def __importModelFromFile(self, file):
"""Return new model instance and set configuration and parameters from file."""
nemoa.log('import model from file')
nemoa.setLog(indent='+1')
# check file
if not os.path.exists(file):
if os.path.exists(nemoa.workspace.path('models') + file + '.mp'):
file = nemoa.workspace.path('models') + file + '.mp'
else:
return nemoa.log(
'error', """
could not load model '%s':
file does not exist.""" % file)
# load model parameters and configuration from file
nemoa.log('load model: \'%s\'' % file)
modelDict = nemoa.common.dictFromFile(file)
model = self.__getModelInstance(name=modelDict['config']['name'],
config=modelDict['config'],
dataset=modelDict['dataset']['cfg'],
network=modelDict['network']['cfg'],
system=modelDict['system']['config'])
if nemoa.type.isModel(model): model._set(modelDict)
else: return None
nemoa.setLog(indent='-1')
return model
def loadProject(self, project):
"""Import configuration files from user project."""
nemoa.log('import private resources')
nemoa.setLog(indent='+1')
# check if project exists
if not project in self.__listUserWorkspaces():
nemoa.log(
'warning', """
could not open project '%s':
project folder could not be found in '%s'!
""" % (project, self.__basepath['user']))
return False
# set project
self.__project = project
# update paths
self.__updatePaths(base='user')
# update path of cache
self.__updateCachePaths()
# update logger to current logfile
nemoa.initLogger(logfile=self.__path['logfile'])
# import object configurations for current project
self.__scanForConfigFiles()
self.__scanForScripts()
self.__scanForNetworks()
nemoa.setLog(indent='-1')
return True
def __importConfigFile(self, file):
"""Import configuration (.ini) file."""
# search definition file
if os.path.isfile(file):
configFile = file
elif os.path.isfile(self.__basepath['workspace'] + file):
configFile = self.__basepath['workspace'] + file
else:
nemoa.log('warning', "configuration file '%s' does not exist!" % (file))
return False
# logger info
nemoa.log("parsing configuration file: '" + configFile + "'")
nemoa.setLog(indent = '+1')
# import and register objects without testing
importer = configFileImporter(self)
objConfList = importer.load(configFile)
for objConf in objConfList:
self.__addObjToStore(objConf)
self.__check(objConfList)
nemoa.setLog(indent = '-1')
return True
def configure(self):
"""Configure model."""
nemoa.log('configure model \'%s\'' % (self.name()))
nemoa.setLog(indent = '+1')
if not 'check' in self.__config:
self.__config['check'] = {'dataset': False,
'network': False, 'System': False}
self.__config['check']['dataset'] = \
self.dataset.configure(network = self.network)
if not self.__config['check']['dataset']:
nemoa.log('error', """could not configure model: dataset
could not be configured!""")
nemoa.setLog(indent = '-1')
return False
self.__config['check']['network'] = \
self.network.configure(dataset = self.dataset,
system = self.system)
if not self.__config['check']['network']:
nemoa.log('error', """could not configure model: network
could not be configured!""")
nemoa.setLog(indent = '-1')
return False
self.__config['check']['system'] = \
self.system.configure(network = self.network,
dataset = self.dataset)
if not self.__config['check']['system']:
nemoa.log('error', """could not configure model: system
could not be configured!""")
nemoa.setLog(indent = '-1')
return False
nemoa.setLog(indent = '-1')
return True
def loadProject(self, project):
"""Import configuration files from user project."""
nemoa.log('import private resources')
nemoa.setLog(indent = '+1')
# check if project exists
if not project in self.__listUserWorkspaces():
nemoa.log('warning', """
could not open project '%s':
project folder could not be found in '%s'!
""" % (project, self.__basepath['user']))
return False
# set project
self.__project = project
# update paths
self.__updatePaths(base = 'user')
# update path of cache
self.__updateCachePaths()
# update logger to current logfile
nemoa.initLogger(logfile = self.__path['logfile'])
# import object configurations for current project
self.__scanForConfigFiles()
self.__scanForScripts()
self.__scanForNetworks()
nemoa.setLog(indent = '-1')
return True
def __importConfigFile(self, file):
"""Import configuration (.ini) file."""
