Skip to content

Adaptation of the AUCCalculator by Davis & Goadrich. My very small update prints out PR points for the original data. You should cite Davis & Goadrich if you use this tool.

License

Notifications You must be signed in to change notification settings

deborah-chasman/auc_orig_points

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

auc_orig_points

This is an adaptation of the original AUCCalculator by Jesse Davis and Mark Goadrich, which is available here: http://mark.goadrich.com/programs/AUC/

The original paper can be found here: The Relationship Between Precision-Recall and ROC Curves Jesse Davis and Mark Goadrich 23rd International Conference on Machine Learning (ICML), Pittsburgh, PA, USA, 26th - 28th June, 2006

I received the original source from Kendrick Boyd. It has some relationship to the Roc project by kboyd & afbarnard, which is released under the FreeBSD License. Please see the LICENSE.txt file and this project: https://github.com/kboyd/Roc

Finally, my update simply prints out the P/R coordinates for the original data points. They will go into a file name ".opr".

Recommended Usage

The original program had multiple usage options, but I have only tested my update with the following usage: java -jar auc_orig_points -t list -o OUTPUTPREFIX FILE

List file format is tab-delimited: prob outcome [weight]

Where prob is the score of the example (higher is better). outcome is the true classification (0 negative, 1 positive), weight is an optional weight for the example, defaults to 1.0.

Example:

java -jar auc_orig_points.jar -t list -o test/test test/test.list

About

Adaptation of the AUCCalculator by Davis & Goadrich. My very small update prints out PR points for the original data. You should cite Davis & Goadrich if you use this tool.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages