ok

demo/mushroom/README

@@ -1,2 +1,12 @@
-example of UCI dataset
+example of training a binary classifier on UCI dataset
 http://archive.ics.uci.edu/ml/datasets/Mushroom
+
+Run: ./runexp.sh
+
+Format of input: LIBSVM format
+
+Format of featmap.txt:
+<featureid> <featurename> <q or i>\n
+
+q means continuous quantities, i means indicator features.
+Feature id must be from 0 to num_features, in sorted order.

demo/mushroom/dump2json.py

@@ -46,13 +46,13 @@ def dumpjson( fo, trees ):
     fo.write('\n}\n')
         
 fo = sys.stdout
-nmap = loadnmap( 'featname.txt' )
+nmap = loadnmap( 'featmap.txt' )
 stat = loadstats( 'agaricus.txt.test', 'dump.path.txt' )
 
 trees = {'roots':[], 'weights':[], 'nodes':[] }
 idmap = {}
 
-for l in open( 'dump.txt'):
+for l in open( 'dump.raw.txt'):
     if l.startswith('booster['):
         bid = int( l.split('[')[1].split(']')[0] )
         trees['roots'].append( mapid(idmap,bid,0) )

demo/mushroom/runexp.sh

@@ -1,10 +1,12 @@
 #!/bin/bash
+# map feature using indicator encoding, also produce featmap.txt
 python mapfeat.py
+# split train and test
 python mknfold.py agaricus.txt 1
+# training
 ../../xgboost mushroom.conf
 # this is what dump will looklike without feature map
 ../../xgboost mushroom.conf task=dump model_in=0003.model name_dump=dump.raw.txt 
 # this is what dump will looklike with feature map
 ../../xgboost mushroom.conf task=dump model_in=0003.model fmap=featmap.txt name_dump=dump.nice.txt
 cat dump.nice.txt
-