def _handleInput(self, paramInput):
"""
Function to handle the parsed paramInput for this class.
@ In, paramInput, ParameterInput, the already parsed input.
@ Out, None
"""
## By default, we want to name the 'labels' by the name of this
## postprocessor, but that name is not available before processing the XML
## At this point, we have that information
self.initializationOptionDict = {}
for child in paramInput.subparts:
if child.getName() == 'KDD':
if len(child.parameterValues) > 0:
## I'm not sure what this thing is used for, but it seems to make more
## sense to only put data that is not otherwise handled rather than
## put all of the information and then to remove the ones we process.
## - dpm 6/8/16
self.initializationOptionDict[child.getName()] = {}
for key,value in child.parameterValues.items():
if key == 'lib':
self.type = value
elif key == 'labelFeature':
self.labelFeature = value
else:
self.initializationOptionDict[child.getName()][key] = value
else:
self.initializationOptionDict[child.getName()] = utils.tryParse(child.value)
for childChild in child.subparts:
if len(childChild.parameterValues) > 0 and not childChild.getName() == 'PreProcessor':
self.initializationOptionDict[child.getName()][childChild.getName()] = dict(childChild.parameterValues)
else:
self.initializationOptionDict[child.getName()][childChild.getName()] = utils.tryParse(childChild.value)
elif child.getName() == 'pivotParameter':
self.pivotParameter = child.value
if not hasattr(self, 'pivotParameter'):
#TODO, if doing time dependent data mining that needs this, an error
# should be thrown
self.pivotParameter = None
if self.type:
#TODO unSurpervisedEngine needs to be able to handle both methods
# without this if statement.
if self.pivotParameter is not None:
self.unSupervisedEngine = unSupervisedLearning.returnInstance("temporalSciKitLearn", self, **self.initializationOptionDict['KDD'])
else:
self.unSupervisedEngine = unSupervisedLearning.returnInstance(self.type, self, **self.initializationOptionDict['KDD'])
else:
self.raiseAnError(IOError, 'No Data Mining Algorithm is supplied!')
## If the user has not defined a label feature, then we will force it to be
## named by the PostProcessor name followed by:
## the word 'Labels' for clustering/GMM models;
## the word 'Dimension' + a numeric id for dimensionality reduction
## algorithms
if self.labelFeature is None:
if self.unSupervisedEngine.getDataMiningType() in ['cluster','mixture']:
self.labelFeature = self.name+'Labels'
elif self.unSupervisedEngine.getDataMiningType() in ['decomposition','manifold']:
self.labelFeature = self.name+'Dimension'
def _localReadMoreXML(self, xmlNode):
"""
Method that reads the portion of the xml input that belongs to this specialized class
and initializes internal parameters
@ In, xmlNode, xml.etree.Element, Xml element node
@ Out, None
"""
self.distParams = {}
for child in xmlNode:
if child.tag == 'metricType':
self.metricType = child.text
else:
self.distParams[str(child.tag)] = utils.tryParse(child.text)
availableMetrics = pairwise.kernel_metrics().keys(
) + pairwise.distance_metrics().keys() + scores.keys()
if self.metricType not in availableMetrics:
metricList = ', '.join(
availableMetrics[:-1]) + ', or ' + availableMetrics[-1]
self.raiseAnError(
IOError,
'Metric SKL error: metricType ' + str(self.metricType) +
' is not available. Available metrics are: ' + metricList +
'.')
for key, value in self.distParams.items():
try:
newValue = ast.literal_eval(value)
if type(newValue) == list:
newValue = np.asarray(newValue)
self.distParams[key] = newValue
except:
self.distParams[key] = value
def tryStrParse(s):
"""
Trys to parse if it is stringish
@ In, s, string, possible string
@ Out, s, string, original type, or possibly parsed string
"""
return utils.tryParse(s) if type(s).__name__ in ['str','unicode'] else s
def _localInputAndCheckParam(self, inputParam):
"""
Function to read the portion of the xml input
@ In, inputParam, ParameterInput, the xml element node that will be checked against the available options specific to this Sampler
@ Out, initDict, dict, dictionary contains the information about the given xml node
"""
initDict = {}
for child in inputParam.subparts:
if len(child.parameterValues) > 0:
initDict[child.getName()] = dict(child.parameterValues)
else:
initDict[child.getName()] = utils.tryParse(child.value)
return initDict
def _localInputAndCheck(self, xmlNode):
"""
Function to read the portion of the xml input
@ In, xmlNode, xml.etree.ElementTree Element Objects, the xml element node that will be checked against the available options specific to this Sampler
@ Out, initDict, dict, dictionary contains the information about the given xml node
"""
initDict = {}
for child in xmlNode:
if child.attrib:
initDict[child.tag] = dict(child.attrib)
else:
initDict[child.tag] = utils.tryParse(child.text)
return initDict
def _localReadMoreXML(self, xmlNode):
"""
Function that reads the portion of the xml input that belongs to this specialized class
and initializes some elements based on the inputs got
@ In, xmlNode, xml.etree.Element, Xml element node
@ Out, None
"""
# paramInput = DataMining.getInputSpecification()()
# paramInput.parseNode(xmlNode)
## By default, we want to name the 'labels' by the name of this
## postprocessor, but that name is not available before processing the XML
## At this point, we have that information
self.initializationOptionDict = {}
for child in xmlNode:
if child.tag == 'KDD':
if child.attrib:
## I'm not sure what this thing is used for, but it seems to make more
## sense to only put data that is not otherwise handled rather than
## put all of the information and then to remove the ones we process.
## - dpm 6/8/16
self.initializationOptionDict[child.tag] = {}
for key,value in child.attrib.iteritems():
if key == 'lib':
self.type = value
elif key == 'labelFeature':
self.labelFeature = value
else:
self.initializationOptionDict[child.tag][key] = value
else:
self.initializationOptionDict[child.tag] = utils.tryParse(child.text)
for childChild in child:
if childChild.attrib and not childChild.tag == 'PreProcessor':
self.initializationOptionDict[child.tag][childChild.tag] = dict(childChild.attrib)
else:
self.initializationOptionDict[child.tag][childChild.tag] = utils.tryParse(childChild.text)
elif child.tag == 'pivotParameter':
self.pivotParameter = child.text
if not hasattr(self, 'pivotParameter'):
#TODO, if doing time dependent data mining that needs this, an error
# should be thrown
self.pivotParameter = None
if self.type:
#TODO unSurpervisedEngine needs to be able to handle both methods
# without this if statement.
if self.pivotParameter is not None:
self.unSupervisedEngine = unSupervisedLearning.returnInstance("temporalSciKitLearn", self, **self.initializationOptionDict['KDD'])
else:
self.unSupervisedEngine = unSupervisedLearning.returnInstance(self.type, self, **self.initializationOptionDict['KDD'])
else:
self.raiseAnError(IOError, 'No Data Mining Algorithm is supplied!')
## If the user has not defined a label feature, then we will force it to be
## named by the PostProcessor name followed by:
## the word 'Labels' for clustering/GMM models;
## the word 'Dimension' + a numeric id for dimensionality reduction
## algorithms
if self.labelFeature is None:
if self.unSupervisedEngine.getDataMiningType() in ['cluster','mixture']:
self.labelFeature = self.name+'Labels'
elif self.unSupervisedEngine.getDataMiningType() in ['decomposition','manifold']:
self.labelFeature = self.name+'Dimension'
