Closed FinnGu closed 5 years ago
Mahelita
commented
5 years ago Hi Finn, it sounds like your data is terribly unbalanced. Are you using the BalancedBatchSizeIterator (https://github.com/robintibor/braindecode/blob/master/braindecode/datautil/iterators.py)?
FinnGu
commented
5 years ago Hi Finn, it sounds like your data is terribly unbalanced. Are you using the BalancedBatchSizeIterator (https://github.com/robintibor/braindecode/blob/master/braindecode/datautil/iterators.py)?
No, I don't use it yet. Will try that, thank you for the suggestion!
What exactly do you mean by unbalanced? My training set contains class0 103 times, class1 101 times and class2 102 times, so I would not say it is terribly unbalanced.
Mahelita
commented
5 years ago My training set contains class0 103 times, class1 101 times and class2 102 times
That is quite balanced but do still use the balanced batch iterator to be sure to have balanced batches. Something else then goes wrong during the training... Can you please post your learning curves?
robintibor
commented
5 years ago Hi @FinnGu , maybe try to see if a rough normalization happens, so if you only subtract mean and divide by standard deviation. definitely training loss stagnating at a high level is indicative of a problem. What value range is your data in? How large are your trials?
FinnGu
commented
5 years ago What value range is your data in?
Minimum (calculated for each channel separately) is [-1964.543091 -2586.350342 -7390.82373], maximum (calculated for each channel separately) is [2672.656738 3156.856689 4913.586426].
After subtracting each channel's mean and dividing by each channel's SD: [-5.13000551 -5.27367852 -3.78299747] to [8.25487205 6.12262174 4.68192221].
Doing this has improved train_misclass to 0.33!!! (Sadly valid_misclass is still at 0.63) When subtracting the mean, am I supposed to add it to the negative numbers? Otherwise I don't really understand the point of shifting all the values more towards the negative.
How large are your trials?
For each subject I have 360 trials in total, 120 per class. Motor Imagery was performed for 4 seconds per trial at 256Hz, resulting in 1024 samples per trial. -> My trials numpy array is of shape (360, 3, 1024).
Mahelita
commented
5 years ago When subtracting the mean, am I supposed to add it to the negative numbers? Otherwise I don't really understand the point of shifting all the values more towards the negative.
You still subtract. This is supposed to center the distribution of your data around 0. If the original mean was positive then your positive values will get smaller (less positive) and your negative values larger (more negative). Vice versa for a negative original mean. Does this make sense to you?
robintibor
commented
5 years ago train misclass is still quite high. what is train loss? can you share picture of learning curves of train loss, test loss, train misclass and test misclass?
FinnGu
commented
5 years ago This is supposed to center the distribution of your data around 0.
@Mahelita That does make sense, thank you!
@robintibor I just let it train for a couple more epochs and the train_loss and train_misclass seem to be improving further. Only problem is, I am probably overfitting. I only have the dataframe right now, will update my answer with the curve in a couple minutes.Learning Logs
