-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtask4run.py
297 lines (272 loc) · 16.3 KB
/
task4run.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
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
import streamlit as st
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
import torch.nn as nn
import torch
import numpy as np
import joblib
#------ 侧边框的东西-------
st.sidebar.expander('')
st.sidebar.expander('')
st.sidebar.subheader('在下方调节你的参数')
moxing=st.sidebar.radio('请选择是否需要使用已经训练好的模型',['否','是'])
number = st.sidebar.number_input('请输入迭代次数:')
epoch = int(number)
learning = st.sidebar.number_input('请输入学习率:')
learning_rate = learning
seq=st.sidebar.selectbox('请输入考虑之前数据的组数',list(range(2,6)))
fangshi = st.sidebar.selectbox('请选择数据集划分比例', ['8:2', '7:3', '6:4'])
# -------------------
#----- 模型系数及一些超参数-------
input_dim = 7 # 数据的特征数
hidden_dim = 64 # 隐藏层的神经元个数
num_layers = 1 # LSTM的层数
output_dim = 1 # 预测值的特征数
pre_days = 7 # 以1周的数据为一组
seq =seq # 用1周前的数据去预测现在数据
batch_size = 128 # 每组数据量
num_epochs = epoch # 模型遍历次数
# ---------------------
#读数据
df_main = pd.read_csv('data_1.csv')
sel_col = ["power_consumption", "low_temp", "high_temp", "kind", "wind", "level", "holiday"]
# 数据归一化
scaler = MinMaxScaler(feature_range=(-1, 1))
for col in sel_col:
df_main[col] = scaler.fit_transform(df_main[col].values.reshape(-1, 1))
df_main['target'] = df_main['power_consumption'].shift(-1)
# LSTM模型 输入为7输出为1 32个神经元
class LSTM(nn.Module):
def __init__(self, input_dim, hidden_dim, num_layers, output_dim):
super(LSTM, self).__init__()
self.hidden_dim = hidden_dim
self.num_layers = num_layers
self.lstm = nn.LSTM(input_dim, hidden_dim, num_layers, batch_first=True, dropout = 0.5)
# 全连接层
self.fc = nn.Linear(hidden_dim, output_dim)
def forward(self, x):
h0 = torch.zeros(self.num_layers, x.size(0), self.hidden_dim).requires_grad_()
c0 = torch.zeros(self.num_layers, x.size(0), self.hidden_dim).requires_grad_()
out, (hn, cn) = self.lstm(x, (h0.detach(), c0.detach()))
out = self.fc(out)
return out
# 创建两个列表，用来存储数据的特征和标签
data_feat, data_target = [], []
for index in range(len(df_main) - seq):
# 构建特征集
data_feat.append(df_main[["power_consumption", "low_temp", "high_temp", "kind", "wind", "level", "holiday"]][
index: index + seq].values)
# 构建target集
data_target.append(df_main['target'][index:index + seq])
# 将特征集和标签集整理成numpy数组
data_feat = np.array(data_feat)
data_target = np.array(data_target)
# 这里按照8:2的比例划分训练集和测试集
if fangshi=="8:2":
test_set_size = 122 # np.round(1)是四舍五入，
train_size = data_feat.shape[0] - (test_set_size)
elif fangshi=="7:3":
test_set_size = 183 # np.round(1)是四舍五入，
train_size = data_feat.shape[0] - (test_set_size)
elif fangshi=="6:4":
test_set_size = 244 # np.round(1)是四舍五入，
train_size = data_feat.shape[0] - (test_set_size)
trainX = torch.from_numpy(data_feat[:train_size].reshape(-1, seq, 7)).type(torch.Tensor)
testX = torch.from_numpy(data_feat[train_size:].reshape(-1, seq, 7)).type(torch.Tensor)
trainY = torch.from_numpy(data_target[:train_size].reshape(-1, seq, 1)).type(torch.Tensor)
testY = torch.from_numpy(data_target[train_size:].reshape(-1, seq, 1)).type(torch.Tensor)
xieruwenjian=torch.from_numpy(data_feat[:10].reshape(-1, seq, 7)).type(torch.Tensor)
torch.save(xieruwenjian,'任务四的测试')
print('x_train.shape = ', trainX.shape)
print('y_train.shape = ', trainY.shape)
print('x_test.shape = ', testX.shape)
