質問編集履歴
2
現在の状況
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -1,8 +1,8 @@
|
|
|
1
1
|
###前提・実現したいこと
|
|
2
2
|
|
|
3
|
-
|
|
3
|
+
Digtisデータで混合行列を作成するコードようにmnistデータでも同じように混合行列を作成したいのですがどのようにしたらいいですか。
|
|
4
|
-
|
|
5
|
-
|
|
4
|
+
|
|
5
|
+
|
|
6
6
|
|
|
7
7
|
原因と改善策を教えてください。
|
|
8
8
|
|
1
書式な改善と現在のコードの状況
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -8,6 +8,12 @@
|
|
|
8
8
|
|
|
9
9
|
|
|
10
10
|
|
|
11
|
+
また、Digtisデータで混同行列は以下のようにできたのに今回はエラーが出るのでしょうか。
|
|
12
|
+
|
|
13
|
+
|
|
14
|
+
|
|
15
|
+
|
|
16
|
+
|
|
11
17
|
###発生している問題・エラーメッセージ
|
|
12
18
|
|
|
13
19
|
|
|
@@ -18,41 +24,41 @@
|
|
|
18
24
|
|
|
19
25
|
ValueError Traceback (most recent call last)
|
|
20
26
|
|
|
21
|
-
<ipython-input-5-8
|
|
27
|
+
<ipython-input-5-f6ad643389d7> in <module>()
|
|
22
|
-
|
|
28
|
+
|
|
23
|
-
2
|
|
29
|
+
29
|
|
24
|
-
|
|
30
|
+
|
|
25
|
-
|
|
31
|
+
30 if __name__ == '__main__':
|
|
26
|
-
|
|
32
|
+
|
|
27
|
-
---> 3
|
|
33
|
+
---> 31 main()
|
|
28
|
-
|
|
29
|
-
|
|
30
|
-
|
|
34
|
+
|
|
35
|
+
|
|
36
|
+
|
|
31
|
-
<ipython-input-5-8
|
|
37
|
+
<ipython-input-5-f6ad643389d7> in main()
|
|
32
|
-
|
|
38
|
+
|
|
33
|
-
2
|
|
39
|
+
21
|
|
34
|
-
|
|
40
|
+
|
|
35
|
-
2
|
|
41
|
+
22 # 正解率を計算
|
|
36
|
-
|
|
42
|
+
|
|
37
|
-
---> 2
|
|
43
|
+
---> 23 score = accuracy_score(test_dataY, predicted_labels)
|
|
38
|
-
|
|
44
|
+
|
|
39
|
-
2
|
|
45
|
+
24 print("正解率:{}".format(score))
|
|
40
|
-
|
|
46
|
+
|
|
41
|
-
2
|
|
47
|
+
25
|
|
42
|
-
|
|
43
|
-
|
|
44
|
-
|
|
48
|
+
|
|
49
|
+
|
|
50
|
+
|
|
45
|
-
~\Anaconda3\lib\site-packages\sklearn\metrics\classification.py in c
|
|
51
|
+
~\Anaconda3\lib\site-packages\sklearn\metrics\classification.py in accuracy_score(y_true, y_pred, normalize, sample_weight)
|
|
46
|
-
|
|
52
|
+
|
|
47
|
-
|
|
53
|
+
174
|
|
48
|
-
|
|
54
|
+
|
|
49
|
-
|
|
55
|
+
175 # Compute accuracy for each possible representation
|
|
50
|
-
|
|
56
|
+
|
|
51
|
-
-->
|
|
57
|
+
--> 176 y_type, y_true, y_pred = _check_targets(y_true, y_pred)
|
|
52
|
-
|
|
58
|
+
|
|
53
|
-
|
|
59
|
+
177 if y_type.startswith('multilabel'):
|
|
54
|
-
|
|
60
|
+
|
|
55
|
-
|
|
61
|
+
178 differing_labels = count_nonzero(y_true - y_pred, axis=1)
|
|
56
62
|
|
|
57
63
|
|
|
58
64
|
|
|
@@ -84,18 +90,92 @@
|
|
|
84
90
|
|
|
85
91
|
|
|
86
92
|
|
|
87
|
-
ValueError: Found input variables with inconsistent numbers of samples: [21000,
|
|
93
|
+
ValueError: Found input variables with inconsistent numbers of samples: [21000, 1]
|
|
88
94
|
|
|
89
95
|
|
|
90
96
|
|
|
91
97
|
```
|
|
92
98
|
|
|
93
|
-
|
|
94
|
-
|
|
95
|
-
###
|
|
99
|
+
###Digitsデータでのソースコード
|
|
96
100
|
|
|
97
101
|
```python
|
|
98
102
|
|
|
103
|
+
from sklearn import datasets
|
|
104
|
+
|
|
105
|
+
from sklearn.model_selection import LeaveOneOut
|
|
106
|
+
|
|
107
|
+
from sklearn.metrics import accuracy_score, confusion_matrix
|
|
108
|
+
|