epoch train_loss valid_loss train_misclass valid_misclass runtime
0 8.759408 8.379377 0.660000 0.700000 0.000000
1 1.676288 1.651536 0.660000 0.683333 0.748986
2 1.120547 1.199027 0.633333 0.650000 0.694491
3 1.096075 1.179758 0.586667 0.650000 0.687491
4 1.085830 1.176387 0.550000 0.666667 0.697491
5 1.080550 1.152808 0.530000 0.633333 0.693491
6 1.079630 1.201502 0.570000 0.666667 0.679500
7 1.072987 1.156153 0.536667 0.650000 0.679483
8 1.071331 1.133579 0.526667 0.666667 0.688491
9 1.069778 1.159460 0.533333 0.650000 0.687991
10 1.067073 1.159788 0.530000 0.650000 0.681713
11 1.064620 1.135865 0.526667 0.650000 0.684131
12 1.063511 1.141347 0.516667 0.666667 0.688991
13 1.062133 1.129188 0.526667 0.666667 0.675492
14 1.061706 1.173878 0.543333 0.716667 0.688991
15 1.058642 1.158047 0.516667 0.700000 0.711492
16 1.058741 1.152954 0.510000 0.700000 0.690990
17 1.056560 1.168133 0.520000 0.700000 0.679169
18 1.058284 1.180714 0.550000 0.683333 0.688784
19 1.055806 1.192691 0.536667 0.716667 0.674492
20 1.054392 1.200964 0.513333 0.666667 0.679451
21 1.053406 1.179440 0.536667 0.700000 0.676968
22 1.051842 1.221439 0.513333 0.666667 0.698991
23 1.050823 1.234113 0.526667 0.683333 0.682492
24 1.050869 1.209331 0.520000 0.683333 0.677948
25 1.049947 1.230951 0.530000 0.666667 0.685992
26 1.047866 1.241897 0.503333 0.666667 0.683463
27 1.046801 1.254566 0.480000 0.683333 0.685491
28 1.048608 1.281728 0.513333 0.650000 0.681991
29 1.046148 1.241690 0.486667 0.716667 0.682492
30 1.047313 1.327724 0.513333 0.683333 0.682991
31 1.043745 1.267750 0.516667 0.683333 0.686167
32 1.043702 1.279967 0.483333 0.733333 0.682491
33 1.045281 1.296920 0.513333 0.700000 0.681993
34 1.042893 1.363529 0.500000 0.683333 0.684490
35 1.039390 1.303987 0.510000 0.733333 0.683555
36 1.039369 1.316590 0.493333 0.733333 0.678993
37 1.036589 1.338337 0.460000 0.700000 0.702989
38 1.036380 1.332471 0.523333 0.716667 0.686991
39 1.040667 1.315642 0.490000 0.716667 0.677992
40 1.034237 1.360488 0.473333 0.700000 0.679770
41 1.032669 1.366279 0.456667 0.733333 0.683991
42 1.032604 1.369405 0.466667 0.733333 0.683491
43 1.032288 1.383133 0.490000 0.750000 0.686991
44 1.030290 1.367392 0.450000 0.716667 0.680991
45 1.029186 1.398806 0.480000 0.750000 0.678991
46 1.028223 1.384746 0.466667 0.733333 0.680992
47 1.029537 1.427768 0.476667 0.733333 0.684492
48 1.025990 1.420856 0.486667 0.750000 0.677991
49 1.027008 1.429379 0.476667 0.750000 0.685992
50 1.026946 1.447039 0.510000 0.750000 0.686491
51 1.033051 1.397873 0.490000 0.733333 0.719991
52 1.025109 1.414541 0.443333 0.716667 0.780490
53 1.021229 1.384627 0.450000 0.750000 0.752491
54 1.023964 1.478723 0.460000 0.716667 0.732991
55 1.020282 1.410171 0.463333 0.733333 0.700992
56 1.024750 1.458286 0.490000 0.750000 0.688501
57 1.017905 1.419961 0.463333 0.733333 0.720249
58 1.016892 1.481050 0.476667 0.716667 0.728991
59 1.023636 1.424646 0.506667 0.750000 0.779490
60 1.017487 1.455956 0.423333 0.733333 0.829490
61 1.014891 1.495187 0.440000 0.766667 0.801869
62 1.016693 1.446011 0.456667 0.733333 0.705824
63 1.014672 1.426771 0.436667 0.733333 0.697988
64 1.023078 1.529745 0.483333 0.716667 0.721491
65 1.015009 1.496703 0.470000 0.766667 0.724991
66 1.014098 1.474905 0.430000 0.766667 0.699818