print('y_test.shape = ', testY.shape)
# dataloader读入数据
train = torch.utils.data.TensorDataset(trainX, trainY)
test = torch.utils.data.TensorDataset(testX, testY)
train_loader = torch.utils.data.DataLoader(dataset=train,
batch_size=batch_size,
shuffle=False)
test_loader = torch.utils.data.DataLoader(dataset=test,
batch_size=batch_size,
shuffle=False)
# 封装模型
model = LSTM(input_dim=input_dim, hidden_dim=hidden_dim, output_dim=output_dim, num_layers=num_layers)
# 定义优化器和损失函数
optimiser = torch.optim.Adam(model.parameters(), lr=learning_rate) # 使用Adam优化算法
# 模型评估条件
loss_fn = torch.nn.MSELoss(size_average=True) # 使用均方差作为损失函数
st.markdown('''## <b style="color:white;"><center>1.任务简介</center></b>''', unsafe_allow_html=True) # 展示任务
st.markdown('''### <b style="color:white;">需求分析</b>''', unsafe_allow_html=True)
st.write('''<b style="color:white
;">建立电力预测模型，根据以往的用户用电数据预测未来的用户用电情况，保证企业能够及时调整电力市场的需求水平、需求时间，以良好的服务质量满足用户合理用电的要求，实现电力供求之间的相互协调。</b>''',
unsafe_allow_html=True)
st.markdown('''### <b style="color:white;">方案设计</b>''', unsafe_allow_html=True)
st.write('''<b style="color:white;">使用1500
名用户在一年半内的用电情况数据，利用各用户一年半内的用电情况、一年半内的天气情况和一年半内的节假日情况，在完成对数据的预处理和可视化分析后，对其中特征进行选取后，使用LSTM
对数据进行模型训练，通过调整参数以及评估分析，达到一个较优的模型效果。作为最终的企业电力营销模型。</b>''', unsafe_allow_html=True)
st.markdown('''## <b style="color:white;"><center>2.模型数据</center></b>''', unsafe_allow_html=True) # 展示任务
st.write(
'''<b style="color:white;">采用了约1500名用户在一年半内的用电情况数据，总数据包括三个文件，分别介绍了各用户一年半内的用电情况、一年半内的天气情况和一年半内的节假日情况。</b>''',
unsafe_allow_html=True)
df = pd.read_csv('train_data.csv')
st.dataframe(df)
st.markdown('''### <b style="color:white;">数据预处理</b>''', unsafe_allow_html=True)
st.write("""<b style="color:white;">
★ 合并数据集 </b>""", unsafe_allow_html=True)
st.write("""<b style="color:white;">
★ 填充缺失值 </b>""", unsafe_allow_html=True)
st.write("""<b style="color:white;">
★ 字符型特征数值化 </b>""", unsafe_allow_html=True)
st.markdown('''## <b style="color:white;"><center>3.数据可视化分析</center></b>''', unsafe_allow_html=True)
st.write('''<b style="color:white;">从图中可以得出用户id和其余变量之间的相关性系数为0，因此不能将不同用户之间的数据一起进行模型训练。同时，可以得出其余变量对用户用电量都有一定的影响，因此我们不对特征进行筛除，将7个变量对应的数据统一运用到模型训练中。
除此之外，根据数据的特征，我们选取了LSTM模型作为预测模型。</b>''', unsafe_allow_html=True)
st.markdown('''#### <b style="color:white;"><center>各用户的用电量示意图</center></b>''', unsafe_allow_html=True)
st.image('居民用电情况.jpg')
# st.image('''<b style="color:white;"><center>'居民用电情况.jpg'</center></b>''',unsafe_allow_html=True)
st.markdown('''#### <b style="color:white;"><center>各特征之间相关性系数热力图</center></b>''', unsafe_allow_html=True)
st.image('相关系数矩阵.png')
st.markdown('''## <b style="color:white;"><center>4.模型实现与结果</center></b>''', unsafe_allow_html=True)
st.write('''<b style="color:white;">与其他神经网络不同，LSTM通过在原始RNN隐藏层增加单元状态c来保持长期状态，且解决了RNN
存在的长期依赖问题。由于其兼具非线性和时序性的特点，被逐渐应用于电力预测领域并取得了良好的效果。LSTM学习过程由信号的正向传播和误差反向传播两个过程组成。</b>''', unsafe_allow_html=True)
st.markdown('''#### <b style="color:white;"><center>模型结构图</center></b>''', unsafe_allow_html=True)
st.image('模型流程图.png')
# 数据处理
if moxing=='否':
# 训练模型
if epoch!=0:
st.markdown('''#### <b style="color:white;"><center>训练过程损失曲线</center></b>''', unsafe_allow_html=True)
hist = np.zeros(num_epochs)
ls = []
for t in range(num_epochs):
y_train_pred = model(trainX)
loss = loss_fn(y_train_pred, trainY)
ls.append(loss.item())
if t % 10 == 0 and t != 0: # 每训练十次，打印一次均方差
print("Epoch ", t, "MSE: ", loss.item())
hist[t] = loss.item()
