-
Notifications
You must be signed in to change notification settings - Fork 1
/
xgboost_DNN_grid.py
202 lines (186 loc) · 12.7 KB
/
xgboost_DNN_grid.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
######################################################### 加载模块
import sys
from tensorflow import keras as K
import tensorflow as tf
from tensorflow.keras import regularizers
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelBinarizer
from tensorflow.keras.wrappers.scikit_learn import KerasClassifier
from sklearn.model_selection import cross_val_score
from sklearn.model_selection import KFold
import numpy as np
from sklearn.model_selection import StratifiedShuffleSplit
import matplotlib.pyplot as plt
from tensorflow.keras.backend import clear_session
from sklearn.model_selection import PredefinedSplit
import math
import pandas as pd
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.wrappers.scikit_learn import KerasClassifier
from sklearn.preprocessing import MinMaxScaler
from sklearn.model_selection import train_test_split
from sklearn.model_selection import GridSearchCV, RandomizedSearchCV
########################################################################### 构建划分函数
# def load_data(CSV_FILE_PATH):
# IRIS = pd.read_csv(CSV_FILE_PATH)
# target_var = 'project_id' # 目标变量
# # 数据集的特征
# features = list(IRIS.columns)
# features.remove(target_var)
# # 目标变量的类别
# Class = IRIS[target_var].unique()
# # 目标变量的类别字典
# Class_dict = dict(zip(Class, range(len(Class))))
# # 增加一列target, 将目标变量进行编码
# IRIS['target'] = IRIS[target_var].apply(lambda x: Class_dict[x])
# # 对目标变量进行0-1编码(One-hot Encoding)
# lb = LabelBinarizer()
# lb.fit(list(Class_dict.values()))
# transformed_labels = lb.transform(IRIS['target'])
# y_bin_labels = [] # 对多分类进行0-1编码的变量
# for i in range(transformed_labels.shape[1]):
# y_bin_labels.append('y' + str(i))
# IRIS['y' + str(i)] = transformed_labels[:, i]
# # 将数据集分为训练集和测试集
# train_x, test_x, train_y, test_y = train_test_split(IRIS[features], IRIS[y_bin_labels], stratify=IRIS['project_id'],\
# train_size=0.8, test_size=0.2, random_state=1)
# return train_x, test_x, train_y, test_y
def FindLayerNodesLinear(n_layers, first_layer_nodes, last_layer_nodes):
layers = []
nodes_increment = (last_layer_nodes - first_layer_nodes)/ (n_layers-1)
nodes = first_layer_nodes
for i in range(1, n_layers+1):
layers.append(math.ceil(nodes))
nodes = nodes + nodes_increment
return layers
def FinddropoutLinear(n_layers, dropout):
layers = []
nodes_increment = round(dropout/(n_layers-1),2)
nodes = dropout
for i in range(1, n_layers+1):
layers.append(nodes)
nodes = round(nodes - nodes_increment,2)
if(nodes <= 0):
nodes = 0
return layers
def createmodel(n_layers, first_layer_nodes, last_layer_nodes, activation_func, loss_func,dropout):
model = Sequential()
n_nodes = FindLayerNodesLinear(n_layers, first_layer_nodes, last_layer_nodes)
n_dropout = FinddropoutLinear(n_layers,dropout)
for i in range(1, n_layers):
if i==1:
model.add(Dense(first_layer_nodes, input_dim=new_train_x.shape[1], activation=activation_func))
model.add(K.layers.Dropout(rate=n_dropout[0]))
else:
model.add(Dense(n_nodes[i-1], activation=activation_func))
model.add(K.layers.Dropout(rate=n_dropout[i-1]))
model.add(Dense(train_y.shape[1], activation='softmax'))
model.compile(optimizer='adam', loss=loss_func, metrics = ["accuracy"]) #note: metrics could also be 'mse'
return model
n = sys.argv[1]
print(n)
n = int(n)
xgboost_feature = pd.read_csv("/public/slst/home/ningwei/methylation/process_data/feature_importance_12_6/Xgboost_feature"+str(n)+".csv")
data = pd.read_csv("/public/slst/home/ningwei/methylation/process_data/train_data/train_data_smote_12_6.csv")
Class = data['project_id'].unique()
Class_dict = dict(zip(Class, range(len(Class))))
data["target"] = data.iloc[:,0].apply(lambda x: Class_dict[x])
# 对目标变量进行0-1编码(One-hot Encoding)
lb = LabelBinarizer()
lb.fit(list(Class_dict.values()))
transformed_labels = lb.transform(data['target'])
train_y = transformed_labels
train_x = data.iloc[:,1:data.shape[1]]
data = pd.read_csv("/public/slst/home/ningwei/methylation/process_data/train_data/test_data_smote_12_7.csv")
data["target"] = data.iloc[:,0].apply(lambda x: Class_dict[x])
lb = LabelBinarizer()
lb.fit(list(Class_dict.values()))
transformed_labels = lb.transform(data['target'])
test_y = transformed_labels
test_x = data.iloc[:,1:data.shape[1]]
m = sys.argv[2]
print(m)
m = int(m)
