Menu
▾
▴
[Jboost-cvs] jboost/scripts VisualizeScores.README, 1.1, 1.2 VisualizeScores.DEMO.README, 1.1, 1.2
|
From: william b. <wil...@us...> - 2009-01-25 08:52:44
|
Update of /cvsroot/jboost/jboost/scripts In directory fdv4jf1.ch3.sourceforge.com:/tmp/cvs-serv25502/scripts Modified Files: VisualizeScores.README VisualizeScores.DEMO.README Log Message: updated to reflect new output directory structure from nfold.py Index: VisualizeScores.DEMO.README =================================================================== RCS file: /cvsroot/jboost/jboost/scripts/VisualizeScores.DEMO.README,v retrieving revision 1.1 retrieving revision 1.2 diff -C2 -d -r1.1 -r1.2 *** VisualizeScores.DEMO.README 25 Jan 2009 08:19:49 -0000 1.1 --- VisualizeScores.DEMO.README 25 Jan 2009 08:52:38 -0000 1.2 *************** *** 5,11 **** 0. you have java, jython, and python installed. ! 1. you have downloaded the jboost dist from sourceforge.net and built jboost from source. ! ! You CANNOT use the prebuilt jboost.jar (version 1.4 or earlier) as we are using a yet-unreleased version of jboost. 2. You setup your CLASSPATH and JBOOST_DIR environment variables per the install instructions on http://jboost.sourceforge.net/install.html --- 5,9 ---- 0. you have java, jython, and python installed. ! 1. you have downloaded the jboost dist from sourceforge.net and built jboost from source or you downloaded the pre-release. (ie. you have jboost 1.4.1 or greater) 2. You setup your CLASSPATH and JBOOST_DIR environment variables per the install instructions on http://jboost.sourceforge.net/install.html *************** *** 58,62 **** Booster type: jboost.booster.LogLossBoost Read 100 training examples ! Read 200 training examples Read 300 training examples Read 400 training examples --- 56,60 ---- Booster type: jboost.booster.LogLossBoost Read 100 training examples ! Read 200 training examples Read 300 training examples Read 400 training examples *************** *** 66,80 **** [snip] you get the idea. ! 5. the results are placed in the directory ./cvdata-mm-dd-hh-mm-ss/<TREE-TYPE> 6. Now run the visualizer in the scripts. Assuming jython is not on your path, be explicit: ! > ~/jython2.2.1/jython ../scripts/VisualizeScores.py cvdata-mm-dd-hh-mm-ss/ADD_ALL/trial (note, this example output shows the cvdata-mm-dd-hh-mm-ss for my test run) ! ['cvdata-01-16-17-44-23/ADD_ALL/trial0.test.boosting.info', 'cvdata-01-16-17-44-23/ADD_ALL/trial1.test.boosting.info'] ! cvdata-01-16-17-44-23/ADD_ALL/trial0.test.boosting.info ! cvdata-01-16-17-44-23/ADD_ALL/trial1.test.boosting.info index=0, a.length=4600 index=1, a.length=4600 --- 64,78 ---- [snip] you get the idea. ! 5. the results are placed in the directory ./spambase_idx.data.folds_2/cvdata-mm-dd-hh-mm-ss/<TREE-TYPE> 6. Now run the visualizer in the scripts. Assuming jython is not on your path, be explicit: ! > ~/jython2.2.1/jython ../scripts/VisualizeScores.py spambase_idx.data.folds_2/cvdata-mm-dd-hh-mm-ss/ADD_ALL/trial (note, this example output shows the cvdata-mm-dd-hh-mm-ss for my test run) ! ['spambase_idx.data.folds_2/cvdata-01-16-17-44-23/ADD_ALL/trial0.test.boosting.info', 'spambase_idx.data.folds_2/cvdata-01-16-17-44-23/ADD_ALL/trial1.test.boosting.info'] ! spambase_idx.data.folds_2/cvdata-01-16-17-44-23/ADD_ALL/trial0.test.boosting.info ! spambase_idx.data.folds_2/cvdata-01-16-17-44-23/ADD_ALL/trial1.test.boosting.info index=0, a.length=4600 index=1, a.length=4600 Index: VisualizeScores.README =================================================================== RCS file: /cvsroot/jboost/jboost/scripts/VisualizeScores.README,v retrieving revision 1.1 retrieving revision 1.2 diff -C2 -d -r1.1 -r1.2 *** VisualizeScores.README 17 Jan 2009 02:00:31 -0000 1.1 --- VisualizeScores.README 25 Jan 2009 08:52:38 -0000 1.2 *************** *** 6,10 **** ============ * Jython ! * build of jboost from current cvs src tree (NOT THE PREBUILT JBOOST.JAR VERSION 1.4) * test | train.boosting.info files which are indexed (see AddRandomIndex.py) * binary class labels (i.e. no multilabel support at this time) --- 6,10 ---- ============ * Jython ! * build of jboost 1.4.1 (NOT JBOOST VERSION 1.4 found on sourceforge) * test | train.boosting.info files which are indexed (see AddRandomIndex.py) * binary class labels (i.e. no multilabel support at this time) *************** *** 25,29 **** Example ======= ! (this example is run from within Jython dist): [user@host jython2.2.1]$ export JBOOST_DIR=/Users/jsmith/workspace/jboost/ (export for a bash shell. ignore if already done) --- 25,29 ---- Example ======= ! (this example is run from within Jython dist and assumes ): [user@host jython2.2.1]$ export JBOOST_DIR=/Users/jsmith/workspace/jboost/ (export for a bash shell. ignore if already done) *************** *** 33,37 **** (NOTE: ASSUMES you have already added jboost.jar and concurrent.jar to your classpath) ! [user@host jython2.2.1]$ ./jython $JBOOST_DIR/scripts/VisualizeScores.py /Users/jsmith/boostingdata/cvdata/ADD_ALL/trial After loading, parsing, and analyzing the boosting.info files, the GUI will launch. Boosting iteration shown in upper right. A histogram showing example count by boosting score for each label is predominate and in the center of the window. Sliders below the histogram allow you select lower and upper score ranges. The region within this range on the ROC curve is shown in the upper left of the window. Pressing the button in the lower left (save selected examples) will write to a file the example index and boosting score for each example within the selected range. This file is saved to the directory indicated in step 1 above. --- 33,37 ---- (NOTE: ASSUMES you have already added jboost.jar and concurrent.jar to your classpath) ! [user@host jython2.2.1]$ ./jython $JBOOST_DIR/scripts/VisualizeScores.py /Users/jsmith/boostingdata/<path to specific cvdata>/ADD_ALL/trial After loading, parsing, and analyzing the boosting.info files, the GUI will launch. Boosting iteration shown in upper right. A histogram showing example count by boosting score for each label is predominate and in the center of the window. Sliders below the histogram allow you select lower and upper score ranges. The region within this range on the ROC curve is shown in the upper left of the window. Pressing the button in the lower left (save selected examples) will write to a file the example index and boosting score for each example within the selected range. This file is saved to the directory indicated in step 1 above. |