# search definition file
if os.path.isfile(file):
configFile = file
elif os.path.isfile(self.__basepath['workspace'] + file):
configFile = self.__basepath['workspace'] + file
else:
nemoa.log('warning',
"configuration file '%s' does not exist!" % (file))
return False
# logger info
nemoa.log("parsing configuration file: '" + configFile + "'")
nemoa.setLog(indent='+1')
# import and register objects without testing
importer = configFileImporter(self)
objConfList = importer.load(configFile)
for objConf in objConfList:
self.__addObjToStore(objConf)
self.__check(objConfList)
nemoa.setLog(indent='-1')
return True
def __importModelFromFile(self, file):
"""Return new model instance and set configuration and parameters from file."""
nemoa.log('import model from file')
nemoa.setLog(indent = '+1')
# check file
if not os.path.exists(file):
if os.path.exists(
nemoa.workspace.path('models') + file + '.mp'):
file = nemoa.workspace.path('models') + file + '.mp'
else: return nemoa.log('error', """
could not load model '%s':
file does not exist.""" % file)
# load model parameters and configuration from file
nemoa.log('load model: \'%s\'' % file)
modelDict = nemoa.common.dictFromFile(file)
model = self.__getModelInstance(
name = modelDict['config']['name'],
config = modelDict['config'],
dataset = modelDict['dataset']['cfg'],
network = modelDict['network']['cfg'],
system = modelDict['system']['config'])
if nemoa.type.isModel(model): model._set(modelDict)
else: return None
nemoa.setLog(indent = '-1')
return model
def scripts(project):
"""Print list of scripts to standard output."""
logParams = nemoa.getLog()
nemoa.setLog(mode='silent')
workspace = nemoa.open(project)
nemoa.setLog(mode=logParams['mode'])
scripts = workspace.list(type='script', namespace=workspace.project())
print 'Scripts in project %s:\n' % (project)
for script in scripts:
print ' %s' % (script)
print ''
def scripts(project):
"""Print list of scripts to standard output."""
logParams = nemoa.getLog()
nemoa.setLog(mode="silent")
workspace = nemoa.open(project)
nemoa.setLog(mode=logParams["mode"])
scripts = workspace.list(type="script", namespace=workspace.project())
print "Scripts in project %s:\n" % (project)
for script in scripts:
print " %s" % (script)
print ""
def __getModelInstance(self,
name=None,
config=None,
dataset=None,
network=None,
system=None):
"""Return new model instance."""
nemoa.log('create model instance')
nemoa.setLog(indent='+1')
# create dataset instance (if not given)
if not nemoa.type.isDataset(dataset): dataset = \
self.__getInstance(type = 'dataset', config = dataset)
if not nemoa.type.isDataset(dataset):
nemoa.log('error',
'could not create model instance: dataset is invalid!')
nemoa.setLog(indent='-1')
return None
# create network instance (if not given)
if network == None: network = {'type': 'auto'}
if not nemoa.type.isNetwork(network): network = \
self.__getInstance(type = 'network', config = network)
if not nemoa.type.isNetwork(network):
nemoa.log('error',
'could not create model instance: network is invalid!')
nemoa.setLog(indent='-1')
return None
# create system instance (if not given)
if not nemoa.type.isSystem(system): system = \
self.__getInstance(type = 'system', config = system)
if not nemoa.type.isSystem(system):
nemoa.log('error',
'could not create model instance: system is invalid!')
nemoa.setLog(indent='-1')
return None
# create name string (if not given)
if name == None:
name = '-'.join([dataset.name(), network.name(), system.name()])
# create model instance
model = self.__getInstance(type='model',
config=config,
name=name,
dataset=dataset,
network=network,
system=system)
nemoa.setLog(indent='-1')
return model
def __getInstance(self, type = None, config = None, empty = False, **kwargs):
"""Return new instance of given object type and configuration."""
nemoa.log('create%s %s instance' % (' empty' if empty else '', type))
nemoa.setLog(indent = '+1')
# import module
module = importlib.import_module('nemoa.' + str(type))
# get objects configuration as dictionary
config = nemoa.workspace.getConfig(type = type,
config = config, **kwargs)
if not isinstance(config, dict):
nemoa.log('error', """could not create %s instance:
unknown configuration!""" % (type))
nemoa.setLog(indent = '-1')
return None
# create and initialize new instance of given class
instance = module.empty() if empty \
else module.new(config = config, **kwargs)
# check instance class
if not nemoa.type.isInstanceType(instance, type):
nemoa.log('error', """could not create %s instance:
invalid configuration!""" % (type))
nemoa.setLog(indent = '-1')
return None
nemoa.log('name of %s is: \'%s\'' % (type, instance.name()))
nemoa.setLog(indent = '-1')
return instance
def configure(self, dataset=None, network=None, *args, **kwargs):
"""Configure system and subsystems to network and dataset."""
