing.plaf.windows.WindowsClassicLookAndFeel,com.formdev.flatlaf.FlatLightLaf,com.formdev.flatlaf.FlatDarkLaf,com.formdev.flatlaf.FlatIntelliJLaf,com.formdev.flatlaf.FlatDarculaLaf
user.country: US
user.dir: C:\Program Files\Weka-3-9-6
user.home: C:\Users\kjami
user.language: en
user.name: kjami
user.script:
user.timezone: Africa/Cairo
user.variant:
weka.distributed.hadoop.hadoop2: false
weka.started.via.GUIChooser: true
weka.version: 3.9.6

Re: [Weka-users] Predictions limited to 100 results

[Peter R. <fra...@gm...>]

> It's working now!
> you were right, the problem was in csv file.
>
>
> Besides, Is there any simple and easy to understand way to learn more about WEKA?

The late Prof Ian Witten created three MOOCs for Weka, which teach you
a lot about Weka and its algorithms. You can still view the videos on
youtube:
https://ml.cms.waikato.ac.nz/weka/courses.html

> I'm working in Telecom industry and interested in utilizing WEKA in my daily tasks.

If you prefer Python, then you might find my python-weka-wrapper3
library useful, which uses Weka (running in a JVM) behind the scenes:
https://github.com/fracpete/python-weka-wrapper3

For more demanding tasks, a more advanced/flexible Explorer interface
(called Weka Investigator) and a more capable workflow engine you can
also have a look at our ADAMS framework:
https://adams.cms.waikato.ac.nz/

I haven't had time lately to make an ADAMS release, so I recommend
downloading a snapshot, e.g., the "adams-ml-app-snapshot" one:
https://adams.cms.waikato.ac.nz/download/snapshot/

Cheers, Peter

Re: [Weka-users] Predictions limited to 100 results

[Peter R. <fra...@gm...>]

Please reply to the list.

> I have attached Two files that I'm working on.
> Actually I want WEKA to predict "Config" column from "Master File-test set.csv" based on "Master File-Training Set.csv".
> What do you suggest?
>
>
> P.S I used ARFF format as well but didn't work. Maybe I'm going the wrong way...

If things don't work after converting to ARFF, it's usually a good
indicator that something is not quite right.

CSV files are often problematic, as Weka cannot infer the types
properly. E.g., your test set does not contain any labels for the
"Config" attribute.
Also, your there is a discrepancy between your attribute names:
- train: DL SE  - Last Month
- test: DL Spectral efficiency  - Last Month

Also, I recommend removing any trailing blanks/tabs from your attribute names.

Here's what I did to convert the CSV files:
- loaded each into the Explorer using the CSVLoader, using custom
options (check box top right of open dialog):
  specified "1" for the string attributes property, as they only seem
to be an ID and differ between train/test
- saved the file as ARFF

To make the compatible:
- train: changed the name of the 2nd attribute in the header section to
  @attribute 'DL Spectral efficiency  - Last Month   ' numeric
- test: replaced the line with the "Config" attribute definition with
the one from the train set

Having these two ARFF files in place, you get your 500 predictions.

I will send you the converted ARFF files in a separate email.

Cheers, Peter

Re: [Weka-users] Predictions limited to 100 results

[Peter R. <fra...@gm...>]

> hereunder are the steps I took:
>
> Explorer-->Preprocess-->Open File-->Training Set.csv  (my file has 9 columns and 1800 rows)
> Classify-->Classifier-->choose-->J48
> Test Options-->"Use Training set"-->Start

Any particular reason that you are evaluating on the training set?
This will typically give you results that are too optimistic.

> Test Options-->"Supplied Test Set"-->"Set"-->"Test instances"-->"Open file"-->Test Set.csv (my file has 9 columns and 200 rows, the 9th column is empty)--> Close.
> Test Options-->"more options"-->Tick "Output Entropy evaluation measures" THEN "output predictions" -->Choose-->Plain Text --> OK
> Result List-->right Click-->"Re-Evaluate model on current test set"
>
> In "classifier Output" I can see just 100 rows !

I still can't replicate your issue, unfortunately.

Here is what I did:

- Split the waveform-5000 dataset into train (66.7%) and test (33.3%)
using the RemovePercentage filter.
- In the Preprocess tab, I loaded the train set.
- Switched to the Classify tab, selected J48 (default settings)
- Selected "Supplied test set", clicked on "Set" and select the test
set and clicked on "Close"
- Clicked on "More options"
  ~ Ticked "Output entropy eval measures"
  ~ Selected "Plain text" for "Output predictions"
  ~ Accepted the settings with "OK" button
- Clicked on "Start"

The output showed the correct 33.3% number of rows (1665).

