def get_instance(self, instance_path):
name = ""
dimension = ""
veiculos = ""
capacidade = ""
demanda = []
data = []
with open(instance_path, 'r', encoding='utf-8') as file:
name = file.readline().split(' ')[-1]
dimension = int(file.readline().split(' ')[-1])
veiculos = int(file.readline().split(' ')[-1])
capacidade = int(file.readline().split(' ')[-1])
demanda = []
inicio_da_secao_Edge_Weight = 5
file.readline() #DEMAND_SECTION:
for i in range(dimension):
stringAuxiliar = file.readline()
stringAuxiliar2 = (
re.findall('\d+', stringAuxiliar)
) # vai criar uma lista apenas com os valores (removendo os espaços)
demanda.append(
int(stringAuxiliar2[1])
) #vai receber apenas o segundo valor na parte de demanda.
inicio_da_secao_Edge_Weight += 1
inicio_da_secao_Edge_Weight += 2
file.readline() #linha vazia antes do EDGE WEIGHT SECTION
file_type = file.readline().strip()
if file_type == 'EDGE_WEIGHT_SECTION':
data2 = [[0 for x in range(dimension)]
for y in range(dimension)
] #matriz [Dimensao][Dimensao]
fop = open(instance_path, "r")
arquivo = fop.readlines()
for i in range(dimension):
stringAuxiliar = arquivo[inicio_da_secao_Edge_Weight]
stringAuxiliar2 = (re.findall('\d+', stringAuxiliar))
for j in range(dimension):
data2[i][j] = int(stringAuxiliar2[j])
inicio_da_secao_Edge_Weight += 1
data = self.__get_instance(file, dimension)
return Instance(name, dimension, veiculos, capacidade, demanda, data,
data2)
def get_instance(self, instance_path):
instance = None
name = ""
dimension = ""
data = []
with open(instance_path, 'r', encoding='utf-8') as file:
name = file.readline().split(' ')[-1]
dimension = int(file.readline().split(' ')[-1])
data = []
file_type = file.readline().strip()
if file_type == 'EDGE_WEIGHT_SECTION':
data = self.__get_instance_teste(file, dimension)
elif file_type == 'DISPLAY_DATA_SECTION':
data = self.__get_instance_tsp_cup(file, dimension)
return Instance(name, dimension, data)
def leerFichero(self, nombre_fichero): f = open(nombre_fichero,'r') #instancias al estilo WEKA instances = Instances() primeraLinea = f.readline() cadenasLinea = re.split(self.delimiters, primeraLinea) numeroEntradas = int(cadenasLinea[0]) numeroClases = int(cadenasLinea[1]) for i in range(0, numeroEntradas): instances.addColumna(str(i), "REAL") for i in range(0, numeroClases): instances.addClase(str(i)) for line in iter(lambda: f.readline(), ''): tokens = self.privateLimpiaVacioTokens(re.split(self.delimiters, line)) #print tokens if len(tokens) <= 0: break #instancia al estilo WEKA instance = Instance() #se anyaden las entradas del perceptron for i in range(0, numeroEntradas): instance.addElement(float(tokens[i])) #transformacion de 1 0 a 0 por ejemplo y 0 1 a 1 #con la finalidad de no usar un array de clases que no tiene sentido en clasificacion #puede tener sentido en un red neuronal, no lo niego j = 0 for i in range(numeroEntradas, numeroEntradas + numeroClases): if tokens[i] == '1': instance.addElement(str(j)) break j += 1 instances.addInstance(instance) f.close() return instances
def loadDataFromFile(fileName,
skipRowsWithInvalidFeature=False,
defaultValueOfInvalidFeature=1):
instances = []
file = open(fileName, 'r').read()
lines = file.split('\n')
for line in lines:
if '?' in line and skipRowsWithInvalidFeature is True:
continue
row = line.split(',')
if row.__len__() == 11:
instance = Instance(row[columnId],
row[columnFirstFeature:quantityOfFeatures + 1],
row[columnCancerClass],
defaultValueOfInvalidFeature)
instances.append(instance)
return instances