if not hasattr(self.__class__, "_configure") or not callable(getattr(self.__class__, "_configure")):
return True
nemoa.log("configure system '%s'" % (self.name()))
nemoa.setLog(indent="+1")
if not self.checkNetwork(network):
nemoa.log(
"error",
"""
system could not be configured:
network is not valid!""",
)
nemoa.setLog(indent="-1")
return False
if not self.checkDataset(dataset):
nemoa.log(
"error",
"""
system could not be configured:
dataset is not valid!""",
)
nemoa.setLog(indent="-1")
return False
retVal = self._configure(dataset=dataset, network=network, *args, **kwargs)
nemoa.setLog(indent="-1")
return retVal
def __getInstance(self, type=None, config=None, empty=False, **kwargs):
"""Return new instance of given object type and configuration."""
nemoa.log('create%s %s instance' % (' empty' if empty else '', type))
nemoa.setLog(indent='+1')
# import module
module = importlib.import_module('nemoa.' + str(type))
# get objects configuration as dictionary
config = nemoa.workspace.getConfig(type=type, config=config, **kwargs)
if not isinstance(config, dict):
nemoa.log(
'error', """could not create %s instance:
unknown configuration!""" % (type))
nemoa.setLog(indent='-1')
return None
# create and initialize new instance of given class
instance = module.empty() if empty \
else module.new(config = config, **kwargs)
# check instance class
if not nemoa.type.isInstanceType(instance, type):
nemoa.log(
'error', """could not create %s instance:
invalid configuration!""" % (type))
nemoa.setLog(indent='-1')
return None
nemoa.log('name of %s is: \'%s\'' % (type, instance.name()))
nemoa.setLog(indent='-1')
return instance
def __scanForNetworks(self, files=None):
"""Scan for scripts files in current project."""
nemoa.log('scanning for networks')
nemoa.setLog(indent='+1')
# are files given?
if files == None:
files = self.__path['networks'] + '*.ini'
# import definition files
for file in glob.iglob(self.getPath(files)):
self.__registerNetwork(file)
nemoa.setLog(indent='-1')
return True
def __scanForNetworks(self, files = None):
"""Scan for scripts files in current project."""
nemoa.log('scanning for networks')
nemoa.setLog(indent = '+1')
# are files given?
if files == None:
files = self.__path['networks'] + '*.ini'
# import definition files
for file in glob.iglob(self.getPath(files)):
self.__registerNetwork(file)
nemoa.setLog(indent = '-1')
return True
def __scanForConfigFiles(self, files = None):
"""Import all config files from current project."""
nemoa.log('scanning for configuration files')
nemoa.setLog(indent = '+1')
# are files given?
if files == None:
files = self.__path['workspace'] + '*.ini'
# import configuration files
for file in glob.iglob(self.getPath(files)):
self.__importConfigFile(file)
nemoa.setLog(indent = '-1')
return True
def __getModelInstance(self, name = None, config = None,
dataset = None, network = None, system = None):
"""Return new model instance."""
nemoa.log('create model instance')
nemoa.setLog(indent = '+1')
# create dataset instance (if not given)
if not nemoa.type.isDataset(dataset): dataset = \
self.__getInstance(type = 'dataset', config = dataset)
if not nemoa.type.isDataset(dataset):
nemoa.log('error',
'could not create model instance: dataset is invalid!')
nemoa.setLog(indent = '-1')
return None
# create network instance (if not given)
if network == None: network = {'type': 'auto'}
if not nemoa.type.isNetwork(network): network = \
self.__getInstance(type = 'network', config = network)
if not nemoa.type.isNetwork(network):
nemoa.log('error',
'could not create model instance: network is invalid!')
nemoa.setLog(indent = '-1')
return None
# create system instance (if not given)
if not nemoa.type.isSystem(system): system = \
self.__getInstance(type = 'system', config = system)
if not nemoa.type.isSystem(system):
nemoa.log('error',
'could not create model instance: system is invalid!')
nemoa.setLog(indent = '-1')
return None
# create name string (if not given)
if name == None: name = '-'.join(
[dataset.name(), network.name(), system.name()])
# create model instance
model = self.__getInstance(
type = 'model', config = config, name = name,
dataset = dataset, network = network, system = system)
nemoa.setLog(indent = '-1')
return model
def __scanForConfigFiles(self, files=None):
"""Import all config files from current project."""
nemoa.log('scanning for configuration files')
nemoa.setLog(indent='+1')
# are files given?
if files == None:
files = self.__path['workspace'] + '*.ini'
# import configuration files
for file in glob.iglob(self.getPath(files)):
self.__importConfigFile(file)
nemoa.setLog(indent='-1')
return True
def optimizeParams(self, dataset, schedule):
"""Optimize system parameters using data and given schedule."""
