class IMachineLearning(model.Schema):
"""
Behavior interface to make a Dexterity type support Machine Learning
"""
model.fieldset('machinelearning',
label=_(u"Learning"),
fields=[
'machineLearningCluster',
])
machineLearningCluster = schema.Int(
title=_(u'label_machine_learning_cluster', default=u'Related cluster'),
required=False,
)
class ISettingsClustering(Interface):
nClusters = schema.Int(
title=_(u"Number of clusters"),
description=_(u"Maximum number of clusters"),
)
clusteringFile = schema.TextLine(
title=_(u"Cluster model file name"),
description=_(
u"The filename that will be stored into directory storage"),
)
clustersTerms = schema.Dict(
title=_(u"Top terms per cluster"),
description=_(u"Learned clusters and top terms for each. "
u"Needs Vectorizer remembering."),
required=False,
readonly=True,
key_type=schema.Int(required=False),
value_type=schema.List(
value_type=schema.TextLine(required=False),
required=False,
),
)
class ISettingsVectorizer(Interface):
# useTfidf = schema.Bool(
# title=_(u"Tf-IDF"),
# description=_(u"Use Tf-IDF in the vectorizer"),
# )
useNltk = schema.Bool(
title=_(u"NLTK stem"),
description=_(u"Use NLTK stem in the vectorizer"),
)
stopWords = schema.TextLine(
title=_(u"Stop Words"),
description=_(u"Words that are too frequent to be considered. "
u"Example: a the for. "
u"Defaults: english (stop words for english)."),
)
ngramRangeMin = schema.Int(
title=_(u"Minimum N-gram range of words"),
description=_(u"Group words together for the vectorizer. "
u"Minimum group size. "
u"Defaults to 1: only one word."),
)
ngramRangeMax = schema.Int(
title=_(u"Maximum N-gram range of words"),
description=_(u"Group words together for the vectorizer. "
u"Maximum group size."
u"Defaults to 1: only one word."),
)
useRemembering = schema.Bool(
title=_(u"Remember string<->hashes mapping"),
description=_(u"Human vocabulary. Slower and more memory consuming."),
)
class SettingsPanelForm(RegistryEditForm):
"""
Define form logic
"""
form.extends(RegistryEditForm)
schema = ISettings
fields = field.Fields(ISettingsDefault)
groups = (FormVectorizer, FormClustering)
@button.buttonAndHandler(_(u"Compute"), name='compute')
def handleCompute(self, action):
model = ILearningModel(self.context)
# pr = cProfile.Profile()
# pr.enable()
self.status = model.compute()
class FormClustering(group.Group):
label = _(u"Cluster model")
fields = field.Fields(ISettingsClustering)
class SettingsPanelForm(RegistryEditForm):
"""
Define form logic
"""
form.extends(RegistryEditForm)
schema = ISettings
fields = field.Fields(ISettingsDefault)
groups = (FormVectorizer, FormClustering)
@button.buttonAndHandler(_(u"Compute"), name='compute')
def handleCompute(self, action):
model = ILearningModel(self.context)
# pr = cProfile.Profile()
# pr.enable()
self.status = model.compute()
# pr.create_stats()
# pr.print_stats(sort='cumtime')
SettingsPanelView = layout.wrap_form(
SettingsPanelForm, ControlPanelFormWrapper)
SettingsPanelView.label = _(u"Machine Learning settings")
class FormClustering(group.Group):
label = _(u"Cluster model")
fields = field.Fields(ISettingsClustering)
class FormVectorizer(group.Group):
label = _(u"Vectorizer")
fields = field.Fields(ISettingsVectorizer)
class FormClustering(group.Group):
label = _(u"Cluster model")
fields = field.Fields(ISettingsClustering)
class SettingsPanelForm(RegistryEditForm):
"""
Define form logic
"""
form.extends(RegistryEditForm)
schema = ISettings
fields = field.Fields(ISettingsDefault)
groups = (FormVectorizer, FormClustering)
@button.buttonAndHandler(_(u"Compute"), name='compute')
def handleCompute(self, action):
model = ILearningModel(self.context)
# pr = cProfile.Profile()
# pr.enable()
self.status = model.compute()
# pr.create_stats()
# pr.print_stats(sort='cumtime')
SettingsPanelView = layout.wrap_form(SettingsPanelForm,
ControlPanelFormWrapper)
SettingsPanelView.label = _(u"Machine Learning settings")