fix remaining examples

shogun-toolbox · Aug 31, 2011 · b17aeae · b17aeae
1 parent ffb7c54
commit b17aeae
Show file tree

Hide file tree

Showing 3 changed files with 45 additions and 44 deletions.
diff --git a/examples/undocumented/python_modular/modelselection_grid_search_linear_modular.py b/examples/undocumented/python_modular/modelselection_grid_search_linear_modular.py
@@ -69,7 +69,7 @@ def modelselection_grid_search_linear_modular(traindat=traindat, label_traindat=
 
 	# print best parameters
 	#print "best parameters:"
-	best_parameters.print_tree()
+	#best_parameters.print_tree()
 
 	# apply them and print result
 	best_parameters.apply_to_machine(classifier)
@@ -78,4 +78,4 @@ def modelselection_grid_search_linear_modular(traindat=traindat, label_traindat=
 
 if __name__=='__main__':
 	print 'ModelSelection GridSearchLinear'
-	modelselection_grid_search_simple(*parameter_list[0])
+	modelselection_grid_search_linear_modular(*parameter_list[0])
diff --git a/examples/undocumented/python_modular/streaming_vw_createcache_modular.py b/examples/undocumented/python_modular/streaming_vw_createcache_modular.py
@@ -4,46 +4,41 @@
 from modshogun import StreamingVwFeatures
 from modshogun import VowpalWabbit
 
-def create_cache():
-    """Creates a binary cache from an ascii data file."""
-
-    # Open the input file as a StreamingVwFile
-    input_file = StreamingVwFile("../data/fm_train_sparsereal.dat")
-    # Default file name will be vw_cache.dat.cache
-    input_file.set_write_to_cache(True)
-
-    # Tell VW that the file is in SVMLight format
-    # Supported types are T_DENSE, T_SVMLIGHT and T_VW
-    input_file.set_parser_type(T_SVMLIGHT)
-
-    # Create a StreamingVwFeatures object, `True' indicating the examples are labelled
-    features = StreamingVwFeatures(input_file, True, 1024)
-
-    # Create a VW object from the features
-    vw = VowpalWabbit(features)
-    vw.set_no_training(True)
-
-    # Train (in this case does nothing but run over all examples)
-    vw.train()
-
-def train_from_cache():
-    """Train using the generated cache file."""
-
-    # Open the input cache file as a StreamingVwCacheFile
-    input_file = StreamingVwCacheFile("vw_cache.dat.cache");
-
-    # The rest is exactly as for normal input
-    features = StreamingVwFeatures(input_file, True, 1024);
-    vw = VowpalWabbit(features)
-    vw.train()
+parameter_list=[['../data/fm_train_sparsereal.dat']]
+
+def streaming_vw_createcache_modular(fname):
+	# First creates a binary cache from an ascii data file.
+	# and then trains using the StreamingVwCacheFile as input
+
+	# Open the input file as a StreamingVwFile
+	input_file = StreamingVwFile(fname)
+	# Default file name will be vw_cache.dat.cache
+	input_file.set_write_to_cache(True)
+
+	# Tell VW that the file is in SVMLight format
+	# Supported types are T_DENSE, T_SVMLIGHT and T_VW
+	input_file.set_parser_type(T_SVMLIGHT)
+
+	# Create a StreamingVwFeatures object, `True' indicating the examples are labelled
+	features = StreamingVwFeatures(input_file, True, 1024)
+
+	# Create a VW object from the features
+	vw = VowpalWabbit(features)
+	vw.set_no_training(True)
+
+	# Train (in this case does nothing but run over all examples)
+	vw.train()
+
+	#Finally Train using the generated cache file
+
+	# Open the input cache file as a StreamingVwCacheFile
+	input_file = StreamingVwCacheFile("vw_cache.dat.cache");
+
+	# The rest is exactly as for normal input
+	features = StreamingVwFeatures(input_file, True, 1024);
+	vw = VowpalWabbit(features)
+	vw.train()
+	return vw
 
 if __name__ == "__main__":
-    print "Creating cache..."
-    create_cache()
-    print "Done."
-    print
-
-    print "Training using the cache file..."
-    print
-    train_from_cache()
-
+	streaming_vw_createcache_modular(*parameter_list[0])
diff --git a/src/shogun/classifier/vw/VowpalWabbit.cpp b/src/shogun/classifier/vw/VowpalWabbit.cpp
@@ -219,7 +219,7 @@ void CVowpalWabbit::init(CStreamingVwFeatures* feat)
 	SG_REF(env);
 	SG_REF(reg);
 
-	quiet = false;
+	quiet = true;
 	no_training = false;
 	dump_interval = exp(1.);
 	sum_loss_since_last_dump = 0.;
@@ -373,6 +373,12 @@ void CVowpalWabbit::output_prediction(int32_t f, float32_t res, float32_t weight
 	}
 }
 
+void CVowpalWabbit::set_verbose(bool verbose)
+{
+	quiet=verbose==false;
+}
+
+
 float32_t CVowpalWabbit::compute_exact_norm(VwExample* &ex, float32_t& sum_abs_x)
 {
 	// We must traverse the features in _precisely_ the same order as during training.