Hi,
We are currently running a study on some potential usability-related changes that could be applied to Weka and we would like to invite you to help us answer some questions around this area. The survey is estimated to take about 5 minutes of your time. No personally identifying information will be collected and your answers will be anonymous. Survey: https://waikato.qualtrics.com/jfe/form/SV_5jXqK1EZk1ZDinz Regards, Bill _______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html |
Hi, Just answered the survey. Forgot saying one comment. To me the MOST important missing feature of Weka is a nice CSV loader such as the one from Rstudio or RapidMiner, one that can visually help the user to specify the separator field, and can intelligently guess the nature of each attribute. I find the current CSVLoader extremely counterintuitive, and I always end up formatting the arff file by myself in code. This is a big barrier for non-programmers users. Cheers, Felipe On Fri, Nov 20, 2020 at 11:50 PM Bill Li <[hidden email]> wrote:
_______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html |
> Just answered the survey. Forgot saying one comment.
> To me the MOST important missing feature of Weka is a nice CSV loader such as the one from Rstudio or RapidMiner, one that can visually help the user to specify the separator field, and can intelligently guess the nature of each attribute. I find the current CSVLoader extremely counterintuitive, and I always end up formatting the arff file by myself in code. This is a big barrier for non-programmers users. Changing the way the CSVLoader works in Weka would break existing code, unfortunately. Rather than changing the loader, I committed some changes today that allow you to have alternative loaders for file types in the user interface. I also have an unofficial package for CSV loader/saver that uses the Apache Common CSV library: https://github.com/fracpete/common-csv-weka-package The loader is quite basic, as all non-numeric columns get imported as STRING attributes. You need to perform attribute conversions post-import (eg StringToNominal). However, the Common CSV library handles multi-line cells, so quite useful for importing textual data that doesn't have the new-lines escaped. Cheers, Peter -- Peter Reutemann Dept. of Computer Science University of Waikato, NZ +64 (7) 577-5304 http://www.cms.waikato.ac.nz/~fracpete/ http://www.data-mining.co.nz/ _______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html |
In reply to this post by Bill Li
Theme is importand but I would like to see Weka oriented more towards usability. Less academic. For example, the results of algorithms are in a format that are difficult to process.
Same with data import.
Sent: Thursday, November 19, 2020 at 7:26 AM
From: "Bill Li" <[hidden email]> To: [hidden email] Subject: [Wekalist] Weka Survey Hi,
_______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.htmlWe are currently running a study on some potential usability-related changes that could be applied to Weka and we would like to invite you to help us answer some questions around this area. The survey is estimated to take about 5 minutes of your time. No personally identifying information will be collected and your answers will be anonymous. Survey: https://waikato.qualtrics.com/jfe/form/SV_5jXqK1EZk1ZDinz Regards, Bill
_______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html |
In reply to this post by Peter Reutemann
> > Just answered the survey. Forgot saying one comment.
> > To me the MOST important missing feature of Weka is a nice CSV loader such as the one from Rstudio or RapidMiner, one that can visually help the user to specify the separator field, and can intelligently guess the nature of each attribute. I find the current CSVLoader extremely counterintuitive, and I always end up formatting the arff file by myself in code. This is a big barrier for non-programmers users. > > Changing the way the CSVLoader works in Weka would break existing > code, unfortunately. > > Rather than changing the loader, I committed some changes today that > allow you to have alternative loaders for file types in the user > interface. > > I also have an unofficial package for CSV loader/saver that uses the > Apache Common CSV library: > https://github.com/fracpete/common-csv-weka-package > > The loader is quite basic, as all non-numeric columns get imported as > STRING attributes. You need to perform attribute conversions > post-import (eg StringToNominal). > However, the Common CSV library handles multi-line cells, so quite > useful for importing textual data that doesn't have the new-lines > escaped. I just released a new version of my CSV loader/saver alternative: 2020.12.29 - can specify ranges for NOMINAL, STRING, DATE attributes - date format for DATE attributes can be specified - nominal label specs can be specified (e.g., forcing labels and/or order) - handles spreadsheets without header row (uses att-X with X from 1 to last column) - custom column names can be supplied (e.g., to override auto-generated or ones in file) - field-separator and quote character can be specified for non-tab-delimited CSV formats - missing value string can be specified Cheers, Peter -- Peter Reutemann Dept. of Computer Science University of Waikato, NZ +64 (7) 577-5304 http://www.cms.waikato.ac.nz/~fracpete/ http://www.data-mining.co.nz/ _______________________________________________ Wekalist mailing list -- [hidden email] Send posts to [hidden email] To unsubscribe send an email to [hidden email] To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html |
Free forum by Nabble | Edit this page |