67 1.013281 1.499792 0.480000 0.750000 0.700555
68 1.013463 1.523083 0.440000 0.750000 0.731490
69 1.011478 1.462903 0.430000 0.750000 0.706862
70 1.010717 1.568936 0.436667 0.733333 0.694631
71 1.013556 1.474226 0.440000 0.750000 0.683940
72 1.004509 1.521454 0.410000 0.716667 0.699991
73 1.005344 1.578061 0.463333 0.766667 0.726989
74 1.004596 1.560611 0.450000 0.733333 0.704421
75 1.005036 1.463143 0.433333 0.766667 0.725990
76 1.003313 1.538439 0.440000 0.733333 0.711994
77 1.002882 1.529980 0.490000 0.750000 0.764489
78 0.999347 1.501390 0.423333 0.733333 0.703991
79 0.996739 1.588695 0.376667 0.750000 0.713683
80 0.998507 1.480397 0.423333 0.750000 0.689653
81 0.999519 1.551335 0.450000 0.750000 0.734491
82 0.997489 1.525754 0.410000 0.733333 0.739993
83 0.993588 1.581290 0.430000 0.733333 0.702615
84 0.991971 1.526525 0.410000 0.733333 0.712957
85 0.991055 1.555757 0.423333 0.783333 0.683857
86 0.991381 1.536540 0.393333 0.700000 0.713155
87 0.990067 1.556044 0.456667 0.750000 0.705493
88 0.987036 1.508045 0.403333 0.666667 0.701035
89 0.986350 1.585934 0.410000 0.750000 0.716728
90 0.984746 1.549676 0.383333 0.716667 0.697492
91 0.987960 1.593336 0.413333 0.733333 0.715490
92 0.993053 1.584121 0.423333 0.733333 0.713491
93 0.996731 1.579435 0.466667 0.733333 0.708000
94 0.990399 1.620260 0.410000 0.766667 0.720317
95 0.982617 1.540713 0.386667 0.683333 0.708491
96 0.980871 1.639990 0.413333 0.750000 0.727457
97 0.985675 1.573103 0.423333 0.733333 0.732490
98 0.990504 1.668333 0.436667 0.783333 0.775990
99 0.978049 1.622453 0.386667 0.733333 0.720786
100 0.974921 1.651500 0.396667 0.750000 0.745101
101 0.976223 1.627341 0.360000 0.733333 0.719990
102 0.977436 1.646850 0.386667 0.783333 0.719492
103 0.989102 1.593368 0.400000 0.716667 0.716491
104 0.980613 1.599945 0.426667 0.783333 0.712491
105 0.975941 1.600077 0.400000 0.716667 0.736492
106 0.971375 1.620329 0.360000 0.716667 0.695491
107 0.970571 1.549663 0.366667 0.750000 0.711491
108 0.976968 1.630356 0.433333 0.750000 0.702991
109 0.970233 1.586010 0.376667 0.750000 0.706491
110 0.977669 1.544223 0.410000 0.716667 0.707491
111 0.972783 1.677697 0.390000 0.733333 0.704492
112 0.971833 1.586840 0.366667 0.733333 0.696821
113 0.975352 1.618620 0.403333 0.750000 0.704221
114 0.975459 1.714128 0.433333 0.733333 0.705991
115 0.963632 1.651544 0.366667 0.716667 0.696991
116 0.964304 1.623365 0.386667 0.733333 0.704991
117 0.966690 1.683713 0.393333 0.716667 0.716491
118 0.964302 1.665059 0.383333 0.750000 0.702491
119 0.965679 1.604004 0.363333 0.700000 0.702491
120 0.967586 1.689639 0.386667 0.716667 0.709991
121 0.964723 1.600946 0.356667 0.716667 0.708991
122 0.959935 1.627681 0.396667 0.716667 0.703992
123 0.961891 1.674775 0.336667 0.750000 0.705491
124 0.981168 1.730132 0.466667 0.716667 0.718991
125 0.961168 1.630492 0.350000 0.666667 0.711550
126 0.957527 1.661504 0.356667 0.716667 0.702991
127 0.959861 1.692457 0.366667 0.716667 0.706491
128 0.956655 1.717450 0.376667 0.700000 0.703490
129 0.953204 1.703787 0.346667 0.716667 0.720993
130 0.953336 1.593744 0.363333 0.683333 0.701992
131 0.956186 1.678114 0.333333 0.750000 0.743854
132 0.958107 1.654596 0.353333 0.700000 0.701990
133 0.952334 1.680992 0.386667 0.716667 0.704491
134 0.949674 1.641915 0.356667 0.716667 0.706490
