質問編集履歴
5
文法修正
title
CHANGED
|
File without changes
|
body
CHANGED
|
@@ -185,7 +185,7 @@
|
|
|
185
185
|
plt.show()
|
|
186
186
|
```
|
|
187
187
|
|
|
188
|
-
```
|
|
188
|
+
```Error
|
|
189
189
|
Epoch: 1
|
|
190
190
|
---------------------------------------------------------------------------
|
|
191
191
|
AttributeError Traceback (most recent call last)
|
4
変更点追加
title
CHANGED
|
File without changes
|
body
CHANGED
|
@@ -9,7 +9,9 @@
|
|
|
9
9
|
import numpy as np
|
|
10
10
|
import pandas as pd
|
|
11
11
|
import matplotlib.pyplot as plt
|
|
12
|
-
|
|
12
|
+
```
|
|
13
|
+
|
|
14
|
+
```python
|
|
13
15
|
stock_data = pd.read_csv(
|
|
14
16
|
"/content/drive/MyDrive/^GSPC.csv",
|
|
15
17
|
index_col = 0,
|
|
@@ -158,6 +160,8 @@
|
|
|
158
160
|
↓
|
|
159
161
|
train_seq[-test_size:].tolist()
|
|
160
162
|
↓
|
|
163
|
+
```
|
|
164
|
+
```python
|
|
161
165
|
for epoch in range(epochs):
|
|
162
166
|
|
|
163
167
|
print()
|
|
@@ -179,10 +183,9 @@
|
|
|
179
183
|
)
|
|
180
184
|
|
|
181
185
|
plt.show()
|
|
182
|
-
↓
|
|
183
|
-
|
|
186
|
+
```
|
|
184
|
-
↓
|
|
185
187
|
|
|
188
|
+
```エラーコード
|
|
186
189
|
Epoch: 1
|
|
187
190
|
---------------------------------------------------------------------------
|
|
188
191
|
AttributeError Traceback (most recent call last)
|
|
@@ -202,5 +205,4 @@
|
|
|
202
205
|
22 )
|
|
203
206
|
|
|
204
207
|
AttributeError: 'Tensor' object has no attribute 'vierq'
|
|
205
|
-
コード
|
|
206
208
|
```
|
3
コード追加
title
CHANGED
|
File without changes
|
body
CHANGED
|
@@ -2,40 +2,205 @@
|
|
|
2
2
|
株価のデータ可視化の練習中に以下のコードでエラーコードが出てしまいました。
|
|
3
3
|
コードの位置などを再度書き直したのですが、エラーが消えないのでご教授お願いいたします。
|
|
4
4
|
|
|
5
|
+
一連の流れ。
|
|
5
6
|
```python
|
|
7
|
+
import torch
|
|
8
|
+
import torch.nn as nn
|
|
9
|
+
import numpy as np
|
|
10
|
+
import pandas as pd
|
|
11
|
+
import matplotlib.pyplot as plt
|
|
12
|
+
↓
|
|
13
|
+
stock_data = pd.read_csv(
|
|
14
|
+
"/content/drive/MyDrive/^GSPC.csv",
|
|
15
|
+
index_col = 0,
|
|
16
|
+
parse_dates=True
|
|
17
|
+
)
|
|
6
18
|
|
|
19
|
+
stock_data
|
|
20
|
+
↓
|
|
21
|
+
stock_data.drop(
|
|
22
|
+
["Open", "High", "Low", "Close", "Volume"],
|
|
23
|
+
axis="columns",
|
|
24
|
+
inplace=True
|
|
25
|
+
)
|
|
26
|
+
|
|
27
|
+
stock_data
|
|
28
|
+
↓
|
|
29
|
+
stock_data.plot(figsize=(12, 4))
|
|
30
|
+
↓
|
|
31
|
+
y = stock_data["Adj Close"].values
|
|
32
|
+
y
|
|
33
|
+
↓
|
|
34
|
+
from sklearn.preprocessing import MinMaxScaler
|
|
35
|
+
↓
|
|
36
|
+
scaler = MinMaxScaler(feature_range=(-1, 1))
|
|
37
|
+
scaler.fit(y.reshape(-1, 1))
|
|
38
|
+
y =scaler.transform(y.reshape(-1, 1))
|
|
39
|
+
y
|
|
40
|
+
↓
|
|
41
|
+
y = torch.FloatTensor(y).view(-1)
|
|
42
|
+
y
|
|
43
|
+
↓
|
|
44
|
+
test_size = 24
|
|
45
|
+
|
|
46
|
+
train_seq = y[:-test_size]
|
|
47
|
+
test_seq = y[-test_size:]
|
|
48
|
+
↓
|
|
49
|
+
plt.figure(figsize=(12, 4))
|
|
50
|
+
plt.xlim(-20, len(train_seq)+20)
|
|
51
|
+
plt.grid(True)
|
|
52
|
+
plt.plot(train_seq)
|
|
53
|
+
↓
|
|
54
|
+
train_window_size = 12
|
|
55
|
+
↓
|
|
56
|
+
def input_data(seq, ws):
|
|
57
|
+
out = []
|
|
58
|
+
L = len(seq)
|
|
59