# check if optimization schedule exists for current system
# and merge default, existing and given schedule
if not "params" in schedule:
config = self._default("optimize")
nemoa.common.dictMerge(self._config["optimize"], config)
self._config["optimize"] = config
elif not self.getType() in schedule["params"]:
return nemoa.log(
"error",
"""could not optimize model:
optimization schedule '%s' does not include system '%s'
"""
% (schedule["name"], self.getType()),
)
else:
config = self._default("optimize")
nemoa.common.dictMerge(self._config["optimize"], config)
nemoa.common.dictMerge(schedule["params"][self.getType()], config)
self._config["optimize"] = config
################################################################
# System independent optimization settings #
################################################################
# check dataset
if (not "checkDataset" in config or config["checkDataset"] == True) and not self._checkDataset(dataset):
return False
# initialize inspector
inspector = nemoa.system.base.inspector(self)
if "inspect" in config and not config["inspect"] == False:
inspector.setTestData(self._getTestData(dataset))
# optimize system parameters
algorithm = config["algorithm"].title()
nemoa.log("note", "optimize '%s' (%s) using algorithm '%s'" % (self.name(), self.getType(), algorithm))
nemoa.setLog(indent="+1")
retVal = self._optimizeParams(dataset, schedule, inspector)
nemoa.setLog(indent="-1")
return retVal
def plot(self, plot = None, output = 'file', file = None, **kwargs):
"""Create plot of model."""
nemoa.log('create plot of model')
nemoa.setLog(indent = '+1')
# check if model is configured
if not self.__isConfigured():
nemoa.log('error', """could not create plot of model:
model is not yet configured!""")
nemoa.setLog(indent = '-1')
return False
# get plot instance
nemoa.log('create plot instance')
nemoa.setLog(indent = '+1')
if plot == None: plot = self.system.getType() + '.default'
if isinstance(plot, str):
plotName, plotParams = nemoa.common.strSplitParams(plot)
mergeDict = plotParams
for param in kwargs.keys(): plotParams[param] = kwargs[param]
objPlot = self.__getPlot(name = plotName, params = plotParams)
if not objPlot:
nemoa.log('warning', "could not create plot: unknown configuration '%s'" % (plotName))
nemoa.setLog(indent = '-1')
return None
elif isinstance(plot, dict): objPlot = self.__getPlot(config = plot)
else: objPlot = self.__getPlot()
if not objPlot: return None
# prepare filename
if output == 'display': file = None
elif output == 'file' and not file:
file = nemoa.common.getEmptyFile(nemoa.workspace.path('plots') + \
self.name() + '/' + objPlot.cfg['name'] + \
'.' + objPlot.settings['fileformat'])
# create plot
retVal = objPlot.create(self, file = file)
if not file == None: nemoa.log('save plot: ' + file)
nemoa.setLog(indent = '-2')
return retVal
def preprocessData(self, **kwargs):
"""Data preprocessing.
Keyword Arguments:
stratify -- see algorithm in stratifyData
normalize -- see algorithm in normalizeData
transform -- see algorithm in transformData
Description:
Process stratification, normalization and transformation."""
nemoa.log('preprocessing data')
nemoa.setLog(indent = '+1')
if 'stratify' in kwargs.keys(): self.stratifyData(kwargs['stratify'])
if 'normalize' in kwargs.keys(): self.normalizeData(kwargs['normalize'])
if 'transform' in kwargs.keys(): self.transformData(kwargs['transform'])
nemoa.setLog(indent = '-1')
return True
def configure(self, dataset, system):
"""Configure network to dataset and system."""
# check if network instance is empty
if self.isEmpty():
nemoa.log('configuration is not needed: network is \'empty\'')
return True
# check if dataset instance is available
if not nemoa.type.isDataset(dataset):
nemoa.log('error', 'could not configure network: no valid dataset instance given!')
return False
# check if system instance is available
if not nemoa.type.isSystem(system):
nemoa.log('error', 'could not configure network: no valid system instance given!')
return False
nemoa.log('configure network: \'%s\'' % (self.name()))
nemoa.setLog(indent = '+1')
# type: 'auto is used for networks
# wich are created by datasets (visible units)
# and systems (hidden units)
if self.cfg['type'] == 'auto':
self.__getVisibleNodesFromDataset(dataset)
self.__getHiddenNodesFromSystem(system)
self.__getEdgesFromNodesAndLayers()
self.updateGraph()
nemoa.setLog(indent = '-1')
return True
# type: 'autolayer' is used for networks
# wich are created by layers and sizes
if self.cfg['type'] == 'autolayer':
self.__getNodesFromLayers()
self.__getEdgesFromNodesAndLayers()
self.updateGraph()
nemoa.setLog(indent = '-1')
return True
# configure network to dataset
groups = dataset.getColGroups()
changes = []
for group in groups:
if not group in self.cfg['nodes'] \
or not (groups[group] == self.cfg['nodes'][group]):
self.updateGraph(nodelist = {'type': group, 'list': groups[group]})
nemoa.setLog(indent = '-1')
return True
def optimizeParams(self, dataset, schedule):
"""Optimize system parameters using data and given schedule."""