There is one thing that I'd like you to try: convert your CSV files to
ARFF and then follow the steps that I outlined above.

Cheers, Peter

Re: [Weka-users] Predictions limited to 100 results

[Peter R. <fra...@gm...>]

> I’m using Weka 3.8.6 and noticed that in Explorer the “Predictions on test set” output only shows 100 rows.
> Could you please explain how I can export all predictions?
> Please let me know if I provide more information.

I just trained J48 on the waveform-5000.arff UCI dataset and also used
that file for testing. The "Predictions on test set" section contained
5000 rows with predictions as expected (as many as there are in the
test set).

Please provide more details on your datasets (train/test), classifier,
chosen format for outputting the predictions, etc.

Cheers, Peter

Re: [Weka-users] Weka Ranker (search) Attribute without Conjunction of Attribute Evaluators

[Peter R. <fra...@gm...>]

>                                Hope you are doing well , I need some help regarding a feature of Weka, below is the description of
>
> I assign score (weight) to my features using my own feature ranking technique,
>
> Now inside of AttributeSelectedClassifier I want to use the top 10 features to check their classification performance, after then 20 features , and 50 features so on..
>
> I want to use ranking (to select top features) without conjunction of attribute evaluators (ReliefF, GainRatio, InformationGain, Entropy etc)..
>
>
> If there any possibility I can use Ranker (Search) without attribute evaluators ?

No, Ranker works only in tandem with an attribute evaluation scheme.

Have you thought about implementing a *dummy* attribute evaluator (see
below for interface and super class) that just returns the attribute
weights that you have assigned as the merit for an attribute? Then you
can use Ranker in conjunction with this dummy evaluator.

Interface:
https://git.cms.waikato.ac.nz/weka/weka/-/blob/main/trunk/weka/src/main/java/weka/attributeSelection/AttributeEvaluator.java

Superclass:
https://git.cms.waikato.ac.nz/weka/weka/-/blob/main/trunk/weka/src/main/java/weka/attributeSelection/ASEvaluation.java

Cheers, Peter

[Weka-users] Predictions limited to 100 results

[Mohammad J. N. <mne...@gm...>]

Hello dear Weka Team,

I’m using Weka 3.8.6 and noticed that in Explorer the “Predictions on test
set” output only shows 100 rows.
Could you please explain how I can export all predictions?
Please let me know if I provide more information.

Thanks,
Javad

[Weka-users] Weka Ranker (search) Attribute without Conjunction of Attribute Evaluators

[Muhammad N. <ca...@gm...>]

Dear Sir/Madam:

                               Hope you are doing well , I need some help
regarding a feature of Weka, below is the description of

I assign score (weight) to my features using my own feature ranking
technique,

Now inside of AttributeSelectedClassifier I want to use the top 10 features
to check their classification performance, after then 20 features , and 50
features so on..

I want to use ranking (to select top features) without conjunction of
attribute evaluators (ReliefF, GainRatio, InformationGain, Entropy etc)..


If there any possibility I can use Ranker (Search) without attribute
evaluators ?

Looking forward

Regards:
               M.Nasir (+92 303 0300 666)

Re: [Weka-users] How to add a new classifier to the package manager?

[Michael H. <mik...@gm...>]

Might be a start

https://waikato.github.io/weka-wiki/packages/structure/

As I remember there are a couple lines of code to get your own java code to search for the packages. But I’m not finding anything at the moment that does that.
The following seems to show that
java weka.Run
Will do that for you?

https://waikato.github.io/weka-wiki/packages/manager/

> On Jun 4, 2025, at 1:37 PM, Miguel Garcia Torres <mga...@up...> wrote:
> 
> Dear all,
> 
> I have a new Classifier and I would like to add it to the package manager. I am looking for a blog, document, etc. explaining the steps. I am sorry if this information is already available. I have unsuccessfully tried to find information on how to add a classifier to the package manager. I have searched both online and in the mailing list. Thank you in advance.
> 
> Best regards,
> 
> MiguelGT
> 
> -- 
> Miguel García Torres, PhD
> Data Science & Big Data Lab
> Universidad Pablo de Olavide
> https://datalab.upo.es/profile/garcia/
> ResearcherID: K-5782-2014
> Scopus Author ID: 8918249300
> ORCID id: orcid.org/0000-0002-6867-7080
> Google Scholar: https://scholar.google.es/citations?user=x7oxkrAAAAAJ&hl=es
> 
> 
> 
> _______________________________________________
> Weka-users mailing list
> Wek...@li...
> https://lists.sourceforge.net/lists/listinfo/weka-users

[Weka-users] How to add a new classifier to the package manager?

