def __init__(self, sModelName, sModelDir, sComment=None,dLearnerConfigArg=None):
if self.bHTR:
cFeatureDefinition = FeatureDefinition_PageXml_StandardOnes
dFeatureConfig = { 'bMultiPage':False, 'bMirrorPage':False
, 'n_tfidf_node':500, 't_ngrams_node':(2,4), 'b_tfidf_node_lc':False
, 'n_tfidf_edge':250, 't_ngrams_edge':(2,4), 'b_tfidf_edge_lc':False }
else:
cFeatureDefinition = FeatureDefinition_PageXml_StandardOnes_noText
dFeatureConfig = { 'bMultiPage':False, 'bMirrorPage':False
, 'n_tfidf_node':None, 't_ngrams_node':None, 'b_tfidf_node_lc':None
, 'n_tfidf_edge':None, 't_ngrams_edge':None, 'b_tfidf_edge_lc':None }
if sComment is None: sComment = sModelName
DU_ECN_Task.__init__(self
, sModelName, sModelDir
, dFeatureConfig=dFeatureConfig
, dLearnerConfig= dLearnerConfigArg if dLearnerConfigArg is not None else self.dLearnerConfig
, sComment=sComment
, cFeatureDefinition=cFeatureDefinition
, cModelClass=DU_Model_GAT
)
if options.bBaseline:
self.bsln_mdl = self.addBaseline_LogisticRegression() # use a LR model trained by GridSearch as baseline
def __init__(self,
sModelName,
sModelDir,
sComment=None,
dLearnerConfigArg=None):
if sComment is None: sComment = sModelName
if dLearnerConfigArg is not None and "ecn_ensemble" in dLearnerConfigArg:
print('ECN_ENSEMBLE')
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig={},
dLearnerConfig=dLearnerConfigArg
if dLearnerConfigArg is not None else self.dLearnerConfig,
sComment=sComment,
cFeatureDefinition=FeatureDefinition_PageXml_NoNodeFeat_v3,
cModelClass=gcn.DU_Model_ECN.DU_Ensemble_ECN)
else:
#Default Case Single Model
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig={},
dLearnerConfig=dLearnerConfigArg
if dLearnerConfigArg is not None else self.dLearnerConfig,
sComment=sComment,
cFeatureDefinition=FeatureDefinition_PageXml_NoNodeFeat_v3)
def __init__(self,
sModelName,
sModelDir,
sComment=None,
dLearnerConfigArg=None):
traceln(self.bHTR)
if self.bHTR:
cFeatureDefinition = FeatureDefinition_PageXml_StandardOnes
dFeatureConfig = {
'bMultiPage': False,
'bMirrorPage': False,
'n_tfidf_node': 300,
't_ngrams_node': (2, 4),
'b_tfidf_node_lc': False,
'n_tfidf_edge': 300,
't_ngrams_edge': (2, 4),
'b_tfidf_edge_lc': False
}
else:
cFeatureDefinition = FeatureDefinition_PageXml_StandardOnes_noText
# cFeatureDefinition = FeatureDefinition_PageXml_NoNodeFeat_v3
dFeatureConfig = {}
if sComment is None: sComment = sModelName
if dLearnerConfigArg is not None and "ecn_ensemble" in dLearnerConfigArg:
traceln('ECN_ENSEMBLE')
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig=dFeatureConfig,
dLearnerConfig=self.dLearnerConfig
if dLearnerConfigArg is None else dLearnerConfigArg,
sComment=sComment,
cFeatureDefinition=cFeatureDefinition,
cModelClass=gcn.DU_Model_ECN.DU_Ensemble_ECN)
else:
#Default Case Single Model
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig=dFeatureConfig,
dLearnerConfig=self.dLearnerConfig
if dLearnerConfigArg is None else dLearnerConfigArg,
sComment=sComment,
cFeatureDefinition=cFeatureDefinition)
def __init__(self,
sModelName,
sModelDir,
sComment=None,
dLearnerConfigArg=None):
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig={},
dLearnerConfig=dLearnerConfigArg
if dLearnerConfigArg is not None else self.dLearnerConfig,
sComment=sComment,
cFeatureDefinition=FeatureDefinition_PageXml_NoNodeFeat_v3)
if options.bBaseline:
self.bsln_mdl = self.addBaseline_LogisticRegression(
) # use a LR model trained by GridSearch as baseline
def __init__(self,
sModelName,
sModelDir,
sComment=None,
dLearnerConfigArg=None):
if sComment is None: sComment = sModelName
DU_ECN_Task.__init__(
self,
sModelName,
sModelDir,
dFeatureConfig={},
dLearnerConfig=dLearnerConfigArg
if dLearnerConfigArg is not None else self.dLearnerConfig,
sComment=sComment,
cFeatureDefinition=
FeatureDefinition_PageXml_StandardOnes_noText,
cModelClass=DU_Model_GAT)
if options.bBaseline:
self.bsln_mdl = self.addBaseline_LogisticRegression(
) # use a LR model trained by GridSearch as baseline
def predict(self, lsColDir):
"""
Return the list of produced files
"""
self.sXmlFilenamePattern = "*.mpxml"
return DU_ECN_Task.predict(self, lsColDir)
