def index():
l = {"^IXIC": "NASDAQ", "^NYA": "NYSE", "^XAX": "AMEX"}
r = []
for k, v in l.items():
r1 = getdata.get(k, v)
r.append(r1)
cdn_js = CDN.js_files[0]
cdn_css = CDN.css_files[0]
return render_template("index.html", r=r, cdn_css=cdn_css, cdn_js=cdn_js)
def plot(n):
# if n == "Google":
# a = "GOOG"
# elif n == "Facebook":
# a = "FB"
# else:
# a = "DXC"
l = {
"Google": "GOOG",
"Facebook": "FB",
"DXC": "DXC",
"Tesla": "TSLA",
"Apple": "AAPL",
"IBM": "IBM",
"Twitter": "TWTR",
"Amazon": "AMZN",
"Microsoft": "MSFT",
"Dell": "DELL",
"Cisco": "CSCO",
"VmWare": "VMW",
"AMD": "AMD",
"Intel": "INTC",
"Dow Jones": "^DJI",
"NASDAQ": "^IXIC",
"NYSE": "^NYA",
"AMEX": "^XAX",
"Alibaba": "BABA",
"Addiko Bank": "ADKO.VI"
}
a = l[n]
cdn_js = CDN.js_files[0]
cdn_css = CDN.css_files[0]
r1 = getdata.get(a, n)
return render_template("plot.html",
script1=r1[0],
div1=r1[1],
cdn_css=cdn_css,
cdn_js=cdn_js,
n1=r1[2],
n=n)
print('ALL DONE')
def makeonehot(X, dim):
res = []
for i in X:
here = np.zeros(dim)
here[i] = 1
res.append(here)
res = np.asarray(res)
return res
if __name__ == "__main__":
classnum = 99
X, Y = getdata.get('/home/lihang/2017/bdimg/data/train_data2/', 224, 224,
3)
X_train, X_test, Y_train, Y_test = train_test_split(X,
Y,
train_size=0.93,
random_state=42)
Y_train = makeonehot(Y_train, classnum)
Y_test = makeonehot(Y_test, classnum)
train(X_train,
Y_train,
X_test,
Y_test,
depoch=25,
ftepoch=201,
batch_size=32,
classnum=classnum,
out='2inception.model')
sheet.write(0, 7, '公交时间/min')
sheet.write(0, 8, '驾车距离/km')
sheet.write(0, 9, '驾车时间/min')
lists = [[10, 14], [10, 15], [11, 14], [11, 15], [12, 8], [12, 9], [12, 10],
[12, 11], [12, 12], [12, 13]]
index = 0
line = 1 # 当前待写入待行号
while (index < 10):
for i in range(nrows):
for j in range(ncols):
x = lists[index][0] - 1
y = lists[index][1] - 1
if ((i <= 7) and (i >= 3) and (j <= 3) and
(j >= 0)) or ((i == x) and (j == y)):
continue
else:
sheet.write(line, 0, '%d,%d' % (i + 1, j + 1)) # 出发地
sheet.write(line, 1, '%d,%d' % (x + 1, y + 1)) # 目的地
for k in range(4): # 获取并打印四种出行方式的距离、时间
distance, duration = getdata.get(table.cell_value(i, j),
table.cell_value(x, y), k)
sheet.write(line, 2 + k * 2,
round(int(distance) / 1000, 1))
sheet.write(line, 2 + k * 2 + 1, round(int(duration) / 60))
line = line + 1
print(line)
index = index + 1
wbk.save('data_final.xls')
def forward(self, x):
x = self.conv1(x.unsqueeze(1))
#print(x.size())
x = self.conv2(x)
#print(x.size())
x = x.view(x.size(0), -1)
x = self.lc1(x)
#x = self.lc2(x)
x = self.out(x)
return x
net = Net().cuda()
print(net)
bsize = 500
log_train, log_test, label_train, label_test = getdata.get()
Train_data = Data.TensorDataset(log_train, label_train)
Test_data = Data.TensorDataset(log_test, label_test)
train_data = Data.DataLoader(dataset=Train_data,
batch_size=bsize,
shuffle=False)
test_data = Data.DataLoader(dataset=Test_data, batch_size=bsize, shuffle=False)
optimizer = optim.SGD(net.parameters(),
lr=0.01,
momentum=0.5,
weight_decay=1e-9)
loss_function = nn.CrossEntropyLoss()
for epoch in range(1000):
#print("Epoch: {}".format(epoch))
running_loss = 0.0
args = parser.parse_args()
np.random.seed(1) #固定下来随机化shuffle的序列
image_height = 30 #数字图片应该普遍长宽比例是这样
image_width = 120
image_channel = 1
#gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.2)
# 训练，测试，持久化
#with tf.Session(config=tf.ConfigProto(gpu_options=gpu_options)) as sess:
os.environ["CUDA_VISIBLE_DEVICES"] = args.gpu_core
config = tf.ConfigProto()
#config.gpu_options.per_process_gpu_memory_fraction = 0.5 # 占用GPU90%的显存
config.gpu_options.allow_growth = True
with tf.Session(config=config) as sess:
#with tf.Session() as sess:
# 完成数据的读取，使用的是tensorflow的读取图片
X, Y, labellen = getdata.get(image_height, image_width, image_channel)
print(max(labellen))
num_class = getdata.classnum() + 1
print('numclass:', num_class)
# 将数据集shuffle
X, Y = util.shuffledata(X, Y)
# 将数据区分为测试集合和训练集合
X_train, X_test, Y_train, Y_test = train_test_split(X,
Y,
train_size=0.9,
random_state=33)
print('Train: ', len(X_train))
print('Test: ', len(X_test))
model = MultiNet(image_height, image_width, image_channel, num_class)
model.train(sess,
X_train,
print('ALL DONE')
def makeonehot(X, dim):
res = []
for i in X:
here = np.zeros(dim)
here[i] = 1
res.append(here)
res = np.asarray(res)
return res
if __name__ == "__main__":
classnum = 99
X, Y = getdata.get('../data/train_data2/')
X_train, X_test, Y_train, Y_test = train_test_split(X,
Y,
train_size=0.95,
random_state=79)
Y_train = makeonehot(Y_train, classnum)
Y_test = makeonehot(Y_test, classnum)
train(X_train,
Y_train,
X_test,
Y_test,
depoch=35,
ftepoch=50,
batch_size=32,
classnum=classnum,
out='inception.model')
model.save(out)
# X_test = preprocess_input(X_test)
#score, acc = model.evaluate(X_test, Y_test, batch_size=batch_size)
#print('now accu:',acc)
print('ALL DONE')
def makeonehot(X, dim):
res = []
for i in X:
here = np.zeros(dim)
here[i] = 1
res.append(here)
res = np.asarray(res)
return res
if __name__ == "__main__":
X, Y = getdata.get('/home/lihang/2017/bdimg/data/train_data/')
#X_train, X_test, Y_train, Y_test = train_test_split(X, Y, train_size=0.9, random_state=33)
Y_train = makeonehot(Y_train, 100)
#Y_test = makeonehot(Y_test,100)
train(X,
Y_train,
None,
None,
epoch=20,
batch_size=32,
out='inceptionv3-ft.model')