def get_queryset(self, **kwargs):
level = self.kwargs['level']
display_result = 'block'
n_data_normalisasi = normalisasi.get_normalisasi(
level)['n_data_normalisasi']
context = {
'level': level,
'n_data_normalisasi': n_data_normalisasi,
'display_result': display_result,
}
return context
def get_queryset(self, **kwargs):
try:
data = Parameter.objects.get(id='1')
if data is None:
form = NormalisasiForm()
sigma = None
else:
form = NormalisasiForm(initial={'sigma': data.sigma})
sigma = data.sigma
except Parameter.DoesNotExist:
form = NormalisasiForm()
sigma = None
level = self.kwargs['level']
n_data_normalisasi = []
n_list_data_kernel_view = []
if level == 1:
display_form = 'block'
display_result = 'none'
else:
display_form = 'none'
if sigma is not None:
display_result = 'block'
n_data_normalisasi = normalisasi.get_normalisasi(level)['n_data_normalisasi']
n_list_data_kernel_view = kernel.get_kernel(n_data_normalisasi, float(sigma))['n_list_data_kernel_view']
# Save to DB
# normalisasi.save_normalisasi_to_db()
else:
display_result = 'none'
context = {
'level': level,
'n_data_normalisasi': n_data_normalisasi,
'n_list_data_kernel_view': n_list_data_kernel_view,
'display_form': display_form,
'display_result': display_result,
'form': form
}
return context
def post(self, request, *args, **kwargs):
form = NormalisasiForm(request.POST)
level = self.kwargs['level']
if form.is_valid():
sigma = form.cleaned_data['sigma']
try:
param = Parameter.objects.get(id='1')
except Parameter.DoesNotExist:
param = Parameter()
param.id = '1'
param.sigma = sigma
param.save()
n_data_normalisasi = normalisasi.get_normalisasi(level)['n_data_normalisasi']
n_list_data_kernel_view = kernel.get_kernel(n_data_normalisasi, float(sigma))['n_list_data_kernel_view']
# Save to DB
normalisasi.save_normalisasi_to_db()
context = {
'level': level,
'n_data_normalisasi': n_data_normalisasi,
'n_list_data_kernel_view': n_list_data_kernel_view,
'display_form': 'block',
'display_result': 'block',
'form': form
}
return render(request, self.template_name, {self.context_object_name: context})
else:
context = {
'level': level,
'n_data_normalisasi': [],
'n_list_data_kernel_view': [],
'display_form': 'block',
'display_result': 'none',
'form': form
}
return render(request, self.template_name, {self.context_object_name: context})
def get_queryset(self, **kwargs):
try:
data = Parameter.objects.get(id='1')
if data is None:
form = ParameterForm()
lamda = None
complexity = None
gamma = None
iterasi = None
s = '2'
else:
lamda = float(data.lamda)
complexity = float(data.complexity)
gamma = float(data.gamma)
iterasi = int(data.iterasi)
s = data.sigma
if s is None or not s.strip():
s = '2'
form = ParameterForm(
initial={
'lamda': data.lamda,
'complexity': data.complexity,
'gamma': data.gamma,
'iterasi': data.iterasi
})
except Parameter.DoesNotExist:
form = ParameterForm()
lamda = None
complexity = None
gamma = None
iterasi = None
s = '2'
level = self.kwargs['level']
n_data_normalisasi = []
n_list_data_matriks_view = []
data_iterasi = []
data_bobot = []
bias = 0
if level == 1:
display_form = 'block'
display_result = 'none'
else:
display_form = 'none'
if lamda is not None and complexity is not None and gamma is not None and iterasi is not None:
display_result = 'block'
data_normalisasi = normalisasi.get_normalisasi(
level)['n_data_normalisasi']
matriks = training.get_matriks(data_normalisasi, lamda,
float(s))
n_data_normalisasi = matriks['n_data_normalisasi']
n_list_data_kernel = matriks['n_list_data_kernel']
n_list_data_matriks = matriks['n_list_data_matriks']
n_list_data_matriks_view = matriks['n_list_data_matriks_view']
data_iterasi = training.get_iterasi(n_list_data_matriks,
complexity, gamma, iterasi)
if len(data_iterasi) > 0:
dt = training.get_bias(
level, n_data_normalisasi,
data_iterasi[len(data_iterasi) - 1]['data_alfa_baru'],
n_list_data_kernel)
data_bobot = dt['data_bobot']
bias = dt['bias']
else:
display_result = 'none'
context = {
'level': level,
'n_data_normalisasi': n_data_normalisasi,
'n_list_data_matriks_view': n_list_data_matriks_view,
'data_iterasi': data_iterasi,
'data_bobot': data_bobot,
'bias': bias,
'display_form': display_form,
'display_result': display_result,
'form': form
}
return context
def post(self, request, *args, **kwargs):
form = ParameterForm(request.POST)
level = self.kwargs['level']
if form.is_valid():
lamda = float(form.cleaned_data['lamda'])
complexity = float(form.cleaned_data['complexity'])
gamma = float(form.cleaned_data['gamma'])
iterasi = int(form.cleaned_data['iterasi'])
try:
param = Parameter.objects.get(id='1')
s = param.sigma
if s is None or not s.strip():
s = '2'
except Parameter.DoesNotExist:
param = Parameter()
param.id = '1'
s = '2'
param.sigma = s
param.lamda = lamda
param.complexity = complexity