[Miguel G. T. <mga...@up...>]

Dear all,

I have a new Classifier and I would like to add it to the package 
manager. I am looking for a blog, document, etc. explaining the steps. I 
am sorry if this information is already available. I have unsuccessfully 
tried to find information on how to add a classifier to the package 
manager. I have searched both online and in the mailing list. Thank you 
in advance.

Best regards,

MiguelGT

-- 
Miguel García Torres, PhD
Data Science & Big Data Lab
Universidad Pablo de Olavide
https://datalab.upo.es/profile/garcia/
ResearcherID: K-5782-2014
Scopus Author ID: 8918249300
ORCID id: orcid.org/0000-0002-6867-7080
Google Scholar: https://scholar.google.es/citations?user=x7oxkrAAAAAJ&hl=es

[Weka-users] Developing for Java-21

[John R. <jr...@in...>]

I've vibe coded with gemini into a java 21 working version of weka

https://github.com/jrobens/weka


-- 
Full-stack developer java/typescript. Payments, content management systems.
http://interlated.com.au
+61 434 996 607
https://www.linkedin.com/in/johnrobens/

[Weka-users] arff file uploader patch

[John R. <jr...@in...>]

I had trouble processing some arff converted by
https://github.com/anaavila/convert-csv-to-arff/blob/master/csv-to-arff.py
which I have also modified.

The main change I made was to get '%' characters to process. This file now
processes a 25k line green bond classification database export.

I had to filter out all quotes.  I can't get quotes to be quoted, so I've
removed them all. The attribute loader just falls apart. Each word after
creates its own value, inevitably creating a duplicate attribute at the
end. I tried detecting what type of quote the field was quoted in and not
treating embedded quotes of other types as quotes, but there does not
seem to be a way.

1. At the start of parseAttribute set % to ordinary character

m_Tokenizer.ordinaryChar('%');

Set it back at the end


m_Tokenizer.commentChar('%');


2. Ignore nulls


// Nulls confuse things.
// us dollars) will be allocated within ifc's treasury (ifc's)
while (m_Tokenizer.nextToken() != '}') {
  if (m_Tokenizer.sval == null) {
    continue;
  }


3. Quotes.


I could not resolve this.

Change to java.util.Scanner? Personally I'd use a CSV library. Apache
Commons CSV? Pass the string through.;


> Escaping quotes doesn't help


"mixed use under the ifc green bond framework : renewables, energy
efficiency (energy savings in existing industrial operations,
utilities, commerical and residential buildings, technicall loss
reductions in transmission and distribution operations, energy
efficiency improvements in existing thermal power plants,green
buildings, manufacturing, cogeneration, waste to heat recovery),
agriculture and land use, waste management, transport, other
mitigation.",the net proceeds of the issue of the notes (which may be
converted into us dollars) will be allocated within ifc\'s treasury to
a special sub-portfolio that will be linked to ifc\'s lending
operations for climate projects (\"eligible projects\").,"financing
the capital expenditure programme of the energy development
corporation (pse:edc), which focuses on optimising the output of its
geothermal power plants and improving resiliency to climate
impacts.","renewable energy and associated infrastructures, green
buildings and retrofits, landuse (agriculture, forestry,
conservation), waste, wastewater management, manufacturing, low carbon
transport, a&r","renewable energy examples: • solar • wind • waves •
fossil       free    district        heating • geothermal energy
efficiency existing buildings • at      least   25%     reduction
 in      estimated       energy   use    or      documented
support from    enova. energy    efficiency     new      buildings •
new        buildings:      breeam excellent        or
outstanding.   a


> Single inside double doesn't work


Double quotes then single quotes

'non-recourse project finance facility of up to jpy16 billion ($149
million) from shinsei bank, limited ("shinsei bank") for its 53.4 mwp
oita hijimachi solar power project. the non-recourse facility has an
18.5-year tenor, with d



Thanks

John


-- 
Full-stack developer java/typescript. Payments, content management systems.
http://interlated.com.au
+61 434 996 607
https://www.linkedin.com/in/johnrobens/

[Weka-users] Fwd: [Wekalist] Re: adding a new ML algorithm to the WEKA distribution

[Joel R. <ra...@ar...>]

Dear Eibe

Thanks for your explanation.  I created a weka package of my implementation
of my algorithm LW which was published in 2020 (details are in the
documentation).