|
109
|
+
from sklearn.neighbors import KNeighborsClassifier
|
|
110
|
+
|
|
111
|
+
|
|
112
|
+
|
|
113
|
+
def main():
|
|
114
|
+
|
|
115
|
+
dataset = datasets.load_digits()
|
|
116
|
+
|
|
117
|
+
|
|
118
|
+
|
|
119
|
+
features = dataset.data
|
|
120
|
+
|
|
121
|
+
targets = dataset.target
|
|
122
|
+
|
|
123
|
+
|
|
124
|
+
|
|
125
|
+
predicted_labels = []
|
|
126
|
+
|
|
127
|
+
|
|
128
|
+
|
|
129
|
+
loo = LeaveOneOut()
|
|
130
|
+
|
|
131
|
+
for train, test in loo.split(features):
|
|
132
|
+
|
|
133
|
+
train_data = features[train]
|
|
134
|
+
|
|
135
|
+
target_data = targets[train]
|
|
136
|
+
|
|
137
|
+
|
|
138
|
+
|
|
139
|
+
model = KNeighborsClassifier(n_neighbors=1, metric='euclidean')
|
|
140
|
+
|
|
141
|
+
model.fit(train_data, target_data)
|
|
142
|
+
|
|
143
|
+
|
|
144
|
+
|
|
145
|
+
predicted_label = model.predict(features[test])
|
|
146
|
+
|
|
147
|
+
predicted_labels.append(predicted_label)
|
|
148
|
+
|
|
149
|
+
|
|
150
|
+
|
|
151
|
+
print(predicted_labels)
|
|
152
|
+
|
|
153
|
+
score = accuracy_score(targets, predicted_labels)
|
|
154
|
+
|
|
155
|
+
print(score)
|
|
156
|
+
|
|
157
|
+
|
|
158
|
+
|
|
159
|
+
# 混合行列を表示
|
|
160
|
+
|
|
161
|
+
cm = confusion_matrix(targets, predicted_labels)
|
|
162
|
+
|
|
163
|
+
print(cm)
|
|
164
|
+
|
|
165
|
+
|
|
166
|
+
|
|
167
|
+
|
|
168
|
+
|
|
169
|
+
if __name__ == '__main__':
|
|
170
|
+
|
|
171
|
+
main()
|
|
172
|
+
|
|
173
|
+
```
|
|
174
|
+
|
|
175
|
+
###mnistデータでのソースコード
|
|
176
|
+
|
|
177
|
+
```python
|
|
178
|
+
|
|
99
179
|
from collections import Counter
|
|
100
180
|
|
|
101
181
|
from matplotlib import pyplot as plt
|
|
@@ -134,16 +214,12 @@
|
|
|
134
214
|
|
|
135
215
|
|
|
136
216
|
|
|
217
|
+
#データを分割
|
|
218
|
+
|
|
137
219
|
train_dataX, test_dataX, train_dataY, test_dataY = model_selection.train_test_split(features,targets,test_size=0.3)
|
|
138
220
|
|
|
139
221
|
|
|
140
222
|
|
|
141
|
-
|
|
142
|
-
|
|
143
|
-
# 使う近傍数ごとに正解率＆各経過時間を計算
|
|
144
|
-
|
|
145
|
-
accuracy_scores = []
|
|
146
|
-
|
|
147
223
|
predicted_labels = []
|
|
148
224
|
|
|
149
225
|
|
|
@@ -160,15 +236,21 @@
|
|
|
160
236
|
|
|
161
237
|
predicted_label = model.predict(test_dataX)
|
|
162
238
|
|
|
239
|
+
predicted_labels.append(predicted_label)
|
|
240
|
+
|
|
241
|
+
|
|
242
|
+
|
|
243
|
+
## print(predicted_labels)
|
|
244
|
+
|
|
163
245
|
|
|
164
246
|
|
|
165
247
|
# 正解率を計算
|
|
166
248
|
|
|
167
|
-
score = accuracy_score(test_dataY, predicted_label)
|
|
249
|
+
score = accuracy_score(test_dataY, predicted_labels)
|
|
168
250
|
|
|
169
251
|
print("正解率:{}".format(score))
|
|
170
252
|
|
|
171
|
-
|
|
253
|
+
|
|
172
254
|
|
|
173
255
|
# 混合行列を表示
|
|
174
256
|
|
|
@@ -176,7 +258,7 @@
|
|
|
176
258
|
|
|
177
259
|
print(cm)
|
|
178
260
|
|
|
179
|
-
|
|
261
|
+
|
|
180
262
|
|
|
181
263
|
if __name__ == '__main__':
|
|
182
264
|
|