135 0.956540 1.668859 0.353333 0.716667 0.709086
136 0.960456 1.635219 0.370000 0.700000 0.725993
137 0.951328 1.695293 0.350000 0.716667 0.720489
138 0.949198 1.662383 0.306667 0.716667 0.718491
139 0.953648 1.683113 0.380000 0.716667 0.707991
140 0.949129 1.693088 0.330000 0.700000 0.719992
141 0.947668 1.677134 0.370000 0.716667 0.689990
142 0.944520 1.638288 0.336667 0.716667 0.703491
143 0.947478 1.729360 0.383333 0.733333 0.699992
144 0.948553 1.604219 0.350000 0.716667 0.704991
145 0.945172 1.650379 0.320000 0.716667 0.718333
146 0.946116 1.703276 0.376667 0.716667 0.699767
147 0.944659 1.672376 0.343333 0.700000 0.695990
148 0.948694 1.684725 0.383333 0.683333 0.703991
149 0.948101 1.697325 0.386667 0.700000 0.700256
150 0.953270 1.609696 0.363333 0.683333 0.702990
151 0.942221 1.756484 0.346667 0.716667 0.701992
152 0.939685 1.641910 0.350000 0.683333 0.707491
153 0.941101 1.685337 0.313333 0.683333 0.713491
154 0.938229 1.638440 0.343333 0.700000 0.724491
155 0.935531 1.681491 0.320000 0.683333 0.707493
156 0.940005 1.636299 0.346667 0.700000 0.702880
157 0.935673 1.657965 0.343333 0.683333 0.708992
158 0.938110 1.587218 0.326667 0.650000 0.708436
159 0.935539 1.778433 0.303333 0.683333 0.713492
160 0.933138 1.597521 0.323333 0.683333 0.703490
161 0.934763 1.726319 0.350000 0.700000 0.704991
162 0.937828 1.635992 0.333333 0.683333 0.703397
163 0.932190 1.612814 0.340000 0.666667 0.709491
164 0.935297 1.618365 0.326667 0.683333 0.751491
165 0.930964 1.623410 0.333333 0.666667 0.706991
166 0.940075 1.668642 0.366667 0.683333 0.701491
167 0.933793 1.617552 0.346667 0.650000 0.699491
168 0.938766 1.663031 0.333333 0.683333 0.704991
169 0.941232 1.626976 0.350000 0.683333 0.701306
170 0.933341 1.612207 0.320000 0.666667 0.690749
171 0.937339 1.693360 0.383333 0.666667 0.688991
172 0.931701 1.612850 0.346667 0.683333 0.709490
173 0.937925 1.622404 0.366667 0.683333 0.711250
174 0.935320 1.582409 0.353333 0.666667 0.727491
175 0.934324 1.618212 0.326667 0.683333 0.705873
176 0.929455 1.563488 0.336667 0.666667 0.715991
177 0.929729 1.586565 0.290000 0.683333 0.704658
178 0.933451 1.602940 0.370000 0.683333 0.706991
179 0.928717 1.573340 0.350000 0.666667 0.719991
180 0.925290 1.613369 0.300000 0.650000 0.712991
181 0.925307 1.616368 0.333333 0.683333 0.716491
182 0.928762 1.621261 0.360000 0.683333 0.802991
183 0.930573 1.620244 0.316667 0.683333 0.739824
184 0.930362 1.663133 0.370000 0.683333 0.738491
185 0.932049 1.602764 0.316667 0.666667 0.757490
186 0.938214 1.601306 0.350000 0.650000 0.770599
187 0.930034 1.584499 0.383333 0.683333 0.833987
188 0.924395 1.586506 0.306667 0.683333 0.734489
189 0.923696 1.523821 0.326667 0.666667 0.714491
190 0.922114 1.528593 0.310000 0.650000 0.716991
191 0.920627 1.611544 0.320000 0.666667 0.735993
192 0.925430 1.542244 0.340000 0.650000 0.719989
193 0.929689 1.621390 0.340000 0.666667 0.729579
194 0.935166 1.523914 0.400000 0.700000 0.728991
195 0.928755 1.574802 0.346667 0.666667 0.728491
196 0.921467 1.538520 0.316667 0.666667 0.730985
197 0.921368 1.640983 0.330000 0.683333 0.738490
198 0.925157 1.645675 0.323333 0.683333 0.703623
199 0.918965 1.473993 0.303333 0.683333 0.732491
200 0.922900 1.531551 0.330000 0.683333 0.692314
201 0.921041 1.527392 0.300000 0.666667 0.693678
