df['target'] = target
return sns.pairplot(df, hue='target', diag_kind='hist')
if __name__ == '__main__':
dataset = datasets.load_iris()
data, target = dataset.data, dataset.target
lvqnet = algorithms.LVQ3(
# number of features
n_inputs=4,
# number of data points that we want
# to have at the end
n_subclasses=30,
# number of classes
n_classes=3,
verbose=True,
show_epoch=20,
step=0.001,
n_updates_to_stepdrop=150 * 100,
)
lvqnet.train(data, target, epochs=100)
plot_scattermatrix(data, target)
plot_scattermatrix(data=lvqnet.weight, target=lvqnet.subclass_to_class)
plt.show()
lvqnet = algorithms.LVQ2(n_inputs=30,n_classes=8)
lvqnet.train(X,Y, epochs=100)
Y_pred = lvqnet.predict(X)
acc = accuracy_score(Y,Y_pred)
'''
for train_index, test_index in kf.split(X):
# Set up the training and testing sets
X_train, X_test = X[train_index], X[test_index]
Y_train, Y_test = Y[train_index], Y[test_index]
print(X_train.shape)
print(Y_train.shape)
# Train the algorithm using training set
lvqnet = algorithms.LVQ3(n_inputs=col - 1, n_classes=8)
lvqnet.train(X_train, Y_train, epochs=500)
# Test the algororithm using testing set
Y_pred = lvqnet.predict(X_test)
acc[count] = accuracy_score(Y_test, Y_pred)
count = count + 1
# Initialize the codebook vectors to be vectors that are randomly selected
# from the training data we have.
# codebooks = []
# randindex = np.random.randint(0, len(X[0])-1)
# codebooks.append(X[randindex,:])
# for x in range (1,8):
# randindex = np.random.randint(0, len(X[0])-1)
def hello_world():
if request.method == 'POST':
import numpy as np
from neupy import algorithms
#TAHAP PRE-PROCESSING DATA
fiturs_dict = request.get_json()
list_train = []
list_testing = []
list_nama = []
list_id_ktp = []
list_no_gambar = []
i = 0
# PEMBAGIAN DATA TRAIN DAN TEST
for fitur in fiturs_dict:
i = i + 1
# print("\n\nnilai i = ",i)
a = len(fiturs_dict)
# a = Panjang data train + test :
if (i == a):
list_testing.append([
fitur['NILAI_GREYSCALE'], fitur['NILAI_INVARIANCE'],
fitur['NILAI_ENTROPY'], fitur['NILAI_SKEWNESS'],
fitur['RELATIVE_SMOOTHNESS'], fitur['NILAI_ENERGY'],
fitur['NILAI_CONTRAST']
])
else:
list_train.append([
fitur['NILAI_GREYSCALE'], fitur['NILAI_INVARIANCE'],
fitur['NILAI_ENTROPY'], fitur['NILAI_SKEWNESS'],
fitur['RELATIVE_SMOOTHNESS'], fitur['NILAI_ENERGY'],
fitur['NILAI_CONTRAST']
])
list_nama.append([fitur['NAMA']])
list_id_ktp.append([fitur['KTP']])
list_no_gambar.append([fitur['NO_GAMBAR']])
print("\n\nISI data list TRAIN (20 awal) \n", list_train[:20])
print("\n\nISI data list TESTING \n", list_testing)
print("\n\nISI data list id ktp \n", list_id_ktp)
print("\n\nISI data list no gambar : \n", list_no_gambar)
(unique, counts) = np.unique(list_id_ktp, return_counts=True)
(unique) = np.unique(list_id_ktp, return_counts=False)
freq_id_ktp_unik = np.asarray((unique, counts)).T
arr_id_ktp_unik = np.asarray((unique)).T
print("\nISI Unique + count id_ktp", freq_id_ktp_unik)
print("\nISI Unique id_ktp", arr_id_ktp_unik)
print("Banyaknya ID KTP unik", len(freq_id_ktp_unik))
(unique, counts) = np.unique(list_nama, return_counts=True)
(unique) = np.unique(list_nama, return_counts=False)
freq_nama = np.asarray((unique, counts)).T
nama_unik = np.asarray((unique)).T
print("\nISI Unique + count NAMA = \n", freq_nama)
