class DemoTest(unittest.TestCase):
def setUp(self):
self.host_url = r'http://127.0.0.1:12356'
self.comm = Common(self.host_url)
@unittest.skip("demonstrating skipping")
def test_nothing(self):
self.fail("shouldn't happen")
def test_index(self):
uri_index = '/'
respanse_index_test = self.comm.get(uri_index)
self.assertEqual(
respanse_index_test.text,
'please input your username(your english name) and password(your english name)'
)
def test_login(self):
uri = r'/login'
username = '******'
password = '******'
payload = 'username='******'&password='******'please select One Equipment:\n10001:Knife\n10002:Big Sword\n10003:KuiHuaBaoDian'
)
def tearDown(self):
pass
class SampleTest(unittest.TestCase):
def setUp(self):
self.host_url = r'http://127.0.0.1:12356'
self.comm = Common(self.host_url)
def test_selectEq(self):
uri_selectEq = r'/selectEq'
equipmentid = '10003'
payload_selectEq = 'equipmentid=' + equipmentid
res_selectEq = self.comm.post(uri_selectEq, payload_selectEq)
self.assertEqual(
res_selectEq.text,
'{"equipmentid": "10003", "Message": "your pick up equipmentid:10003 please select your enemyid:\\n20001:Terran\\n20002:ORC\\n20003:Undead"}'
)
def test_kill(self):
equipmentid = '10003'
enmyid = '20001'
uri_kill = '/kill'
payload_enmyid = 'enmyid=' + enmyid + '&equipmentid=' + equipmentid
res_enmyid = self.comm.post(uri_kill, payload_enmyid)
self.assertEqual(res_enmyid.text, 'Error 9904: Your kill yourself!!')
def tearDown(self):
pass
def __init__(self):
self.driver = WebDriver()
self.driver.implicitly_wait(2)
self.driver.get(Config.main_page)
self.common = Common(self)
self.stf = Stf(self)
self.common.get_seed()
self.common.write_seed_to_file()
class Application:
def __init__(self):
self.driver = WebDriver()
self.driver.implicitly_wait(2)
self.driver.get(Config.main_page)
self.common = Common(self)
self.stf = Stf(self)
self.common.get_seed()
self.common.write_seed_to_file()
def destroy(self):
self.driver.close()
self.driver.quit()
def content_cost(input, target):
# First normalize both the input and target (preprocess for VGG16)
input_norm = normalize_batch(input)
target_norm = normalize_batch(target)
input_layers = Common.forward_vgg(input_norm, False)
target_layers = Common.forward_vgg(target_norm, False)
accumulated_loss = 0
for layer in range(len(input_layers)):
accumulated_loss = accumulated_loss + torch.mean(
torch.square(input_layers[layer] - target_layers[layer]))
return accumulated_loss
def test_common():
"""this is not a real test
It's just for see print statements.
Try exchange del and print statement positions
"""
obj = Common()
del obj
print('hi')
def style_cost(input, target):
# First normalize both the input and target (preprocess for VGG16)
input_norm = normalize_batch(input)
target_norm = normalize_batch(target)
input_layers = Common.forward_vgg(input_norm, True)
target_layers = Common.forward_vgg(target_norm, True)
# layer weights
layer_weights = [0.3, 0.7, 0.7, 0.3]
# The accumulated losses for the style
accumulated_loss = 0
for layer in range(len(input_layers)):
accumulated_loss = accumulated_loss + layer_weights[layer] * \
torch.mean(torch.square(compute_gram(input_layers[layer]) -
compute_gram(target_layers[layer])))
return accumulated_loss
def show_img(img):
# Convert to tensor
img = torch.from_numpy(img.reshape(1, 3, 256, 256)).float()
# Put through network
gen_img = Common.transformation_net(img)
gen_img = gen_img.detach().numpy()
# Clip the floats
gen_img = np.clip(gen_img, 0, 255)
# Convert to ints (for images)
gen_img = gen_img.astype('uint8')
gen_img = gen_img.reshape(3, 256, 256).transpose(1, 2, 0)
# Show the image
plt.imshow(gen_img)
plt.show()
def training_show_img():
# Get an image from the validation set
img = load_training_batch(0, 10, 'val')[4]
# Convert to tensor
train_img = torch.from_numpy(img.reshape(1, 3, 256, 256)).float()
# Put through network
gen_img = Common.transformation_net(train_img)
gen_img = gen_img.detach().numpy()
# Clip the floats
gen_img = np.clip(gen_img, 0, 255)
# Convert to ints (for images)
gen_img = gen_img.astype('uint8')
gen_img = gen_img.reshape(3, 256, 256).transpose(1, 2, 0)
# Show the image
plt.imshow(gen_img)
plt.show()
def setUp(self):
self.host_url = r'http://127.0.0.1:12356'
self.comm = Common(self.host_url)
from src.common import Common
# 实例化Common
host_url = r'http://127.0.0.1:12356'
comm = Common(host_url)
# 访问首页
uri_index = '/'
res_index = comm.get(uri_index)
print('访问首页' + res_index.text)
# 存储战场的登录
uri = r'/login'
username = '******'
password = '******'
payload = 'username='******'&password='******'/selectEq'
equipmentid = '10003'
payload_selectEq = 'equipmentid=' + equipmentid
res_selectEq = comm.post(uri_selectEq, payload_selectEq)
print(u'存储战场的武器选择' + res_selectEq.text)
# 存储战场另一个武器选择
enmyid = '20001'
uri_kill = '/kill'
payload_enmyid = 'enmyid=' + enmyid + '&equipmentid=' + equipmentid
def test_str():
obj = Common()
assert str(obj) == 'Common'
def test_repr():
obj = Common()
assert repr(obj) == '<Common>'
# The content batch is the same as the train batch, except train batch has noise added to it
train_batch = load_training_batch(batch, Common.BATCH_SIZE, 'train')
content_batch = np.copy(train_batch)
# Add noise to the training batch
train_batch = add_noise(train_batch)
# Convert the batches to tensors
train_batch = torch.from_numpy(train_batch).float()
content_batch = torch.from_numpy(content_batch).float()
# Zero the gradients
opt.zero_grad()
# Forward propagate
gen_images = Common.transformation_net(train_batch)
# Compute loss
loss = total_cost(gen_images,
[content_batch, Common.STYLE_IMAGE_TENSOR])
# Backprop
loss.backward()
# Clip the gradient to minimize chance of exploding gradients
torch.nn.utils.clip_grad_norm_(Common.transformation_net.parameters(),
1.0)
# Apply gradients
opt.step()
def test_get_stock_codes_from_tse(self):
stock_codes = Common.get_stock_codes_from_tse((2017, 9, 7))
self.assertEqual(912, len(stock_codes))
def test_is_in_future(self):
tomorrow_obj = datetime.date.today() + datetime.timedelta(1)
self.assertTrue(Common.is_in_future(tomorrow_obj))
with self.assertRaises(TypeError):
Common.is_in_future('abc')