new_train_x = train_x[xgboost_feature.iloc[0:m,0]]
new_test_x = test_x[xgboost_feature.iloc[0:m,0]]
#train_x1, value_x, train_y1, value_y = train_test_split(new_train_x, train_y,train_size=0.8, test_size=0.2, random_state=1)
train_x1 = new_train_x
value_x = new_test_x
train_y1 = train_y
value_y = test_y
# ############################################################### 对数据进行标准化
# train_x1_zcore=(train_x1-new_train_x.mean(axis=0))/new_train_x.std(axis=0)
# value_x_zcore=(value_x-new_train_x.mean(axis=0))/new_train_x.std(axis=0)
# ############################################################################ 构建函数
# train_val_features = np.concatenate((train_x1,value_x),axis=0)
# train_val_labels = np.concatenate((train_y1,value_y ),axis=0)
# test_fold = np.zeros(train_val_features.shape[0])
# test_fold[:train_x1.shape[0]] = -1
# ps = PredefinedSplit(test_fold=test_fold)
#####################################################################设置参数范围
model = KerasClassifier(build_fn=createmodel, verbose = False)
activation_funcs = ['relu']
#activation_funcs = ['relu']
loss_funcs = ['categorical_crossentropy']
if m ==5 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [50,40], last_layer_nodes = [40], dropout=[0,0.1], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m ==10 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [100,80,60], last_layer_nodes = [50,40], dropout=[0,0.3], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 15 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [150,130,110,90], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 20 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [200,180,160,140], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 25 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [250,230,200,180,160,130], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 30 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [300,280,250,230,200,180,150], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 35 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [350,330,310,280,250,230,200,180], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 40 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [400,380,360,340,310,280,250,230,200], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 45 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [450,430,410,400,380,360,340,310,280,250], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
if m == 50 :
#param_grid = dict(n_layers=[2,3,4,5,6,7], first_layer_nodes = [10000,8000,6000,5000,4000], last_layer_nodes = [60,90,120], dropout=[0,0.1,0.3,0.5,0.7,0.9], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50])
param_grid = dict(n_layers=[3,4,5,6], first_layer_nodes = [500,470,450,430,410,400,380,360,340,310,280,250], last_layer_nodes = [50,40], dropout=[0,0.3,0.5], activation_func = activation_funcs, loss_func = loss_funcs, batch_size = [100,50], epochs = [50,100])
grid = GridSearchCV(estimator = model, param_grid = param_grid,n_jobs=1)
#grid = RandomizedSearchCV (estimator = model, param_grid = param_grid,cv=3,n_jobs=5)
################################################################ 进行训练
grid.fit(train_x1, train_y1)
############################################################### 输出最高精确度及相应参数
print(grid.best_score_)
print(grid.best_params_)
f = open("/public/slst/home/ningwei/methylation/process_data/DNN_train_params_12_8/"+"xgboost_"+str(n)+"_"+str(m)+".txt",'a') #若文件不存在，系统自动创建。'a'表示可连续写入到文件，保留原内容，在原内容之后写入。可修改该模式（'w+','w','wb'等）
f.write(str(grid.best_score_)) #将字符串写入文件中
f.write("\n") #换行
f.write(str(grid.best_params_)) #将字符串写入文件中
f.write("\n") #换行
f.close()
R_predict = grid.predict(value_x)
R_predict = np.argmax(R_predict, axis=-1)
y_really = np.argmax(value_y, axis=-1)
accuracy = accuracy_score(y_really,R_predict)
F1_score = f1_score(y_really,R_predict,average='weighted')
recall = recall_score(y_really, R_predict, average='weighted')
precision = precision_score(y_really, R_predict,average='weighted')
f = open("/public/slst/home/ningwei/methylation/process_data/DNN_train_data_12_6/"+"xgboost_"+str(n)+"_"+str(m)+".txt",'a') #若文件不存在，系统自动创建。'a'表示可连续写入到文件，保留原内容，在原内容之后写入。可修改该模式（'w+','w','wb'等）
f.write("test_acc: "+str(accuracy)) #将字符串写入文件中
f.write("\n") #换行
f.write("test_recall: "+str(recall)) #将字符串写入文件中
f.write("\n") #换行
f.write("test_precision: "+str(precision)) #将字符串写入文件中
f.write("\n") #换行
f.write("test_F1: "+str(F1_score)) #将字符串写入文件中
f.write("\n") #换行
f.close()