def create_micromodel(self):
pred_index = _to_cpp_array_int(self.prediction_column)
cond_index= _to_cpp_array_int(self.conditional_column)
self._saved_model = _IscMarkovGaussMicroModel(pred_index, len(self.prediction_column),
cond_index, len(self.conditional_column))
pyisc._free_array_int(pred_index)
pyisc._free_array_int(cond_index)
return self._saved_model
def create_micromodel(self):
pred_index = _to_cpp_array_int(self.prediction_column)
cond_index = _to_cpp_array_int(self.conditional_column)
self._saved_model = _IscMarkovGaussMicroModel(
pred_index, len(self.prediction_column), cond_index,
len(self.conditional_column))
pyisc._free_array_int(pred_index)
pyisc._free_array_int(cond_index)
return self._saved_model
def _anomaly_score_intfloat(self, x_intfloat, length, data_object):
deviations = pyisc._double_array(self.num_of_partitions)
min = pyisc._intfloat_array(length)
max = pyisc._intfloat_array(length)
peak = pyisc._intfloat_array(length)
anom = pyisc._double_array(1)
cla = pyisc._int_array(1)
clu = pyisc._int_array(1)
self._anomaly_detector._CalcAnomalyDetails(x_intfloat,anom, cla, clu, deviations, peak, min, max)
if self.is_clustering and self.class_column > -1:
result = [pyisc._get_double_value(anom,0),
pyisc._get_int_value(cla,0),
pyisc._get_int_value(clu,0),
list(pyisc._to_numpy_array(deviations,self.num_of_partitions)),
list(data_object._convert_to_numpyarray(peak, length)),
list(data_object._convert_to_numpyarray(min, length)),
list(data_object._convert_to_numpyarray(max, length))]
elif self.is_clustering:
result = [pyisc._get_double_value(anom,0),
pyisc._get_int_value(clu,0),
list(pyisc._to_numpy_array(deviations,self.num_of_partitions)),
list(data_object._convert_to_numpyarray(peak, length)),
list(data_object._convert_to_numpyarray(min, length)),
list(data_object._convert_to_numpyarray(max, length))]
elif self.class_column > -1:
result = [pyisc._get_double_value(anom,0),
pyisc._get_int_value(cla,0),
list(pyisc._to_numpy_array(deviations,self.num_of_partitions)),
list(data_object._convert_to_numpyarray(peak, length)),
list(data_object._convert_to_numpyarray(min, length)),
list(data_object._convert_to_numpyarray(max, length))]
else:
result = [pyisc._get_double_value(anom,0),
list(pyisc._to_numpy_array(deviations,self.num_of_partitions)),
list(data_object._convert_to_numpyarray(peak, length)),
list(data_object._convert_to_numpyarray(min, length)),
list(data_object._convert_to_numpyarray(max, length))]
pyisc._free_array_double(deviations);
pyisc._free_array_intfloat(min)
pyisc._free_array_intfloat(max)
pyisc._free_array_intfloat(peak)
pyisc._free_array_double(anom)
pyisc._free_array_int(cla)
pyisc._free_array_int(clu)
return result
def create_micromodel(self):
column_array = _to_cpp_array_int(self.column_index)
self._saved_model = _IscMultiGaussianMicroModel(len(self.column_index), column_array)
pyisc._free_array_int(column_array)
return self._saved_model
def create_micromodel(self):
value_array = _to_cpp_array_int(self.value_columns)
self._saved_model = _IscMarkovGaussMatrixMicroModel(value_array, len(self.value_columns), self.slots_per_row)
pyisc._free_array_int(value_array)
return self._saved_model
def create_micromodel(self):
column_array = _to_cpp_array_int(self.column_index)
self._saved_model = _IscMultiGaussianMicroModel(
len(self.column_index), column_array)
pyisc._free_array_int(column_array)
return self._saved_model
def create_micromodel(self):
value_array = _to_cpp_array_int(self.value_columns)
self._saved_model = _IscMarkovGaussMatrixMicroModel(
value_array, len(self.value_columns), self.slots_per_row)
pyisc._free_array_int(value_array)
return self._saved_model