The package has three native libraries, for Windows, Linux and Mac OS. I
have tested it by reading the package (a .zip file) as an unofficial one
into Weka's package manager.

I would like to submit it for inclusion  in the weka repository.

Do I need to upload the .zip somewhere?

Thank you,

Joel


On Mon, 9 Sept 2024 at 04:11, Eibe Frank <eib...@wa...> wrote:

>
>

> New algorithms can be made available as packages:
>
>
>
>   https://waikato.github.io/weka-wiki/packages/
>
>
>
> Ideally, your native code would be compiled for all the platforms that
> WEKA commonly runs on. Yes, it is possible to have native code.
>
>
>
> Cheers,
>
> Eibe
>
>
>

Re: [Weka-users] GUI bug in latest stable WEKA

[Michael H. <mik...@gm...>]

> On Mar 24, 2025, at 4:17 PM, Peter Reutemann <fra...@wa...> wrote:
> 
> I'm not a Windows user,

I’m not either. But out of curiosity I tried current Weka 3.8.6 on my old Intel Mac with both VirtualBox Windows 10 and Bootcamp Windows 10 and couldn’t recreate.

So I’m not sure JDK is the OP’s 3.8.6 problem. 

But that is still not exactly native Windows. 

Re: [Weka-users] GUI bug in latest stable WEKA

[Eibe F. <eib...@wa...>]

Thanks for looking into this. It does not sound like it is an issue in the WEKA code. It is probably an interaction between your particular Windows installation and the Java Virtual Machine that comes with WEKA.

If you need to get WEKA to run properly on your version of Windows, it would be worth trying to run it with a recent release of a Java (8) Virtual Machine, e.g., using

  java -jar weka.jar

from a terminal window (assuming you are in the folder containing weka.jar). Simply double-clicking on weka.jar should also work.

This assumes you have a Java 8 Virtual Machine installed on your system. If you want to use a version > 8, you may need additional command-line arguments or adapt one of the start-up scripts that come with the WEKA distribution.

Cheers,
Eibe

From: Valtteri Asikainen <val...@tu...>
Date: Tuesday, 25 March 2025 at 9:56 AM
To: Michael Hall <mik...@gm...>
Cc: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] GUI bug in latest stable WEKA
Jotkut tämän viestin vastaanottaneet henkilöt eivät usein saa sähköpostia lähettäjältä val...@tu.... Lue, miksi tämä on tärkeää<https://aka.ms/LearnAboutSenderIdentification>
I have now tested the different versions on Windows 10 and 11.

On Windows 10 3.7.13 works, but both 3.8.6 and 3.9.6 are broken.

On Windows 11 I couldn't replicate the bug.

I think I'll be fine using 3.7 or another machine, but this does seem like a bug in the current stable and development versions that affects at least Windows 10 (and possibly older Windows versions which are past EOL anyway). While Windows 10 EOL is closing in, LTS will still last at least until 2027 and people and organizations could be using it.


________________________________
From: Michael Hall <mik...@gm...>
Sent: Sunday, March 23, 2025 2:51:51 AM
To: Valtteri Asikainen <val...@tu...>
Cc: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] GUI bug in latest stable WEKA

You could try an older version of Weka, or the current 3.9 development version.

Other than that I’m not sure.


On Mar 22, 2025, at 3:43 PM, Valtteri Asikainen <val...@tu...> wrote:

I've tried with my data or no data at all, just an empty Explorer window. Happens with both.

Could be a Windows specific bug, or even Windows 10 specific. I'll try on some other devices as well when I have the time.


________________________________
From: Michael Hall <mik...@gm...>
Sent: Saturday, March 22, 2025 10:33:21 PM
To: Valtteri Asikainen <val...@tu...>
Cc: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] GUI bug in latest stable WEKA

Only on this data or any data?

You are doing nothing else on the window?

Otherwise it would seem to be a general bug in Weka or Swing.

I can’t recreate on MacOS 3.9.6. I either get tooltips or nothing. But no display overlays.

On Mar 21, 2025, at 9:04 AM, Valtteri Asikainen <val...@tu...> wrote:


Possible triggers I've identified so far are 1) hovering over the greyed-out filter "Stop" button long enough to get the tooltip pop-up, and 2) often but not always - hovering over the visualization in the bottom right corner.