# check if optimization schedule exists for current system
# and merge default, existing and given schedule
if not 'params' in schedule:
config = self._default('optimize')
nemoa.common.dictMerge(self._config['optimize'], config)
self._config['optimize'] = config
elif not self.getType() in schedule['params']: return nemoa.log(
'error', """could not optimize model:
optimization schedule '%s' does not include system '%s'
""" % (schedule['name'], self.getType()))
else:
config = self._default('optimize')
nemoa.common.dictMerge(self._config['optimize'], config)
nemoa.common.dictMerge(schedule['params'][self.getType()], config)
self._config['optimize'] = config
################################################################
# System independent optimization settings #
################################################################
# check dataset
if (not 'checkDataset' in config or config['checkDataset'] == True) \
and not self._checkDataset(dataset): return False
# initialize inspector
inspector = nemoa.system.base.inspector(self)
if 'inspect' in config and not config['inspect'] == False:
inspector.setTestData(self._getTestData(dataset))
# optimize system parameters
algorithm = config['algorithm'].title()
nemoa.log('note', "optimize '%s' (%s) using algorithm '%s'" % \
(self.name(), self.getType(), algorithm))
nemoa.setLog(indent = '+1')
retVal = self._optimizeParams(dataset, schedule, inspector)
nemoa.setLog(indent = '-1')
return retVal
def save(self, file = None):
"""Save model settings to file and return filepath."""
nemoa.log('save model to file')
nemoa.setLog(indent = '+1')
# get filename
if file == None:
fileName = '%s.mp' % (self.name())
filePath = nemoa.workspace.path('models')
file = filePath + fileName
file = nemoa.common.getEmptyFile(file)
# save model parameters and configuration to file
nemoa.common.dictToFile(self._get(), file)
# create console message
nemoa.log("save model as: '%s'" % (os.path.basename(file)[:-3]))
nemoa.setLog(indent = '-1')
return file
def export(self, file, **kwargs):
"""Export data to file."""
file = nemoa.common.getEmptyFile(file)
type = nemoa.common.getFileExt(file).lower()
nemoa.log('export data to file')
nemoa.setLog(indent = '+1')
nemoa.log('exporting data to file: \'%s\'' % (file))
if type in ['gz', 'data']: retVal = self.save(file)
elif type in ['csv', 'tsv', 'tab', 'txt']:
cols, data = self.getData(output = ('cols', 'recarray'))
retVal = nemoa.common.csvSaveData(file, data,
cols = [''] + cols, **kwargs)
else: retVal = nemoa.log('error', """
could not export dataset:
unsupported file type '%s'""" % (type))
nemoa.setLog(indent = '-1')
return retVal
def __createNewModel(self,
name,
config=None,
dataset=None,
network=None,
system=None,
configure=True,
initialize=True):
nemoa.log('create new model')
nemoa.setLog(indent='+1')
model = self.__getModelInstance(name=name,
config=config,
dataset=dataset,
network=network,
system=system)
if not nemoa.type.isModel(model):
nemoa.log('error', 'could not create new model!')
nemoa.setLog(indent='-1')
return False
# configure model
if configure: model.configure()
# initialize model parameters
if initialize: model.initialize()
nemoa.setLog(indent='-1')
return model
def loadCommon(self):
"""Import common projects."""
nemoa.log('import shared resources')
nemoa.setLog(indent = '+1')
# get current project
curProject = self.__project
#
for project in self.__listCommonProjects():
# set common project, update paths and import workspaces
self.__project = project
self.__updatePaths(base = 'common')
self.__scanForConfigFiles()
self.__scanForScripts()
self.__scanForNetworks()
# reset to previous project
self.__project = curProject
nemoa.setLog(indent = '-1')
return True
def loadCommon(self):
"""Import common projects."""
nemoa.log('import shared resources')
nemoa.setLog(indent='+1')
# get current project
curProject = self.__project
#
for project in self.__listCommonProjects():
# set common project, update paths and import workspaces
self.__project = project
self.__updatePaths(base='common')
self.__scanForConfigFiles()
self.__scanForScripts()
self.__scanForNetworks()
# reset to previous project
self.__project = curProject
nemoa.setLog(indent='-1')
return True
def configure(self, dataset = None, network = None, *args, **kwargs):
"""Configure system and subsystems to network and dataset."""
