matlab矩阵的一个小问题

夏夜 · 发表于 2014-8-23 19:36:53

登录后查看更多精彩内容~

您需要登录才可以下载或查看，没有帐号？立即注册

x

例如这样一个矩阵：a=[0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 6 6.3 1.1 1.2 9 1.4 1.5 1.6 10 15 1.9 2];
如何找出异常值6 6.3 9 10 15？
比如find(a(i)>3)可以找出来，但是在其他数据中3这个判断值不好确定。
怎么根据前后两个值的差来判别？
根据数据的变化趋势，把这几个异常值插值为合理的值，但是现在还没找出这些值。
有兴趣的来看看。
谢谢各位。

二爷名声在外 · 发表于 2014-8-23 20:06:12

没太看懂楼主的意思，我举个例子楼主看合适不合适，就是find那些值，然后比如用interp插值代替那些值，楼主看看是不是这种情况

夏夜 · 发表于 2014-8-23 20:35:13

二爷名声在外发表于 2014-8-23 20:06
没太看懂楼主的意思，我举个例子楼主看合适不合适，就是find那些值，然后比如用interp插值代替那些值，楼主 ...

谢谢回复。就是find那些值不好找，比如一组数半部分可以取>3 后半部分又取>7来判断，就是这个判别值标准不一样。能不能根据前后附近的几个数来判定

二爷名声在外 · 发表于 2014-8-23 20:54:03

夏夜发表于 2014-8-23 20:35
谢谢回复。就是find那些值不好找，比如一组数半部分可以取>3 后半部分又取>7来判断，就是这个判别值标准 ...

这组数有规律吗

夏夜 · 发表于 2014-8-24 09:21:53

二爷名声在外发表于 2014-8-23 20:54
这组数有规律吗

比如其中一组像这样的：

187.409
188.216
188.522
188.936
189.739
191.319
194.335
196.793
198.816
200.165
201.982
204.803
205.652
206.452
206.986
207.925
209.363
211.374
211.96
212.572
213.669
215.439
216.27
217.161
217.026
217.486
218.497
218.698
219.126
218.867
218.606
217.656
217.39
217.438
217.454
216.955
216.571
216.416
216.657
216.479
216.128
216.26
216.11
215.968
215.471
214.837
213.958
212.555
211.225
210.497
209.795
209.495
208.974
208.506
207.608
207.04
206.067
205.735
204.837
204.364
203.848
202.989
202.342
201.427
201.126
200.728
200.771
200.577
200.304
200.147
200.197
200.053
200.164
200.002
199.847
199.796
199.679
199.854
199.732
199.594
199.524
199.069
199.041
198.867
198.67
198.799
198.712
198.435
198.561
198.47
198.28
197.969
198.013
197.987
198.161
198.017
198.475
198.499
198.464
198.497
198.669
198.396
198.409
198.421
197.997
183.269
197.5
197.013
197.05
196.902
196.663
196.327
196.125
195.972
195.943
195.885
195.781
195.576
416.016
194.498
1520.34
193.655
193.197
192.726
191.948
191.356
190.792
189.977
189.315
189.161
188.941
188.827
188.661
188.374
188.075
187.671
187.254
186.817
186.422
185.837
184.745
183.688
183.504
183.331
1368.12
183.495
183.347
183.328
183.189
182.91
182.711
182.312
181.833
181.202
180.614
180.522
180.156
179.644
179.25
179.374
179.408
179.258
179.308
179.579
179.82
179.841
179.905
180.082
179.943
179.851
179.686
179.385
179.277
179.166
179.132
178.958
178.62
178.156
177.906
177.605
177.444
177.299
177.321
177.531
177.285
176.939
176.608
176.319
176.274
176.095
175.927
175.772
175.492
175.108
174.779
174.539
174.259
174.431
173.905
173.054
172.936
172.884
172.67
172.613
172.47
172.326
172.183
172.369
172.532
172.635
172.575
172.428
172.317
172.243
172.173
172.078
172.064
172.282
172.309
172.121
172.022
172.148
172.338
172.395
172.47
172.292
172.047
171.667
171.254
170.793
170.333
169.823
169.268
168.552
168.104
168.548
168.642
168.785
168.623
168.649
168.396
168.163
167.988
167.902
167.717
167.544
167.352
167.146
166.788
166.362
166.132
165.792
165.533
165.375
165.036
164.731
164.523
164.262
164.18
163.731
162.889
162.234
161.892
161.568
161.176
161.063
160.978
160.566
160.356
160.205
160.485
160.652
160.963
161.249
161.452
160.664
159.761
158.922
158.07
156.227
155.956
155.097
153.998
152.501
150.652
150.311
149.61
458.937
147.192
145.854
144.971
143.996
142.545
141.179
140.53
139.593
138.64
137.447
136.331
135.424
133.313
128.384
126.02
123.674
120.563
117.835
113.878
109.756
105.791
101.123
97.4025
93.691
90.7181
89.4139
86.6175
83.9813
81.0088
78.2499
76.1636
74.7982
73.226
72.1281
71.1975
69.7793
68.3262
66.5006
64.3221
62.6416
61.4448
60.5883
59.5077
58.2773
56.6985
55.5485
54.8744
53.8494
53.5269
53.4027
52.8515
51.8756
51.5619

