def loadFileKMeans(file,classNameIndex):
k = KMeans(constants.getN())
fileHelper = FileHelper()
c1 = KMeansClass(0,"Iris-setosa")
c1.setVCenter([4.6,3.0,4.0,0.0])
c2 = KMeansClass(1,"Iris-versicolor")
c2.setVCenter([6.8,3.4,4.6,0.7])
k.addClass(c1)
k.addClass(c2)
try:
f = fileHelper.openReadOnlyFile(file)
lineas = f.readlines()
uMatrix = constants.getKMeansInitializeUMAtrix(len(lineas))
k.setUMatrix(uMatrix)
xVector = []
for linea in lineas:
xVector = linea.strip("\r\n").split(",")
del xVector[classNameIndex-1]
xVector = [float(x) for x in xVector]
k.addXVector(xVector)
return k
except:
print("Error al leer el fichero")
def k_means(data, sample_to_guess, log_file, tolerance=0.01, b=2):
def getKMeansInitializeUMAtrix(xVectorsCount):
random.seed()
matrix = []
for i in range(xVectorsCount):
elem = random.uniform(0, 1)
matrix.append(elem)
matrix.append(1 - elem)
return np.reshape(matrix, (xVectorsCount, 2))
k = KMeans(4)
c1 = KMeansClass(0, "Iris-setosa")
c1.setVCenter([4.6, 3.0, 4.0, 0.0])
c2 = KMeansClass(1, "Iris-versicolor")
c2.setVCenter([6.8, 3.4, 4.6, 0.7])
k.addClass(c1)
k.addClass(c2)
uMatrix = getKMeansInitializeUMAtrix(
len(data['Iris-setosa']) + len(data['Iris-versicolor']))
k.setUMatrix(uMatrix)
for x in data['Iris-setosa']:
k.addXVector(x)
for x2 in data['Iris-versicolor']:
k.addXVector(x2)
k.doTraining(log_file=log_file, epsilonLimit=tolerance, b=b)
prediction = k.clasifyEuclideanDistance(sample_to_guess, log_file)
log_file.write("La prediccion para la muestra es {}".format(prediction))
print("La predicción para la muestra es {}".format(prediction))
return prediction