if not hasattr(self.__class__, '_configure') \
or not callable(getattr(self.__class__, '_configure')):
return True
nemoa.log("configure system '%s'" % (self.name()))
nemoa.setLog(indent = '+1')
if not self.checkNetwork(network):
nemoa.log('error', """
system could not be configured:
network is not valid!""")
nemoa.setLog(indent = '-1')
return False
if not self.checkDataset(dataset):
nemoa.log('error', """
system could not be configured:
dataset is not valid!""")
nemoa.setLog(indent = '-1')
return False
retVal = self._configure(dataset = dataset, network = network,
*args, **kwargs)
nemoa.setLog(indent = '-1')
return retVal
def __createNewModel(self, name, config = None,
dataset = None, network = None, system = None,
configure = True, initialize = True):
nemoa.log('create new model')
nemoa.setLog(indent = '+1')
model = self.__getModelInstance(name = name, config = config,
dataset = dataset, network = network, system = system)
if not nemoa.type.isModel(model):
nemoa.log('error', 'could not create new model!')
nemoa.setLog(indent = '-1')
return False
# configure model
if configure: model.configure()
# initialize model parameters
if initialize: model.initialize()
nemoa.setLog(indent = '-1')
return model
def main(argv):
os.system('cls' if os.name == 'nt' else 'clear')
nemoa.setLog(mode='shell')
nemoa.welcome()
def main(argv):
os.system('cls' if os.name=='nt' else 'clear')
nemoa.setLog(mode = 'shell')
nemoa.welcome()
def optimize(self, schedule = None, **kwargs):
"""Optimize system parameters."""
nemoa.log('optimize model')
nemoa.setLog(indent = '+1')
# check if model is empty
if self.isEmpty():
nemoa.log('warning', "empty models can not be optimized!")
nemoa.setLog(indent = '-1')
return self
# check if model is configured
if not self.__isConfigured():
nemoa.log('error',
'could not optimize model: model is not yet configured!')
nemoa.setLog(indent = '-1')
return False
# get optimization schedule
if schedule == None: schedule = self.system.getType() + '.default'
elif not '.' in schedule: schedule = \
self.system.getType() + '.' + schedule
schedule = nemoa.workspace.getConfig(
type = 'schedule', config = schedule,
merge = ['params', self.system.getType()],
**kwargs)
if not schedule:
nemoa.log('error', """
could not optimize system parameters:
optimization schedule is not valid!""")
nemoa.setLog(indent = '-1')
return self
# optimization of system parameters
nemoa.log("starting optimization schedule: '%s'" % (schedule['name']))
nemoa.setLog(indent = '+1')
# 2DO: find better solution for multistage optimization
if 'stage' in schedule and len(schedule['stage']) > 0:
for stage, params in enumerate(config['stage']):
self.system.optimizeParams(self.dataset, **params)
elif 'params' in schedule:
self.system.optimizeParams(
dataset = self.dataset, schedule = schedule)
nemoa.setLog(indent = '-1')
# update network
self.system.updateNetwork(self.network)
nemoa.setLog(indent = '-1')
return self
def transformData(self, algorithm = 'system', system = None, mapping = None, **kwargs):
"""Transform dataset.
Keyword Arguments:
algorithm -- name of algorithm used for data transformation
'system':
Transform data using nemoa system instance
'gaussToBinary':
Transform Gauss distributed values to binary values in {0, 1}
'gaussToWeight':
Transform Gauss distributed values to weights in [0, 1]
'gaussToDistance': ??
Transform Gauss distributed values to distances in [0, 1]
system -- nemoa system instance (nemoa object root class 'system')
used for model based transformation of data
mapping -- ..."""
if not isinstance(algorithm, str): return False
# system based data transformation
if algorithm.lower() == 'system':
if not nemoa.type.isSystem(system): return nemoa.log('error',
"""could not transform data using system:
parameter 'system' is invalid!""")