与前后几个相比变化比较异常的值

二爷名声在外 · 发表于 2014-8-24 09:31:48

夏夜发表于 2014-8-24 09:21
比如其中一组像这样的：

亲，你这数据能不能横着展示一下…完全就看不清的样子…

飞云侠客 · 发表于 2014-8-24 09:32:01

楼主问的不是Matlab编程问题，是一个统计诊断问题，先定义你的异常，或称离群的指标，根据这个指标确定异常值，剔除即可。参见统计诊断的书。

夏夜 · 发表于 2014-8-24 09:51:11

二爷名声在外发表于 2014-8-24 09:31
亲，你这数据能不能横着展示一下…完全就看不清的样子…

刚才那组有误，我忘了把其中有几个负号加上了。
187.409000000000 188.216000000000 188.522000000000 188.936000000000 189.739000000000 191.319000000000 194.335000000000 196.793000000000 198.816000000000 200.165000000000 201.982000000000 204.803000000000 205.652000000000 206.452000000000 206.986000000000 207.925000000000 209.363000000000 211.374000000000 211.960000000000 212.572000000000 213.669000000000 215.439000000000 216.270000000000 217.161000000000 217.026000000000 217.486000000000 218.497000000000 218.698000000000 219.126000000000 218.867000000000 218.606000000000 217.656000000000 217.390000000000 217.438000000000 217.454000000000 216.955000000000 216.571000000000 216.416000000000 216.657000000000 216.479000000000 216.128000000000 216.260000000000 216.110000000000 215.968000000000 215.471000000000 214.837000000000 213.958000000000 212.555000000000 211.225000000000 210.497000000000 209.795000000000 209.495000000000 208.974000000000 208.506000000000 207.608000000000 207.040000000000 206.067000000000 205.735000000000 204.837000000000 204.364000000000 203.848000000000 202.989000000000 202.342000000000 201.427000000000 201.126000000000 200.728000000000 200.771000000000 200.577000000000 200.304000000000 200.147000000000 200.197000000000 200.053000000000 200.164000000000 200.002000000000 199.847000000000 199.796000000000 199.679000000000 199.854000000000 199.732000000000 199.594000000000 199.524000000000 199.069000000000 199.041000000000 198.867000000000 198.670000000000 198.799000000000 198.712000000000 198.435000000000 198.561000000000 198.470000000000 198.280000000000 197.969000000000 198.013000000000 197.987000000000 198.161000000000 198.017000000000 198.475000000000 198.499000000000 198.464000000000 198.497000000000 198.669000000000 198.396000000000 198.409000000000 198.421000000000 197.997000000000 -183.269000000000 197.500000000000 197.013000000000 197.050000000000 196.902000000000 196.663000000000 196.327000000000 196.125000000000 195.972000000000 195.943000000000 195.885000000000 195.781000000000 195.576000000000 -416.016000000000 194.498000000000 1520.34000000000 193.655000000000 193.197000000000 192.726000000000 191.948000000000 191.356000000000 190.792000000000 189.977000000000 189.315000000000 189.161000000000 188.941000000000 188.827000000000 188.661000000000 188.374000000000 188.075000000000 187.671000000000 187.254000000000 186.817000000000 186.422000000000 185.837000000000 184.745000000000 183.688000000000 183.504000000000 183.331000000000 1368.12000000000 183.495000000000 183.347000000000 183.328000000000 183.189000000000 182.910000000000 182.711000000000 182.312000000000 181.833000000000 181.202000000000 180.614000000000 180.522000000000 180.156000000000 179.644000000000 179.250000000000 179.374000000000 179.408000000000 179.258000000000 179.308000000000 179.579000000000 179.820000000000 179.841000000000 179.905000000000 180.082000000000 179.943000000000 179.851000000000 179.686000000000 