|
+
|
|
60
|
+
for i in range(L-ws):
|
|
61
|
+
window = seq[i:i+ws]
|
|
62
|
+
label = seq[i+ws:i+ws+1]
|
|
63
|
+
out.append((window, label))
|
|
64
|
+
|
|
65
|
+
return out
|
|
66
|
+
↓
|
|
67
|
+
train_data = input_data(train_seq, train_window_size)
|
|
68
|
+
↓
|
|
69
|
+
print("The NUmber of Training Data: ",len(train_data))
|
|
70
|
+
↓
|
|
71
|
+
class Model(nn.Module):
|
|
72
|
+
|
|
73
|
+
def __init__(self, input=1, h=50, output=1):
|
|
74
|
+
|
|
75
|
+
super().__init__()
|
|
76
|
+
self.hidden_size = h
|
|
77
|
+
|
|
78
|
+
self.lstm = nn.LSTM(input, h)
|
|
79
|
+
self.fc = nn.Linear(h, output)
|
|
80
|
+
|
|
81
|
+
self.hidden = (
|
|
82
|
+
torch.zeros(1, 1, h),
|
|
83
|
+
torch.zeros(1, 1, h)
|
|
84
|
+
)
|
|
85
|
+
|
|
86
|
+
|
|
87
|
+
def forward(self, seq):
|
|
88
|
+
|
|
89
|
+
out, _=self.lstm(
|
|
90
|
+
seq.vierq(len(seq), 1, -1),
|
|
91
|
+
self.hidden
|
|
92
|
+
)
|
|
93
|
+
|
|
94
|
+
out = self.fc(
|
|
95
|
+
out.view(len(seq), -1)
|
|
96
|
+
)
|
|
97
|
+
|
|
98
|
+
return out[-1]
|
|
99
|
+
↓
|
|
100
|
+
torch.manual_seed(123)
|
|
101
|
+
model = Model()
|
|
102
|
+
criterion = nn.MSELoss()
|
|
103
|
+
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)
|
|
104
|
+
↓
|
|
105
|
+
epochs = 10
|
|
106
|
+
train_losses = []
|
|
107
|
+
test_losses = []
|
|
108
|
+
↓
|
|
109
|
+
def run_train():
|
|
110
|
+
model.train()
|
|
111
|
+
|
|
112
|
+
for train_window, correct_label in train_data:
|
|
113
|
+
optimizer.zero_grad()
|
|
114
|
+
|
|
115
|
+
model.hidden = (
|
|
116
|
+
torch.zeros(1, 1, model.hidden_size),
|
|
117
|
+
torch.zeros(1, 1, model.hidden_size),
|
|
118
|
+
|
|
119
|
+
)
|
|
120
|
+
|
|
121
|
+
train_predicred_label = model.forward(train_window)
|
|
122
|
+
train_loss = criterion(train_predicted_label, correct_label)
|
|
123
|
+
|
|
124
|
+
train_loss.backward()
|
|
125
|
+
optimizer.step()
|
|
126
|
+
|
|
127
|
+
train_losses.append(train_loss)
|
|
128
|
+
↓
|
|
129
|
+
a = torch.tensor([3])
|
|
130
|
+
a.item()
|
|
131
|
+
↓
|
|
132
|
+
def run_test():
|
|
133
|
+
model.eval()
|
|
134
|
+
|
|
135
|
+
for i in range(test_size):
|
|
136
|
+
|
|
137
|
+
test_window = torch.FloatTensor(extending_seq[-test_size:])
|
|
138
|
+
|
|
139
|
+
with torch.no_grad():
|
|
140
|
+
|
|
141
|
+
model.hidden = (
|
|
142
|
+
torch.zeros(1, 1, model.hidden_size),
|
|
143
|
+
torch.zeros(1, 1, model.hidden_size),
|
|
144
|
+
|
|
145
|
+
)
|
|
146
|
+
|
|
147
|
+
test_predicted_label = mode.forward(test_window)
|
|
148
|
+
extending_seq.append(test_predicted_label())
|
|
149
|
+
|
|
150
|
+
test_loss = criterion(
|
|
151
|
+
torch.FloatTensor(extending_seq[-test_size:]),
|
|
152
|
+
y[len(y)-test_size:]
|
|
153
|
+
)
|
|
154
|
+
|
|
155
|
+
test_losses.append(test_loss)
|
|
156
|
+
↓
|
|
157
|
+
train_seq[-test_size:]
|
|
158
|
+
↓
|
|
159
|
+
train_seq[-test_size:].tolist()
|
|
160
|
+
↓
|
|
7
161
|
for epoch in range(epochs):
|
|
162
|
+
|
|
8
163
|
print()
|
|
9
164
|
print(f'Epoch: {epoch+1}')
|
|
10
165
|
|
|
11
|
-
run_train()
|
|
166
|
+
run_train()
|
|
12
167
|
|
|
13
|
-
extending_seq = train_seq[-test_size:].
