Skip to content

Commit

Permalink
added remaining case in unittest + code style changes
Browse files Browse the repository at this point in the history
  • Loading branch information
@mazumdarparijat
mazumdarparijat committed Mar 11, 2014
1 parent 87ca78d commit 41d3666
Show file tree
Hide file tree
Showing 2 changed files with 44 additions and 40 deletions.
74 changes: 36 additions & 38 deletions src/shogun/multiclass/tree/ID3ClassifierTree.cpp
View file
Expand Up @@ -47,9 +47,9 @@ CID3ClassifierTree::~CID3ClassifierTree()

CMulticlassLabels* CID3ClassifierTree::apply_multiclass(CFeatures* data)
{
REQUIRE(data, "Data required for classification in apply_multiclass")
REQUIRE(data, "Data required for classification in apply_multiclass\n")

CDenseFeatures<float64_t>* feats = (CDenseFeatures<float64_t>*) data;
CDenseFeatures<float64_t>* feats = dynamic_cast<CDenseFeatures<float64_t>*>(data);
int32_t num_vecs = feats->get_num_vectors();
SGVector<float64_t> labels = SGVector<float64_t>(num_vecs);

Expand All @@ -64,7 +64,7 @@ CMulticlassLabels* CID3ClassifierTree::apply_multiclass(CFeatures* data)
int32_t flag = 0;
for (int32_t j=0; j<children->get_num_elements(); j++)
{
node_t* child = (node_t*) children->get_element(j);
node_t* child = dynamic_cast<node_t*>(children->get_element(j));
if (child->data.transit_if_feature_value
== sample[node->data.attribute_id])
{
Expand Down Expand Up @@ -98,14 +98,14 @@ CMulticlassLabels* CID3ClassifierTree::apply_multiclass(CFeatures* data)

bool CID3ClassifierTree::train_machine(CFeatures* data)
{
REQUIRE(data,"Data required for training")
REQUIRE(data->get_feature_class()==C_DENSE, "Dense data required for training")
REQUIRE(data,"Data required for training\n")
REQUIRE(data->get_feature_class()==C_DENSE, "Dense data required for training\n")

int32_t num_features = ((CDenseFeatures<float64_t>*) data)->get_num_features();
int32_t num_features = (dynamic_cast<CDenseFeatures<float64_t>*>(data))->get_num_features();
SGVector<int32_t> feature_ids = SGVector<int32_t>(num_features);
feature_ids.range_fill();

set_root(id3train(data, (CMulticlassLabels*) m_labels, feature_ids, 0));
set_root(id3train(data, dynamic_cast<CMulticlassLabels*>(m_labels), feature_ids, 0));

return true;
}
Expand All @@ -114,18 +114,18 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
CMulticlassLabels* class_labels, SGVector<int32_t> feature_id_vector, int32_t level)
{
node_t* node = new node_t();
CDenseFeatures<float64_t>* feats = (CDenseFeatures<float64_t>*) data;
CDenseFeatures<float64_t>* feats = dynamic_cast<CDenseFeatures<float64_t>*>(data);
int32_t num_vecs = feats->get_num_vectors();

// if all samples belong to the same class
if(class_labels->get_unique_labels().size() == 1)
if (class_labels->get_unique_labels().size() == 1)
{
node->data.class_label = class_labels->get_unique_labels()[0];
return node;
}

// if no feature is left
if(feature_id_vector.vlen == 0)
if (feature_id_vector.vlen == 0)
{
// decide label - label occuring max times
SGVector<float64_t> labels = class_labels->get_labels();
Expand All @@ -134,9 +134,8 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
int32_t most_label = labels[0];
int32_t most_num = 1;
int32_t count = 1;
int32_t i = 1;

while (i<labels.vlen)
for (int32_t i=1; i<labels.vlen; i++)
{
while ((labels[i] == labels[i-1]) && (i<labels.vlen))
{
Expand All @@ -151,7 +150,6 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
}

count = 1;
i++;
}

node->data.class_label = most_label;
Expand All @@ -161,11 +159,11 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
// else get the feature with the highest informational gain
float64_t max = 0;
int32_t best_feature_index = -1;
for(int32_t i=0; i<feats->get_num_features(); i++)
for (int32_t i=0; i<feats->get_num_features(); i++)
{
float64_t gain = informational_gain_attribute(i,feats,class_labels);

if(gain > max)
if (gain >= max)
{
max = gain;
best_feature_index = i;
Expand All @@ -174,21 +172,21 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,

//get feature values for the best feature chosen
SGVector<float64_t> best_feature_values = SGVector<float64_t>(num_vecs);
for(int32_t i=0; i<num_vecs; i++)
for (int32_t i=0; i<num_vecs; i++)
best_feature_values[i] = (feats->get_feature_vector(i))[best_feature_index];

CMulticlassLabels* best_feature_labels = new CMulticlassLabels(best_feature_values);
SGVector<float64_t> best_labels_unique = best_feature_labels->get_unique_labels();

for(int32_t i=0; i<best_labels_unique.vlen; i++)
for (int32_t i=0; i<best_labels_unique.vlen; i++)
{
//compute the number of vectors with active attribute value
int32_t num_cols = 0;
float64_t active_feature_value = best_labels_unique[i];

for(int32_t j=0; j<num_vecs; j++)
for (int32_t j=0; j<num_vecs; j++)
{
if( active_feature_value == best_feature_values[j])
if ( active_feature_value == best_feature_values[j])
num_cols++;
}