179.385000000000 179.277000000000 179.166000000000 179.132000000000 178.958000000000 178.620000000000 178.156000000000 177.906000000000 177.605000000000 177.444000000000 177.299000000000 177.321000000000 177.531000000000 177.285000000000 176.939000000000 176.608000000000 176.319000000000 176.274000000000 176.095000000000 175.927000000000 175.772000000000 175.492000000000 175.108000000000 174.779000000000 174.539000000000 174.259000000000 174.431000000000 173.905000000000 173.054000000000 172.936000000000 172.884000000000 172.670000000000 172.613000000000 172.470000000000 172.326000000000 172.183000000000 172.369000000000 172.532000000000 172.635000000000 172.575000000000 172.428000000000 172.317000000000 172.243000000000 172.173000000000 172.078000000000 172.064000000000 172.282000000000 172.309000000000 172.121000000000 172.022000000000 172.148000000000 172.338000000000 172.395000000000 172.470000000000 172.292000000000 172.047000000000 171.667000000000 171.254000000000 170.793000000000 170.333000000000 169.823000000000 169.268000000000 168.552000000000 168.104000000000 168.548000000000 168.642000000000 168.785000000000 168.623000000000 168.649000000000 168.396000000000 168.163000000000 167.988000000000 167.902000000000 167.717000000000 167.544000000000 167.352000000000 167.146000000000 166.788000000000 166.362000000000 166.132000000000 165.792000000000 165.533000000000 165.375000000000 165.036000000000 164.731000000000 164.523000000000 164.262000000000 164.180000000000 163.731000000000 162.889000000000 162.234000000000 161.892000000000 161.568000000000 161.176000000000 161.063000000000 160.978000000000 160.566000000000 160.356000000000 160.205000000000 160.485000000000 160.652000000000 160.963000000000 161.249000000000 161.452000000000 160.664000000000 159.761000000000 158.922000000000 158.070000000000 156.227000000000 155.956000000000 155.097000000000 153.998000000000 152.501000000000 150.652000000000 150.311000000000 149.610000000000 458.937000000000 147.192000000000 145.854000000000 144.971000000000 143.996000000000 142.545000000000 141.179000000000 140.530000000000 139.593000000000 138.640000000000 137.447000000000 136.331000000000 135.424000000000 133.313000000000 128.384000000000 126.020000000000 123.674000000000 120.563000000000 117.835000000000 113.878000000000 109.756000000000 105.791000000000 101.123000000000 97.4025000000000 93.6910000000000 90.7181000000000 89.4139000000000 86.6175000000000 83.9813000000000 81.0088000000000 78.2499000000000 76.1636000000000 74.7982000000000 73.2260000000000 72.1281000000000 71.1975000000000 69.7793000000000 68.3262000000000 66.5006000000000 64.3221000000000 62.6416000000000 61.4448000000000 60.5883000000000 59.5077000000000 58.2773000000000 56.6985000000000 55.5485000000000 54.8744000000000 53.8494000000000 53.5269000000000 53.4027000000000 52.8515000000000 51.8756000000000 51.5619000000000

夏夜 · 发表于 2014-8-24 09:52:46

飞云侠客发表于 2014-8-24 09:32
楼主问的不是Matlab编程问题，是一个统计诊断问题，先定义你的异常，或称离群的指标，根据这个指标确定异常 ...

嗯就这个异常不知道怎么定义，它是随机产生的。。

二爷名声在外 · 发表于 2014-8-24 09:53:33

夏夜发表于 2014-8-24 09:51
刚才那组有误，我忘了把其中有几个负号加上了。
187.409000000000 188.216000000000 188.522000000000 1 ...

请楼主参考7楼的观点…

		自动登录	找回密码
密码			立即注册

[程序设计] matlab矩阵的一个小问题

登录后查看更多精彩内容~