def __init__(self, record_generator, model_drifter, nodes=1):
self.record_generator = record_generator
self.model_drifter = model_drifter
self.model = model_drifter.construct_init_model()
self.examples_per_round = nodes
self.examples_in_current_round = 0
InputStream.__init__(self, self.getIdentifier(), nodes)
def __init__(self,
identifier,
nodes,
dataset,
label,
targetStock='NASDAQ:GOOG',
repetitions=0,
repetition_interval=7,
delay=False,
prediction_index=0):
self.prediction_index = prediction_index
if isinstance(targetStock, list):
self.m_stTargetStock = targetStock.pop(0)
self.m_aTargetStocks = targetStock
else:
self.m_stTargetStock = targetStock
self.configureFiles(dataset, label)
self.readPrices()
self.readAnnotations()
self.preprocessAvgPrices()
self.m_oCorrel = correlation(self.m_aPrices, [self.m_stTargetStock])
self.calcFeatureNames()
self.parameters = "Financial(%s)" % dataset
self.m_bDelay = False
self.m_iRepetitions = repetitions
self.m_iRepetitionInterval = repetition_interval
self.m_iCurrRepitions = 0
InputStream.__init__(self, identifier, nodes)
def __init__(self,
nodes,
prediction_index=1,
bPreprocess=False,
numberOfStocks=None):
self.m_lFeatures = []
self.m_lLabels = []
self.m_aDates = []
self.m_aStocks = []
self.m_lAvgPrices11 = cl.defaultdict(list)
self.m_lAvgPrices50 = cl.defaultdict(list)
self.m_lAvgPrices200 = cl.defaultdict(list)
self.m_hPrices = cl.defaultdict(list)
self.m_bPreprocess = bPreprocess
self.prediction_index = prediction_index
self.configureFiles()
self.readPrices(numberOfStocks)
if self.m_bPreprocess:
self.preprocessFeatures()
self.m_hPrices.clear() #clear off some memory
self.m_lAvgPrices11.clear()
self.m_lAvgPrices50.clear()
self.m_lAvgPrices200.clear()
else:
self.m_stActStock = self.m_hPrices.keys()[0]
InputStream.__init__(self, self.getIdentifier, nodes)
def __init__(self, dim, drift_prob,num_nodes):
InputStream.__init__(self, "HiddenLayer" + "(" + str(dim) + ")",num_nodes)
self.label_distribution = bernoulli(0.5)
self.dim = dim
self.drift_prob = drift_prob
self.hidden_layer_size = int(log(dim, 2))
self.set_random_parameters()
self.examples_in_current_macro_round=0
def __init__(self, file_name, target_attribute_name, num_nodes=1, positive_target_value=None,
nominal_attributes_default_value=1.0):
InputStream.__init__(self, file_name, num_nodes)
uci_dataset_path = _get_path_to_uci_data_set(file_name)
self.file_name = file_name
self.target_attribute_name = target_attribute_name
self.positive_target_value = positive_target_value
self.nominal_attributes_default_value = nominal_attributes_default_value
self._arff_description = describe(uci_dataset_path)
self.data = self._read_data(uci_dataset_path, self._arff_description)
self.arff_stream = open_stream(uci_dataset_path, self._arff_description)
self._number_of_generated_examples = 0
def __init__(self, file_name, target_attribute_name, num_nodes=1, positive_target_value=None,
nominal_attributes_default_value=1.0, iterations=1):
InputStream.__init__(self, file_name, num_nodes)
uci_dataset_path = _get_path_to_uci_data_set(file_name)
# self.identifier = self._generatate_identifier(file_name)
self.file_name = file_name
self.target_attribute_name = target_attribute_name
self.positive_target_value = positive_target_value
self.nominal_attributes_default_value = nominal_attributes_default_value
self._arff_description = describe(uci_dataset_path)
self.arff_stream = open_stream(uci_dataset_path, self._arff_description)
self._generated_examples = 0
self.iterations = iterations
self.current_iteration = 1
def __init__(self,
nodes,
driftProb=0.028,
prediction_index=1,
numberOfStocks=None,
normalize=False):
self.m_lFeatures = []
self.m_aDates = []
self.m_aStocks = []
self.m_lAvgPrices11 = cl.defaultdict(list)
self.m_lAvgPrices50 = cl.defaultdict(list)
self.m_lAvgPrices200 = cl.defaultdict(list)
self.m_hPrices = cl.defaultdict(list)
self.bNormalize = normalize
self.drift_prob = driftProb
self.prediction_index = prediction_index
self.configureFiles()
self.readPrices(numberOfStocks)
self.preprocessFeatures()
self.m_lAvgPrices11.clear()
self.m_lAvgPrices50.clear()
self.m_lAvgPrices200.clear()
self.m_stActStock = self.m_hPrices.keys()[0]
InputStream.__init__(self, self.getIdentifier, nodes)
def __init__(self, record_generator, model, nodes=1):
self.record_generator = record_generator
self.model = model
InputStream.__init__(self, self.getIdentifier(), nodes)
def __init__(self, data_file='higgs/higgs.svm', nodes = 1, repetitions = 1):
InputStream.__init__(self,"libsvm_reader(data_file = %s)" % data_file, nodes)
self.file_name = os.path.join(config.PATH_TO_DATASETS, data_file)
self.labels = set([1,-1])
self.input_file_handle = open(self.file_name, 'r')
self.current_line = self._fetch_line()