Expand All @@ -197,15 +195,15 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,

int32_t cnt = 0;
//choose the samples that have the active feature value
for(int32_t j=0; j<num_vecs; j++)
for (int32_t j=0; j<num_vecs; j++)
{
SGVector<float64_t> sample = feats->get_feature_vector(j);
if(active_feature_value == sample[best_feature_index])
if (active_feature_value == sample[best_feature_index])
{
int32_t idx = -1;
for(int32_t k=0; k<sample.size(); k++)
for (int32_t k=0; k<sample.size(); k++)
{
if(k != best_feature_index)
if (k != best_feature_index)
mat(++idx, cnt) = sample[k];
}

Expand All @@ -217,9 +215,9 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
//remove the best_attribute from the remaining attributes index vector
SGVector<int32_t> new_feature_id_vector = SGVector<int32_t>(feature_id_vector.vlen-1);
cnt = -1;
for(int32_t j=0;j<feature_id_vector.vlen;j++)
for (int32_t j=0;j<feature_id_vector.vlen;j++)
{
if(j!=best_feature_index)
if (j!=best_feature_index)
new_feature_id_vector[++cnt] = feature_id_vector[j];
}

Expand All @@ -243,40 +241,40 @@ CTreeMachineNode<id3TreeNodeData>* CID3ClassifierTree::id3train(CFeatures* data,
float64_t CID3ClassifierTree::informational_gain_attribute(int32_t attr_no, CFeatures* data,
CMulticlassLabels* class_labels)
{
REQUIRE(data,"Data required for information gain calculation")
REQUIRE(data,"Data required for information gain calculation\n")
REQUIRE(data->get_feature_class()==C_DENSE,
"Dense data required for information gain calculation")
"Dense data required for information gain calculation\n")

float64_t gain = 0;
CDenseFeatures<float64_t>* feats = (CDenseFeatures<float64_t>*) data;
CDenseFeatures<float64_t>* feats = dynamic_cast<CDenseFeatures<float64_t>*>(data);
int32_t num_vecs = feats->get_num_vectors();

//get attribute values for attribute
SGVector<float64_t> attribute_values = SGVector<float64_t>(num_vecs);

for(int32_t i=0; i<num_vecs; i++)
for (int32_t i=0; i<num_vecs; i++)
attribute_values[i] = (feats->get_feature_vector(i))[attr_no];

CMulticlassLabels* attribute_labels = new CMulticlassLabels(attribute_values);
SGVector<float64_t> attr_val_unique = attribute_labels->get_unique_labels();

for(int32_t i=0; i<attr_val_unique.vlen; i++)
for (int32_t i=0; i<attr_val_unique.vlen; i++)
{
//calculate class entropy for the specific attribute_value
int32_t attr_count=0;

for(int32_t j=0; j<num_vecs; j++)
for (int32_t j=0; j<num_vecs; j++)
{
if(attribute_values[j] == attr_val_unique[i])
if (attribute_values[j] == attr_val_unique[i])
attr_count++;
}

SGVector<float64_t> sub_class = SGVector<float64_t>(attr_count);
int32_t count = 0;

for(int32_t j=0; j<num_vecs; j++)
for (int32_t j=0; j<num_vecs; j++)
{
if(attribute_values[j] == attr_val_unique[i])
if (attribute_values[j] == attr_val_unique[i])
sub_class[count++] = class_labels->get_label(j);
}

Expand All @@ -300,19 +298,19 @@ float64_t CID3ClassifierTree::entropy(CMulticlassLabels* labels)
SGVector<float64_t> log_ratios = SGVector<float64_t>
(labels->get_unique_labels().size());

for(int32_t i=0;i<labels->get_unique_labels().size();i++)
for (int32_t i=0;i<labels->get_unique_labels().size();i++)
{
int32_t count = 0;

for(int32_t j=0;j<labels->get_num_labels();j++)
for (int32_t j=0;j<labels->get_num_labels();j++)
{
if(labels->get_unique_labels()[i] == labels->get_label(j))
if (labels->get_unique_labels()[i] == labels->get_label(j))
count++;
}

log_ratios[i] = (count-0.f)/(labels->get_num_labels()-0.f);

if(log_ratios[i] != 0)
if (log_ratios[i] != 0)
log_ratios[i] = CMath::log(log_ratios[i]);
}

Expand Down
10 changes: 8 additions & 2 deletions tests/unit/multiclass/tree/ID3ClassifierTree_unittest.cc
View file
Expand Up @@ -54,7 +54,7 @@ TEST(ID3ClassifierTree, classify_simple)
{
/* Example from http://www.cise.ufl.edu/~ddd/cap6635/Fall-97/Short-papers/2.htm */

SGMatrix<float64_t> data(4,14);
SGMatrix<float64_t> data(4,15);

//vector = [Outlook Temperature Humidity Wind]
data(0,0)=sunny;
Expand Down Expand Up @@ -127,10 +127,15 @@ TEST(ID3ClassifierTree, classify_simple)
data(2,13)=high;
data(3,13)=strong;

data(0,14)=overcast;
data(1,14)=mild;
data(2,14)=high;
data(3,14)=strong;

CDenseFeatures<float64_t>* feats=new CDenseFeatures<float64_t>(data);

// yes 1. no 0.
SGVector<float64_t> lab(14);
SGVector<float64_t> lab(15);
lab[0]=0.0;
lab[1]=0.0;
lab[2]=1.0;
Expand All @@ -145,6 +150,7 @@ TEST(ID3ClassifierTree, classify_simple)
lab[11]=1.0;
lab[12]=1.0;
lab[13]=0.0;
lab[14]=0.0;

CMulticlassLabels* labels=new CMulticlassLabels(lab);

Expand Down

0 comments on commit 41d3666

Please sign in to comment.