|
|
168
|
+
extending_seq = train_seq[-test_size:].tolist()
|
|
14
169
|
|
|
15
|
-
run_test()
|
|
170
|
+
run_test()
|
|
16
171
|
|
|
17
|
-
plt.figure(figsize=(12, 4))
|
|
172
|
+
plt.figure(figsize=(12, 4))
|
|
18
|
-
plt.xlim(-
|
|
173
|
+
plt.xlim(-21, len(y)+20)
|
|
19
|
-
plt.grid(True)
|
|
174
|
+
plt.grid(True)
|
|
20
175
|
|
|
21
|
-
plt.plot(y.numpy())
|
|
22
|
-
|
|
23
|
-
plt.plot(
|
|
176
|
+
plt.plot(
|
|
24
177
|
range(len(y)-test_size, len(y)),
|
|
25
178
|
extending_seq[-test_size:]
|
|
26
179
|
)
|
|
27
180
|
|
|
28
|
-
plt.show()
|
|
181
|
+
plt.show()
|
|
29
|
-
|
|
182
|
+
↓
|
|
30
|
-
↓エラー
|
|
183
|
+
↓エラー内容
|
|
31
|
-
```error
|
|
32
|
-
NameError Traceback (most recent call last)
|
|
33
|
-
<ipython-input-11-24515b2a1814> in <module>()
|
|
34
|
-
----> 1 for epoch in range(epochs):
|
|
35
|
-
2 print()
|
|
36
|
-
3 print(f'Epoch: {epoch+1}')
|
|
37
|
-
|
|
184
|
+
↓
|
|
38
|
-
5 run_train()
|
|
39
185
|
|
|
186
|
+
Epoch: 1
|
|
187
|
+
---------------------------------------------------------------------------
|
|
188
|
+
AttributeError Traceback (most recent call last)
|
|
189
|
+
<ipython-input-85-98e8d85b09e9> in <module>()
|
|
40
|
-
|
|
190
|
+
4 print(f'Epoch: {epoch+1}')
|
|
191
|
+
5
|
|
192
|
+
----> 6 run_train()
|
|
193
|
+
7
|
|
194
|
+
8 extending_seq = train_seq[-test_size:].tolist()
|
|
195
|
+
|
|
196
|
+
1 frames
|
|
197
|
+
<ipython-input-77-b3af88443259> in forward(self, seq)
|
|
198
|
+
18
|
|
199
|
+
19 out, _=self.lstm(
|
|
200
|
+
---> 20 seq.vierq(len(seq), 1, -1),
|
|
201
|
+
21 self.hidden
|
|
202
|
+
22 )
|
|
203
|
+
|
|
204
|
+
AttributeError: 'Tensor' object has no attribute 'vierq'
|
|
205
|
+
コード
|
|
41
206
|
```
|
2
変更点追加
title
CHANGED
|
@@ -1,1 +1,1 @@
|
|
|
1
|
-
|
|
1
|
+
NameError: name 'epochs' is not defined python エラーコード
|
body
CHANGED
|
File without changes
|
1
文法修正
title
CHANGED
|
@@ -1,1 +1,1 @@
|
|
|
1
|
-
株
|
|
1
|
+
python 株分析 練習 エラーコード
|
body
CHANGED
|
File without changes
|