def Evaluate():
open('input/points.csv',"w")
labels = [['s.no', 'id']]
for i in range(10000):
labels[0].append('x'+ str(i+1))
labels[0].append('y'+ str(i+1))
csvData = labels
with open('input/points.csv', 'w') as csvFile:
writer = csv.writer(csvFile)
writer.writerows(csvData)
filepath = os.path.join(app.config['UPLOAD_FOLDER'], "input.txt")
fo = open(os.path.join(app.config['UPLOAD_FOLDER'], "input.txt"), "r")
filename = os.path.basename(fo.name)
test = "HEGSE"
csv_list = []
if filename.find(test) == -1:
# sno = sno+1
return_list = getpoints(filepath,1)
csv_list.append(return_list)
with open('input/points.csv',"a") as csvFile:
writer = csv.writer(csvFile)
writer.writerows(csv_list)
getvalues=[0,0,0,0]
pathtocsv = os.path.join(app.config['UPLOAD_FOLDER'], "points.csv")
pathtoMODEL = os.path.join(app.config['MODEL_FOLDER'], "")
typ=""
typ, data_x, scaler_rob_x, X, fid = data_pre(pathtocsv, pathtoMODEL)
getvalues= get_val(data_x, typ, scaler_rob_x, X,pathtoMODEL)
if(typ == "on"):
k = "ON"
else:
k="OFF"
emit('Predict',{ 'typ': str( k), 'val2': float(getvalues[1]), 'val3':float(getvalues[2]), 'val4': float(getvalues[3]), 'fid': str(fid) })
emit("Y")
def train_model(self, train_x, train_y):
# train_y = train_y['0'].values
train_y = tf.one_hot(train_y, depth=2)
tensorboard_callback = tf.keras.callbacks.TensorBoard(log_dir=data_pre().loadLog() + "people_npz-32-8-2",
histogram_freq=1)
network = Sequential([
layers.Dense(32, activation="sigmoid"),
# layers.Dense(16, activation="sigmoid"),
layers.Dense(8, activation="sigmoid"),
layers.Dense(2)
])
network.build(input_shape=(None, 400))
network.summary()
network.compile(optimizer=optimizers.Adam(0.001),
loss=losses.CategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
checkpoint = tf.keras.callbacks.ModelCheckpoint(
filepath='my_model.h5',
monitor='val_loss',
verbose=1,
save_weights_only=False,
save_best_only=True,
mode='min',
)
network.fit(x=train_x, y=train_y, epochs=100, validation_split=0.1, validation_freq=1, verbose=1,
callbacks=[checkpoint, tensorboard_callback])
print(network.predict(train_x[:200]))
network.save("my_model_npz.h5")
test = "HEGSE"
csv_list = []
if filename.find(test) == -1:
# sno = sno+1
return_list = getpoints(filepath,1)
csv_list.append(return_list)
with open('input/points.csv',"a") as csvFile:
writer = csv.writer(csvFile)
writer.writerows(csv_list)
getvalues=[0,0,0,0]
pathtocsv = os.path.join(app.config['UPLOAD_FOLDER'], "points.csv")
pathtoMODEL = os.path.join(app.config['MODEL_FOLDER'], "")
typ=""
typ, data_x, scaler_rob_x, X, fid = data_pre(pathtocsv, pathtoMODEL)
getvalues= get_val(data_x, typ, scaler_rob_x, X,pathtoMODEL)
if(typ == "on"):
k = "ON"
else:
k="OFF"
typ = str( k)
val2 = float(getvalues[1])
val3 = float(getvalues[2])
from data_pre import data_pre
import lightgbm as lgb
import pandas as pd
import numpy as np
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import KFold
import xgboost as xgb
train, test = data_pre()
label = train['收率']
test_id = test['样本id']
del test['样本id']
del test['收率']
del train['样本id']
del train['收率']
train.fillna(-1, inplace=True)
test.fillna(-1, inplace=True)
# 五折交叉验证
folds = KFold(n_splits=5, shuffle=True, random_state=2018)
oof_lgb = np.zeros(len(train))
predictions_lgb = np.zeros(len(test))
param = {
'num_leaves': 120,
'min_data_in_leaf': 30,
'objective': 'regression',
'max_depth': -1,
'learning_rate': 0.01,
"min_child_samples": 30,
network.compile(optimizer=optimizers.Adam(0.001),
loss=losses.CategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
checkpoint = tf.keras.callbacks.ModelCheckpoint(
filepath='my_model.h5',
monitor='val_loss',
verbose=1,
save_weights_only=False,
save_best_only=True,
mode='min',
)
network.fit(x=train_x, y=train_y, epochs=100, validation_split=0.1, validation_freq=1, verbose=1,
callbacks=[checkpoint, tensorboard_callback])
print(network.predict(train_x[:200]))
network.save("my_model_npz.h5")
pass
pass
if __name__ == '__main__':
npz_filepath = r"/home/jry/MicroWave_right_npz/mydataset.npz"
# x, y = data_pre().loadData()
x, y =data_pre().load_npz_data(path=npz_filepath)
bp_model().train_model(train_x=x, train_y=y)