# 梯度归零
optimiser.zero_grad()
# Backward
loss.backward()
# 更新参数
optimiser.step()
loss_data = pd.DataFrame(
ls,
columns=["loss值"]
)
st.line_chart(loss_data)
else:
st.markdown('''#### <b style="color:white;"><center>训练过程损失曲线</center></b>''', unsafe_allow_html=True)
ls = [0.5947179794311523, 0.4507795572280884, 0.32666218280792236, 0.22864139080047607, 0.1810951679944992,
0.20072788000106812, 0.219834104180336, 0.1996694654226303, 0.16286492347717285, 0.1347438246011734,
0.12425760924816132, 0.12667830288410187, 0.1316240429878235, 0.13067269325256348, 0.1217602789402008,
0.10766869783401489, 0.0929853618144989, 0.08185668289661407, 0.07632635533809662, 0.07534956187009811,
0.07535826414823532, 0.07276003807783127, 0.06653910875320435, 0.05853329598903656, 0.051560405641794205,
0.047397494316101074, 0.04598104581236839, 0.04590829461812973, 0.04557082802057266, 0.04411137476563454,
0.041716318577528, 0.03926895931363106, 0.03768862038850784, 0.03732573613524437, 0.037740547209978104,
0.038026466965675354, 0.037466708570718765, 0.03600345924496651, 0.03416650742292404, 0.032608285546302795,
0.03166688606142998, 0.031239405274391174, 0.03095424361526966, 0.03045312874019146, 0.029595160856842995,
0.02849569357931614, 0.0274167750030756, 0.026590043678879738, 0.02607543207705021, 0.02574039250612259,
0.02537364326417446, 0.02484879642724991, 0.024207692593336105, 0.023601701483130455, 0.023153360933065414,
0.02285877615213394, 0.022600146010518074, 0.022242626175284386, 0.021733202040195465, 0.02113015204668045,
0.020548896864056587, 0.020070672035217285, 0.019687429070472717, 0.01932711713016033, 0.018933316692709923,
0.018521204590797424, 0.018156995996832848, 0.017887337133288383, 0.017692873254418373,
0.017508702352643013, 0.017285726964473724, 0.017031125724315643, 0.016792116686701775,
0.016604460775852203, 0.01646040380001068, 0.01632745936512947, 0.016191935166716576, 0.01607416197657585,
0.01599947363138199, 0.015963546931743622, 0.015934403985738754, 0.015886016190052032, 0.015821749344468117,
0.015762269496917725, 0.01571686938405037, 0.015674743801355362, 0.015623818151652813, 0.015568578615784645,
0.015522986650466919, 0.015490490943193436, 0.015459226444363594, 0.015417144633829594, 0.01536527182906866,
0.015312861651182175, 0.015263956040143967, 0.015214378945529461, 0.015160984359681606,
0.015107913874089718, 0.01506117731332779, 0.015020622871816158, 0.014980807900428772, 0.014938442967832088,
0.01489558070898056, 0.01485530473291874, 0.014817298389971256, 0.014779355376958847, 0.014741463586688042,
0.014706015586853027, 0.014674236066639423, 0.014644456095993519, 0.014614459127187729, 0.0145840710029006,
0.01455447357147932, 0.014525962062180042, 0.014497620984911919, 0.01446906104683876, 0.014441151171922684,
0.014414721168577671, 0.01438938733190298, 0.014364199712872505, 0.014338935725390911, 0.014314086176455021,
0.0142898578196764, 0.014265892095863819, 0.014241994358599186, 0.014218521304428577, 0.014195832423865795,
0.014173734001815319, 0.014151797629892826, 0.0141299432143569, 0.014108359813690186, 0.01408708281815052,
0.014065933413803577, 0.014044865034520626, 0.01402406394481659, 0.014003628864884377, 0.013983413577079773,
0.013963254168629646, 0.013943170197308064, 0.013923232443630695, 0.013903412036597729,
0.013883628882467747, 0.01386391930282116, 0.013844364322721958, 0.013824944384396076, 0.013805573806166649,