param.gamma = gamma
param.iterasi = iterasi
param.save()
# Save to DB
normalisasi.save_normalisasi_to_db()
l_n_data_normalisasi = []
l_n_list_data_matriks_view = []
l_data_iterasi = []
l_data_bobot = []
l_bias = 0
for i in range(7):
lv = i + 1
data_normalisasi = normalisasi.get_normalisasi(
lv)['n_data_normalisasi']
matriks = training.get_matriks(data_normalisasi, lamda,
float(s))
n_data_normalisasi = matriks['n_data_normalisasi']
n_list_data_kernel = matriks['n_list_data_kernel']
n_list_data_matriks = matriks['n_list_data_matriks']
n_list_data_matriks_view = matriks['n_list_data_matriks_view']
data_iterasi = training.get_iterasi(n_list_data_matriks,
complexity, gamma, iterasi)
data_bobot = []
bias = 0
if len(data_iterasi) > 0:
dt = training.get_bias(
lv, n_data_normalisasi,
data_iterasi[len(data_iterasi) - 1]['data_alfa_baru'],
n_list_data_kernel)
data_bobot = dt['data_bobot']
bias = dt['bias']
if lv == level:
l_n_data_normalisasi = n_data_normalisasi
l_n_list_data_matriks_view = n_list_data_matriks_view
l_data_iterasi = data_iterasi
l_data_bobot = data_bobot
l_bias = bias
context = {
'level': level,
'n_data_normalisasi': l_n_data_normalisasi,
'n_list_data_matriks_view': l_n_list_data_matriks_view,
'data_iterasi': l_data_iterasi,
'data_bobot': l_data_bobot,
'bias': l_bias,
'display_form': 'block',
'display_result': 'block',
'form': form
}
return render(request, self.template_name,
{self.context_object_name: context})
else:
context = {
'level': level,
'n_data_normalisasi': [],
'n_list_data_matriks_view': [],
'data_iterasi': [],
'display_form': 'block',
'display_result': 'none',
'form': form
}
return render(request, self.template_name,
{self.context_object_name: context})
def post(self, request, *args, **kwargs):
form = ParameterFormKfold(request.POST)
if form.is_valid():
sigma = float(form.cleaned_data['sigma'])
lamda = float(form.cleaned_data['lamda'])
complexity = float(form.cleaned_data['complexity'])
gamma = float(form.cleaned_data['gamma'])
iterasi = int(form.cleaned_data['iterasi'])
split = int(form.cleaned_data['split'])
try:
param = Parameter.objects.get(id='1')
s = param.sigma
if s is None or not s.strip():
s = '2'
except Parameter.DoesNotExist:
param = Parameter()
param.id = '1'
s = '2'
param.sigma = sigma
param.lamda = lamda
param.complexity = complexity
param.gamma = gamma
param.iterasi = iterasi
param.save()
list_data_bias = []
cv = StratifiedKFold(n_splits=split, shuffle=True, random_state=42)
for i in range(7):
lv = i + 1
dt_normalisasi = normalisasi.get_normalisasi(lv)
x = dt_normalisasi['data_normalisasi_x']
y = dt_normalisasi['data_normalisasi_y']
xK = np.array(x)
yK = np.array(y)
data_bias = []
for train_index, test_index in cv.split(xK, yK):
matriks = training.get_matriks(xK[train_index].tolist(),
lamda, float(s), True)
n_data_normalisasi = matriks['n_data_normalisasi']
n_list_data_kernel = matriks['n_list_data_kernel']
n_list_data_matriks = matriks['n_list_data_matriks']
data_iterasi = training.get_iterasi(
n_list_data_matriks, complexity, gamma, iterasi)
data_alpha = data_iterasi[len(data_iterasi) -
1]['data_alfa_baru']
for j, x in enumerate(n_data_normalisasi):
x[5] = data_alpha[j]
bias = 0
if len(data_iterasi) > 0:
dt = training.get_bias(lv, n_data_normalisasi,
data_alpha, n_list_data_kernel,
False)
bias = dt['bias']
data_bias.append(bias)
list_data_bias.append(data_bias)
accuracy = []
for i in range(7):
lv = i + 1
dt_normalisasi = normalisasi.get_normalisasi(lv)
x = dt_normalisasi['data_normalisasi_x']
y = dt_normalisasi['data_normalisasi_y']
xK = np.array(x)
yK = np.array(y)
data_akurasi = []
y = 0
for train_index, test_index in cv.split(xK, yK):
y = y + 1
matriks = training.get_matriks(xK[train_index].tolist(),
lamda, float(s), True)
n_data_normalisasi = matriks['n_data_normalisasi']
n_list_data_matriks = matriks['n_list_data_matriks']
data_iterasi = training.get_iterasi(
n_list_data_matriks, complexity, gamma, iterasi)
data_alpha = data_iterasi[len(data_iterasi) -
1]['data_alfa_baru']
for j, x in enumerate(n_data_normalisasi):
x[5] = data_alpha[j]
data_testing = testing.proses_testing(
True, xK[test_index].tolist(), n_data_normalisasi,
list_data_bias[i][y - 1])
akurasi = []
for j, x in enumerate(data_testing):
akurasi.append(float(x['data_akurasi']))
data_akurasi.append(sum(akurasi) / len(akurasi))
ac = sum(data_akurasi) / y
accuracy.append(ac)
context = {
'scores': [],
'scores_mean': (sum(accuracy) / len(accuracy)),
'data_evaluasi': [],
'display': 'block',
'form': form
}
return render(request, self.template_name,
{self.context_object_name: context})
else:
context = {
'scores': [],
'scores_mean': 0,
'data_evaluasi': [],
'display': 'none',
'form': form
}
return render(request, self.template_name,
{self.context_object_name: context})