把两个结果and下

stock/cnn_predict.py

@@ -9,7 +9,7 @@ from keras.models import load_model
 lines = []
 def read_data(path):
     with open(path) as f:
-        for line in f.readlines()[:2000]:
+        for line in f.readlines()[:]:
             lines.append(eval(line.strip()))
 
     train_x=[s[:-2] for s in lines]
@@ -17,29 +17,44 @@ def read_data(path):
     return np.array(train_x),np.array(train_y)
 
 
-test_x,test_y=read_data("D:\\data\\quantization\\stock6_test.log")
-test_x = test_x.reshape(test_x.shape[0], 1,80,5)
-
-path="15min_model_seq.h5"
-model=load_model(path)
-score = model.evaluate(test_x, test_y)
-print(score)
+def predict():
+    test_x,test_y=read_data("D:\\data\\quantization\\stock6_test.log")
+    test_x = test_x.reshape(test_x.shape[0], 1,80,5)
+
+    path="15min_cnn_seq.h5"
+    model=load_model(path)
+    score = model.evaluate(test_x, test_y)
+    print(score)
+
+    result=model.predict(test_x)
+    print(result)
+
+    up_num = 0
+    up_right = 0
+    i = 0
+    for r in result:
+        fact = test_y[i]
+        if fact[0] == 1:
+            if r[0] > 0.5:
+                up_right = up_right + 1
+            elif r[1] > 0.5:
+                up_right = up_right + 0.5
+            up_num = up_num + 1
+
+        i = i + 1
+    print(up_right, up_num, up_right/up_num)
+
+    i = 0
+    win_dnn = []
+    for r in result:
+        if r[0] > 0.5:
+            print(lines[i][-2])
+            win_dnn.append([lines[i][-2], lines[i][-1]])
+        i = i + 1
 
-result=model.predict(test_x)
-print(result)
+    return win_dnn
 
-up_num = 0
-up_right = 0
-i = 0
-for r in result:
-    fact = test_y[i]
-    if fact[0] == 1:
-        if r[0] > 0.5:
-            up_right = up_right + 1
-        elif r[1] > 0.5:
-            up_right = up_right + 0.9
-        up_num = up_num + 1
 
-    i = i + 1
+if __name__ == '__main__':
+    predict()
 
-print(up_right, up_num, up_right/up_num)

stock/cnn_train.py

@@ -17,7 +17,7 @@ from keras.layers import Conv2D, Activation, MaxPool2D, Flatten, Dense
 def read_data(path):
     lines = []
     with open(path) as f:
-        for x in range(10000):
+        for x in range(40000):
             lines.append(eval(f.readline().strip()))
 
     random.shuffle(lines)
@@ -66,7 +66,7 @@ model.add(Activation('relu'))
 
 # 设置下采样(池化层）
 model.add(MaxPool2D(
-    pool_size=(2,2),  # 下采样格为2*2
+    pool_size=(4,1),  # 下采样格为2*2
     strides=(2,2),  # 向右向下的步长
     padding='same', # padding mode is 'same'
 ))
@@ -93,7 +93,7 @@ print(score)
 print('Test score:', score[0])
 print('Test accuracy:', score[1])
 
-path="15min_model_seq.h5"
+path="15min_cnn_seq.h5"
 model.save(path)
 model=None
 

stock/compont_predict.py

@@ -0,0 +1,21 @@
+from stock import cnn_predict
+from stock import dnn_predict
+
+
+def and_predict():
+    cnn_result = cnn_predict.predict()
+    dnn_result = dnn_predict.predict(file_path='D:\\data\\quantization\\stock6_5_test.log', model_path='5d_dnn_seq.h5')
+    print('计算完成')
+
+    with open('and_predict.txt', 'a') as f:
+        for m in cnn_result:
+            # print('find', m)
+            for n in dnn_result:
+                if m[0][0] == n[0][0] and int(m[0][1][:10].replace('-', '')) == n[0][1]:
+                    print('AND', m, n)
+                    f.write(str(m) + '\n')
+                    break
+
+
+if __name__ == '__main__':
+    and_predict()

stock/dnn_predict.py

@@ -17,21 +17,40 @@ def read_data(path):
     train_y=[s[size-1] for s in lines]
     return np.array(train_x),np.array(train_y)
 