nemoa.log('transform data using system \'%s\'' % (system.name()))
nemoa.setLog(indent = '+1')
if mapping == None: mapping = system.getMapping()
sourceColumns = system.getUnits(group = mapping[0])[0]
targetColumns = system.getUnits(group = mapping[-1])[0]
self.setColLabels(sourceColumns)
for src in self.data:
data = self.data[src]['array']
dataArray = data[sourceColumns].view('<f8').reshape(
data.size, len(sourceColumns))
transArray = system.mapData(dataArray, mapping = mapping, **kwargs)
# create empty record array
numRows = self.data[src]['array']['label'].size
colNames = ('label',) + tuple(targetColumns)
colFormats = ('<U12',) + tuple(['<f8' for x in targetColumns])
newRecArray = numpy.recarray((numRows,),
dtype = zip(colNames, colFormats))
# set values in record array
newRecArray['label'] = self.data[src]['array']['label']
for colID, colName in enumerate(newRecArray.dtype.names[1:]):
# update source data columns
newRecArray[colName] = (transArray[:, colID]).astype(float)
# set record array
self.data[src]['array'] = newRecArray
self.setColLabels(targetColumns)
nemoa.setLog(indent = '-1')
return True
# gauss to binary data transformation
elif algorithm.lower() in ['gausstobinary', 'binary']:
nemoa.log('transform data using \'%s\'' % (algorithm))
for src in self.data:
# update source per column (recarray)
for colName in self.data[src]['array'].dtype.names[1:]:
# update source data columns
self.data[src]['array'][colName] = \
(self.data[src]['array'][colName] > 0.0).astype(float)
return True
# gauss to weight in [0, 1] data transformation
elif algorithm.lower() in ['gausstoweight', 'weight']:
nemoa.log('transform data using \'%s\'' % (algorithm))
for src in self.data:
# update source per column (recarray)
for colName in self.data[src]['array'].dtype.names[1:]:
# update source data columns
self.data[src]['array'][colName] = \
(2.0 / (1.0 + numpy.exp(-1.0 * \
self.data[src]['array'][colName] ** 2))).astype(float)
return True
# gauss to distance data transformation
# ????
elif algorithm.lower() in ['gausstodistance', 'distance']:
nemoa.log('transform data using \'%s\'' % (algorithm))
for src in self.data:
# update source per column (recarray)
for colName in self.data[src]['array'].dtype.names[1:]:
# update source data columns
self.data[src]['array'][colName] = \
(1.0 - (2.0 / (1.0 + numpy.exp(-1.0 * \
self.data[src]['array'][colName] ** 2)))).astype(float)
return True
return nemoa.log('error'
"could not transform data: unknown algorithm '%s'!" % (algorithm))
def configure(self, network, useCache = False, **kwargs):
"""Configure dataset to a given network object
Keyword arguments:
network -- nemoa network object
useCache -- shall data be cached"""
nemoa.log("configure dataset '%s' to network '%s'" % \
(self.name(), network.name()))
nemoa.setLog(indent = '+1')
# load data from cachefile (if caching is used and cachefile exists)
cacheFile = self.searchCacheFile(network) if useCache else None
if cacheFile and self.load(cacheFile):
nemoa.log('load cachefile: \'%s\'' % (cacheFile))
# preprocess data
if 'preprocessing' in self.cfg.keys():
self.preprocessData(**self.cfg['preprocessing'])
nemoa.setLog(indent = '-1')
return True
# create table with one record for every single dataset files
if not 'table' in self.cfg:
conf = self.cfg.copy()
self.cfg['table'] = {}
self.cfg['table'][self.cfg['name']] = conf
self.cfg['table'][self.cfg['name']]['fraction'] = 1.0
################################################################
# Annotation #
################################################################
# get nodes from network and convert to common format
if network.cfg['type'] == 'auto': netGroups = {'v': None}
else:
# get grouped network node labels and label format
netGroups = network.getNodeGroups(type = 'visible')
netGroupsOrder = []
for layer in netGroups:
netGroupsOrder.append((network.layer(layer)['id'], layer))
netGroupsOrder = sorted(netGroupsOrder)
# convert network node labels to common format
nemoa.log('search network nodes in dataset sources')
convNetGroups = {}
convNetGroupsLost = {}
convNetNodes = []
convNetNodesLost = []
netLblFmt = network.cfg['label_format']
for id, group in netGroupsOrder:
convNetGroups[group], convNetGroupsLost[group] = \
nemoa.annotation.convert(netGroups[group], input = netLblFmt)
convNetNodes += convNetGroups[group]
convNetNodesLost += convNetGroupsLost[group]
# notify if any network node labels could not be converted
if convNetNodesLost:
nemoa.log('%s of %s network nodes could not be converted! (see logfile)'
% (len(convNetNodesLost), len(convNetNodes)))
## 2DO get original node labels for log file
nemoa.log('logfile', nemoa.common.strToList(convNetNodesLost))