Re: [Weka-users] GUI bug in latest stable WEKA

[Peter R. <fra...@wa...>]

> I have now tested the different versions on Windows 10 and 11.
>
> On Windows 10 3.7.13 works, but both 3.8.6 and 3.9.6 are broken.
>
> On Windows 11 I couldn't replicate the bug.
>
> I think I'll be fine using 3.7 or another machine, but this does seem 
> like a bug in the current stable and development versions that affects 
> at least Windows 10 (and possibly older Windows versions which are 
> past EOL anyway). While Windows 10 EOL is closing in, LTS will still 
> last at least until 2027 and people and organizations could be using it.

I'm not a Windows user, so can't really comment. I'm just wondering 
whether the difference is the JVM version that is being distributed with 
the Windows installs. You could try installing another JDK and start up 
Weka directly from the command-prompt with that specific JDK.

Steps:
- download a JDK from here (ZIP):
https://adoptium.net/temurin/releases/?variant=openjdk11&os=windows&arch=x64&package=jdk
- unzip the archive into a directory
- open a command-prompt
- set the JAVA_HOME variable to the JDK directory that sits just above 
the "bin" directory
   set JAVA_HOME=C:\....
- navigate to the directory that holds your weka.jar
- launch Weka
   "%JAVA_HOME%"\bin\java -jar weka.jar

Cheers, Peter
-- 
Peter Reutemann
Dept. of Computer Science
University of Waikato, Hamilton, NZ
Mobile +64 22 190 2375
https://profiles.waikato.ac.nz/peter.reutemann
http://www.data-mining.co.nz/

Re: [Weka-users] GUI bug in latest stable WEKA

[Valtteri A. <val...@tu...>]

I have now tested the different versions on Windows 10 and 11.

On Windows 10 3.7.13 works, but both 3.8.6 and 3.9.6 are broken.

On Windows 11 I couldn't replicate the bug.

I think I'll be fine using 3.7 or another machine, but this does seem like a bug in the current stable and development versions that affects at least Windows 10 (and possibly older Windows versions which are past EOL anyway). While Windows 10 EOL is closing in, LTS will still last at least until 2027 and people and organizations could be using it.


________________________________
From: Michael Hall <mik...@gm...>
Sent: Sunday, March 23, 2025 2:51:51 AM
To: Valtteri Asikainen <val...@tu...>
Cc: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] GUI bug in latest stable WEKA

You could try an older version of Weka, or the current 3.9 development version.

Other than that I’m not sure.

On Mar 22, 2025, at 3:43 PM, Valtteri Asikainen <val...@tu...> wrote:

I've tried with my data or no data at all, just an empty Explorer window. Happens with both.

Could be a Windows specific bug, or even Windows 10 specific. I'll try on some other devices as well when I have the time.


________________________________
From: Michael Hall <mik...@gm...>
Sent: Saturday, March 22, 2025 10:33:21 PM
To: Valtteri Asikainen <val...@tu...>
Cc: wek...@li... <wek...@li...>
Subject: Re: [Weka-users] GUI bug in latest stable WEKA

Only on this data or any data?

You are doing nothing else on the window?

Otherwise it would seem to be a general bug in Weka or Swing.

I can’t recreate on MacOS 3.9.6. I either get tooltips or nothing. But no display overlays.

On Mar 21, 2025, at 9:04 AM, Valtteri Asikainen <val...@tu...> wrote:


Possible triggers I've identified so far are 1) hovering over the greyed-out filter "Stop" button long enough to get the tooltip pop-up, and 2) often but not always - hovering over the visualization in the bottom right corner.

Re: [Weka-users] GUI bug in latest stable WEKA

[Michael H. <mik...@gm...>]

Sounds like you’ve nailed it down some.
Hopefully you hear from the Weka people following this.
You might try something on StackOverflow if you don’t.

> On Mar 24, 2025, at 3:58 AM, Valtteri Asikainen <val...@tu...> wrote:
> 
> I have now tested the different versions on Windows 10 and 11. 
> 
> On Windows 10 3.7.13 works, but both 3.8.6 and 3.9.6 are broken. 
> 
> On Windows 11 I couldn't replicate the bug.
> 
> I think I'll be fine using 3.7 or another machine, but this does seem like a bug in the current stable and development versions that affects at least Windows 10 (and possibly older Windows versions which are past EOL anyway). While Windows 10 EOL is closing in, LTS will still last at least until 2027 and people and organizations could be using it.
>