-# test_x,test_y=read_data("D:\\data\\quantization\\stock_test.log")
-test_x,test_y=read_data("D:\\data\\quantization\\s\\stock_2020-01-07.log")
-
-path="model_seq.h5"
-model=load_model(path)
-# score = model.evaluate(test_x, test_y)
-# print(score)
-
-result=model.predict(test_x)
-# print(result)
-i = 0
-for x in result:
-    # print(str(i) + ":" + str(x))
-    if x[0] > 0.8:
-        print(lines[i][-2], x, 1)
-    # elif x[0] + x[1] > 0.9:
-    #     print(lines[i][-2], x, 2)
-    i = i + 1
+
+def predict(file_path='', model_path='15min_dnn_seq.h5'):
+    test_x,test_y=read_data(file_path)
+
+    model=load_model(model_path)
+    score = model.evaluate(test_x, test_y)
+    print(score)
+
+    up_num = 0
+    up_right = 0
+    i = 0
+    result=model.predict(test_x)
+    for r in result:
+        fact = test_y[i]
+        if fact[0] == 1:
+            if r[0] > 0.5:
+                up_right = up_right + 1
+            elif r[1] > 0.5:
+                up_right = up_right + 0.5
+            up_num = up_num + 1
+
+        i = i + 1
+    print(up_right, up_num, up_right/up_num)
+
+    i = 0
+    win_dnn = []
+    for r in result:
+        if r[0] > 0.5:
+            print(lines[i][-2])
+            win_dnn.append([lines[i][-2], lines[i][-1]])
+
+        i = i + 1
+    return win_dnn
+
+
+if __name__ == '__main__':
+    predict(file_path='D:\\data\\quantization\\stock6_5_test.log', model_path='5d_dnn_seq.h5')

stock/dnn_train.py

@@ -9,11 +9,9 @@ from imblearn.over_sampling import RandomOverSampler
 
 
 def read_data(path):
-    train_x=[]
-    train_y=[]
     lines = []
     with open(path) as f:
-        for x in range(20000):
+        for x in range(40000):
             lines.append(eval(f.readline().strip()))
 
     random.shuffle(lines)
@@ -36,47 +34,38 @@ def read_data(path):
 
     return X_resampled,y_resampled,np.array(test_x),np.array(test_y)
 
-train_x,train_y,test_x,test_y=read_data("D:\\data\\quantization\\stock4.log")
-
-model = Sequential()
-model.add(Dense(units=425, input_dim=166,  activation='relu'))
-model.add(Dense(units=325, activation='relu'))
-model.add(Dense(units=325, activation='relu'))
-model.add(Dropout(0.2))
-# model.add(Dense(units=325, activation='relu'))
-# model.add(Dropout(0.2))
-# model.add(Dense(units=325, activation='relu'))
-# model.add(Dropout(0.2))
-# model.add(Dense(units=325, activation='relu'))
-# model.add(Dropout(0.2))
-# model.add(Dense(units=325, activation='relu'))
-# model.add(Dropout(0.2))
-# model.add(Dense(units=225, activation='relu'))
-# model.add(Dense(units=225, activation='relu'))
-# model.add(Dense(units=225, activation='relu'))
-model.add(Dense(units=225, activation='relu'))
-model.add(Dropout(0.2))
-# model.add(Dense(units=225, activation='relu'))
-model.add(Dense(units=125, activation='relu'))
-model.add(Dropout(0.2))
-# model.add(Dense(units=125, activation='relu'))
-model.add(Dense(units=166, activation='relu'))
-model.add(Dense(units=8, activation='softmax'))
-model.compile(loss='categorical_crossentropy', optimizer="adam",metrics=['accuracy'])
-
-print("Starting training ")
-h=model.fit(train_x, train_y, batch_size=64, epochs=26, shuffle=True)
-score = model.evaluate(test_x, test_y)
-print(score)
-print('Test score:', score[0])
-print('Test accuracy:', score[1])
-
-path="model_seq.h5"
-model.save(path)
-model=None
-
-
-model=load_model(path)
-result=model.predict(test_x)
-print(result)
-print(test_y)
+
+def train(input_dim=400, result_class=3, file_path="D:\\data\\quantization\\stock6.log", model_name=''):
+    train_x,train_y,test_x,test_y=read_data(file_path)
+
+    model = Sequential()
+    model.add(Dense(units=625, input_dim=input_dim,  activation='relu'))
+    model.add(Dense(units=525, activation='relu'))
+    model.add(Dense(units=525, activation='relu'))
+    model.add(Dropout(0.2))
+    model.add(Dense(units=525, activation='relu'))
+    model.add(Dropout(0.2))
+    model.add(Dense(units=325, activation='relu'))
+    model.add(Dropout(0.2))
+    model.add(Dense(units=266, activation='relu'))
+    model.add(Dense(units=result_class, activation='softmax'))
+    model.compile(loss='categorical_crossentropy', optimizer="adam",metrics=['accuracy'])
+
+    print("Starting training ")
+    h=model.fit(train_x, train_y, batch_size=64, epochs=26, shuffle=True)
+    score = model.evaluate(test_x, test_y)
+    print(score)
+    print('Test score:', score[0])
+    print('Test accuracy:', score[1])
+
+    model.save(model_name)
+
+    model=None
+    model=load_model(model_name)
+    result=model.predict(test_x)
+    print(result)
+    print(test_y)
+
+
+if __name__ == '__main__':
+    train(input_dim=46, result_class=5, file_path="D:\\data\\quantization\\stock6_5.log", model_name='5d_dnn_seq.h5')