# get columns from dataset files and convert to common format
colLabels = {}
nemoa.log('configure data sources')
nemoa.setLog(indent = '+1')
for src in self.cfg['table']:
nemoa.log("configure '%s'" % (src))
srcCnf = self.cfg['table'][src]
# get column labels from csv-file
if 'csvtype' in srcCnf['source']:
csvType = srcCnf['source']['csvtype'].strip().lower()
else: csvType = None
origColLabels = nemoa.common.csvGetColLabels(
srcCnf['source']['file'], type = csvType)
if not origColLabels: continue
# set annotation format
format = srcCnf['source']['columns'] \
if 'columns' in srcCnf['source'] else 'generic:string'
# convert column labes
convColLabels, convColLabelsLost = \
nemoa.annotation.convert(origColLabels, input = format)
# notify if any dataset columns could not be converted
if convColLabelsLost:
nemoa.log('warning',
"%i of %i dataset columns could not be converted! (logfile)"
% (len(convColLabelsLost), len(convColLabels)))
nemoa.log("logfile", ", ".join([convColLabels[i] \
for i in convColLabelsLost]))
if not network.cfg['type'] == 'auto':
# search converted network nodes in converted column labels
numLost = 0
numAll = 0
lostNodes = {}
for id, group in netGroupsOrder:
lostNodesConv = \
[val for val in convNetGroups[group] \
if val not in convColLabels]
numAll += len(convNetGroups[group])
if not lostNodesConv: continue
numLost += len(lostNodesConv)
# get original labels
lostNodes[group] = [netGroups[group][
convNetGroups[group].index(val)] for val in lostNodesConv]
# notify if any network nodes could not be found
if numLost:
nemoa.log('warning', """
%i of %i network nodes could not be found in
dataset source! (logfile)""" % (numLost, numAll))
for group in lostNodes:
nemoa.log("logfile", """
missing nodes (group %s): """ % (group)
+ ", ".join(lostNodes[group]))
# prepare dictionary for column source ids
colLabels[src] = {
'conv': convColLabels,
'usecols': (),
'notusecols': convColLabelsLost }
nemoa.setLog(indent = '-1')
# intersect converted dataset column labels
interColLabels = colLabels[colLabels.keys()[0]]['conv']
for src in colLabels:
list = colLabels[src]['conv']
blackList = [list[i] for i in colLabels[src]['notusecols']]
interColLabels = [val for val in interColLabels \
if val in list and not val in blackList]
# if network type is 'auto', set network visible nodes
# to intersected data from database files (without label column)
if network.cfg['type'] == 'auto':
netGroups['v'] = [label for label in interColLabels \
if not label == 'label']
convNetGroups = netGroups
# search network nodes in dataset columns
self.cfg['columns'] = ()
for groupid, group in netGroupsOrder:
found = 0
for id, col in enumerate(convNetGroups[group]):
if not col in interColLabels:
continue
found += 1
# add column (use network label and group)
self.cfg['columns'] += ((group, netGroups[group][id]), )
for src in colLabels: colLabels[src]['usecols'] \
+= (colLabels[src]['conv'].index(col), )
if not found:
nemoa.log('error', """
no node from network group '%s'
could be found in dataset source!
""" % (group))
nemoa.setLog(indent = '-1')
return False
# update source file config
for src in colLabels: self.cfg['table'][src]['source']['usecols'] \
= colLabels[src]['usecols']
################################################################
# Column & Row Filters #
################################################################
# add column filters and partitions from network node groups
self.cfg['colFilter'] = {'*': ['*:*']}
self.cfg['colPartitions'] = {'groups': []}
for group in netGroups:
self.cfg['colFilter'][group] = [group + ':*']
self.cfg['colPartitions']['groups'].append(group)
# add row filters and partitions from sources
self.cfg['rowFilter'] = {'*': ['*:*']}
self.cfg['rowPartitions'] = {'source': []}
for source in self.cfg['table']:
self.cfg['rowFilter'][source] = [source + ':*']
self.cfg['rowPartitions']['source'].append(source)
################################################################
# Import data from CSV-files into numpy arrays #
################################################################
# import data from sources
nemoa.log('import data from sources')
nemoa.setLog(indent = '+1')
self.data = {}
for src in self.cfg['table']:
self.data[src] = {
'fraction': self.cfg['table'][src]['fraction'],
'array': self.csvGetData(src) }
nemoa.setLog(indent = '-1')
# save cachefile
if useCache:
cacheFile = self.createCacheFile(network)
nemoa.log('save cachefile: \'%s\'' % (cacheFile))
self.save(cacheFile)
# preprocess data
if 'preprocessing' in self.cfg.keys():
self.preprocessData(**self.cfg['preprocessing'])
nemoa.setLog(indent = '-1')
return True