print("\nISI Unique NAMA \n", nama_unik)
print("\nBanyaknya NAMA \n", len(freq_nama))
array_train = np.array(list_train)
print("\n\nIsi array train (20 awal) : ", array_train[:20])
array_id_ktp = np.array(list_id_ktp)
list_id_ktp_unik = arr_id_ktp_unik.tolist()
list_id_ktp = array_id_ktp.tolist()
print(f'\n\nList \n: {list_id_ktp_unik}')
print(f'\n\nList id ktp \n: {list_id_ktp}')
i = 0
persamaan_kelas_awal = []
for a in arr_id_ktp_unik:
str_id_ktp_unik = ''.join(a)
print("\nPrint str id unik ", str_id_ktp_unik)
persamaan_kelas_awal.append(str(i))
#print("\n\ninisialisasi normalisasi : \n", persamaan_kelas_awal)
i = i + 1
a = list_id_ktp
list_label_id_ktp = []
for n, i in enumerate(a):
str1 = ''.join(a[n])
list_label_id_ktp.append(str1)
print("LIST STRING ID KTP SEMUA : \n", list_label_id_ktp)
list_str_no_gambar = []
for n, i in enumerate(list_no_gambar):
str1 = ''.join(list_no_gambar[n])
list_str_no_gambar.append(str1)
print("LIST STRING NO GAMBAR : \n", list_str_no_gambar)
results = list(map(int, list_label_id_ktp))
print("\n\nlist int ID ktp :", results)
results2 = list(map(str, list_nama))
print("\n\nlist str nama:", results2)
list_label_nama_pemilik = []
for n, i in enumerate(results):
list_label_id_ktp[n] = i - 1
list_label_nama_pemilik.append(list_nama[n])
print("\n\nLIST Pelabelan NAMA : ", list_label_nama_pemilik)
print("\n\nLIST STRING kelas di jadikan urut mulai dari nol : \n",
list_label_id_ktp)
list_int_target = list(map(int, list_label_id_ktp))
array_target = np.array(list_int_target)
list_int_testing = []
for n, i in enumerate(list_testing):
list_int = list(map(int, i))
list_int_testing.append(list_int)
print("\n\nLIST INT testing: \n", list_int_testing)
array_testing = np.array(list_testing)
scaler = MinMaxScaler(feature_range=(0, 1))
array_minmax_testing = np.array(array_testing)
arr_minmax_testing = scaler.fit_transform(array_minmax_testing)
print("\n\n Minmaxscaler Testing isi : \n", arr_minmax_testing)
array_minmax_train = np.vstack(list_train)
arr_minmax_train = scaler.fit_transform(array_minmax_train)
print("\nn\ Minmaxscaler Train isi (20 awal) : \n",
arr_minmax_train[:20])
list_testing_minmax = arr_minmax_testing.tolist()
print("\n\nIsi array train : \n", array_train[:20])
print("\n\nIsi list test : \n", list_testing_minmax)
print("\n\nISI data list TESTING \n", list_testing)
print("\n\nIsi array test : \n", array_testing)
print("\nIsi array target : \n", array_target)
#TAHAP PROCESSING LVQ
lvqnet = algorithms.LVQ3(n_inputs=len(fiturs_dict),
n_classes=len(list_id_ktp_unik),
verbose=True,
step=0.005,
epsilon=0.05)
lvqnet.train(array_train, array_target, epochs=100)
output_kelas = lvqnet.predict(list_testing)
for i, a in enumerate(list_label_id_ktp):
if (output_kelas == list_label_id_ktp[i]):
hasil_nama = list_label_nama_pemilik[i]
str_nama = ''.join(hasil_nama)
str_output_kelas = np.array_str(output_kelas)
str_output_kelas = str_output_kelas.replace("[", "").replace("]", "")
print("\n\nMasuk ke kelas : ", str_output_kelas)
kelas_hasil = str_output_kelas, str_nama
print(kelas_hasil)
print(
"==========================================================================================="
)
return str(kelas_hasil)
else:
return "bukan POST"