0.01378621719777584, 0.013766906224191189, 0.013747649267315865, 0.01372840628027916, 0.013709168881177902,
0.013689970597624779, 0.013670817017555237, 0.013651669025421143, 0.013632498681545258,
0.013613317161798477, 0.013594131916761398, 0.013574929907917976, 0.013555699028074741, 0.01353644859045744,
0.013517189770936966, 0.013497901149094105, 0.013478566892445087, 0.013459189794957638,
0.013439776375889778, 0.013420315459370613, 0.013400801457464695, 0.013381239026784897, 0.01336162630468607,
0.01334195677191019, 0.013322215527296066, 0.01330240722745657, 0.013282528147101402, 0.013262572698295116,
0.01324253436177969, 0.01322241686284542, 0.013202212750911713, 0.013181917369365692, 0.013161523267626762,
0.013141029514372349, 0.013120434246957302, 0.013099730014801025, 0.013078916817903519,
0.013057989999651909, 0.013036944903433323, 0.013015778735280037, 0.012994487769901752,
0.012973069213330746, 0.012951517477631569, 0.012929831631481647, 0.012908006086945534,
0.012886039912700653, 0.012863928452134132, 0.0128416633233428, 0.01281924732029438, 0.012796674855053425,
0.012773944064974785, 0.012751052156090736, 0.012727992609143257, 0.012704764492809772]
loss_data = pd.DataFrame(
ls,
columns=["loss值"]
)
st.line_chart(loss_data)
st.markdown('''#### <b style="color:white;"><center>测试集用户用电实际值与预测值对比曲线</center></b>''', unsafe_allow_html=True)
if moxing=='否':
# 模型预测
y_test_pred = model(testX)
list1 = []
list2 = []
y_test_pred = model(testX)
y_train_pred=model(trainX)
loss1 = loss_fn(y_test_pred[:-pre_days], testY[:-pre_days]).item()
print(loss1)
pred_value = y_train_pred.detach().numpy()[:, -1, 0]
true_value = trainY.detach().numpy()[:, -1, 0]
pt=[]
test_pre=y_test_pred.detach().numpy()[:, -1, 0]
test_ture=testY.detach().numpy()[:, -1, 0]
pt.append(test_pre)
pt.append(test_ture)
listb = [[r[col] for r in pt] for col in range(len(pt[0]))]
chart_data = pd.DataFrame(
listb,
columns=["预测值", "实际值"]
)
st.line_chart(chart_data)
else:
model=joblib.load('企业电力营销模型.mkl')
list1 = []
list2 = []
y_test_pred = model(testX)
y_train_pred = model(trainX)
loss1 = loss_fn(y_test_pred[:-pre_days], testY[:-pre_days]).item()
print(loss1)
pred_value = y_train_pred.detach().numpy()[:, -1, 0]
true_value = trainY.detach().numpy()[:, -1, 0]
pt = []
test_pre = y_test_pred.detach().numpy()[:, -1, 0]
test_ture = testY.detach().numpy()[:, -1, 0]
pt.append(test_pre)
pt.append(test_ture)
listb = [[r[col] for r in pt] for col in range(len(pt[0]))]
chart_data = pd.DataFrame(
listb,
columns=["预测值", "实际值"]
)
st.line_chart(chart_data)
st.markdown('''#### <b style="color:white;"><center>上传需要预测的数据,可以得到预测值:</center></b>''', unsafe_allow_html=True)
file=st.file_uploader("", type=None, accept_multiple_files=False, key=None, \
help=None, on_change=None)
if file is not None:
data=torch.load(file)
pred=model(data)
pred_value = pred.detach().numpy()[:, -1, 0]
ptyu=[]
ptyu.append(pred_value)
listbyu = [[r[col] for r in ptyu] for col in range(len(ptyu[0]))]
chart_data_yu = pd.DataFrame(
listbyu,
columns=["预测值"]
)
st.write(chart_data_yu)
st.line_chart(chart_data_yu)
img_url = 'https://img.zcool.cn/community/0156cb59439764a8012193a324fdaa.gif'
# 修改背景样式
st.markdown('''<style>.css-fg4pbf{background-image:url(''' + img_url + ''');
background-size:100% 100%;background-attachment:fixed;}</style>
''', unsafe_allow_html=True)