202 0.918985 1.549988 0.336667 0.666667 0.721510
203 0.918650 1.554068 0.320000 0.683333 0.693866
204 0.924302 1.537482 0.346667 0.650000 0.688446
205 0.919061 1.485955 0.320000 0.683333 0.710490
206 0.919414 1.466837 0.300000 0.633333 0.715461
207 0.930215 1.513564 0.393333 0.666667 0.717491
208 0.915739 1.544081 0.306667 0.633333 0.703442
209 0.914533 1.509360 0.293333 0.666667 0.711991
210 0.913064 1.572206 0.296667 0.683333 0.710491
211 0.916282 1.538930 0.300000 0.683333 0.749119
212 0.915102 1.472220 0.303333 0.650000 0.691659
213 0.913911 1.599235 0.320000 0.683333 0.708489
214 0.912637 1.557416 0.310000 0.666667 0.706991
215 0.913841 1.541138 0.326667 0.666667 0.725992
216 0.915548 1.472203 0.303333 0.666667 0.729060
217 0.911305 1.563302 0.296667 0.666667 0.697493
218 0.912342 1.497310 0.333333 0.666667 0.697990
219 0.912140 1.565773 0.323333 0.633333 0.727492
220 0.915402 1.476618 0.323333 0.683333 0.711489
221 0.912820 1.581619 0.333333 0.666667 0.735675
222 0.914776 1.536006 0.340000 0.666667 0.705491
223 0.914127 1.575079 0.316667 0.666667 0.742490
224 0.911796 1.538983 0.330000 0.666667 0.715991
225 0.910560 1.508662 0.320000 0.666667 0.714628
226 0.911272 1.507964 0.316667 0.633333 0.702491
227 0.910172 1.539469 0.293333 0.666667 0.704113
228 0.911500 1.495480 0.330000 0.683333 0.710490
229 0.913281 1.470183 0.316667 0.650000 0.703991
230 0.909021 1.516184 0.300000 0.666667 0.705072
231 0.914272 1.548039 0.353333 0.683333 0.712493
232 0.910205 1.508645 0.303333 0.666667 0.702990
233 0.914380 1.496030 0.323333 0.650000 0.706491
234 0.915586 1.559450 0.306667 0.666667 0.698991
235 0.909686 1.503510 0.300000 0.650000 0.716455
236 0.908588 1.577522 0.343333 0.666667 0.703492
237 0.905469 1.453737 0.283333 0.666667 0.710490
238 0.906312 1.457462 0.300000 0.683333 0.721491
239 0.909069 1.563199 0.323333 0.666667 0.710492
240 0.910898 1.512830 0.353333 0.650000 0.712990
241 0.905804 1.510240 0.300000 0.666667 0.707991
242 0.904808 1.533014 0.306667 0.650000 0.716588
243 0.906220 1.567175 0.316667 0.666667 0.712492
244 0.904994 1.558251 0.316667 0.666667 0.707990
245 0.909664 1.462771 0.316667 0.650000 0.710491
246 0.902728 1.554450 0.276667 0.650000 0.722123
247 0.901789 1.493736 0.300000 0.633333 0.724491
248 0.901129 1.519779 0.286667 0.650000 0.713491
249 0.908045 1.529070 0.336667 0.683333 0.714491
250 0.909137 1.527716 0.366667 0.616667 0.712490
251 0.907243 1.496473 0.326667 0.666667 0.712991
252 0.903324 1.475295 0.280000 0.683333 0.716491
253 0.906966 1.566307 0.313333 0.650000 0.712991
254 0.905584 1.496057 0.323333 0.650000 0.720991
255 0.901105 1.522956 0.313333 0.666667 0.718417
256 0.899776 1.476555 0.296667 0.666667 0.713491
257 0.901022 1.443628 0.310000 0.650000 0.715991
258 0.901759 1.544258 0.286667 0.666667 0.698493
259 0.899705 1.510944 0.296667 0.633333 0.757990
260 0.902706 1.567773 0.283333 0.666667 0.713832
261 0.900594 1.481073 0.300000 0.633333 0.701421
262 0.900492 1.504038 0.310000 0.633333 0.706534
263 0.909880 1.484438 0.330000 0.633333 0.702991
264 0.902280 1.524400 0.293333 0.666667 0.716990
265 0.898809 1.484860 0.286667 0.650000 0.710222
266 0.900685 1.486311 0.293333 0.650000 0.715816
267 0.904731 1.581427 0.323333 0.650000 0.715670
268 0.899616 1.439474 0.280000 0.633333 0.713489
269 0.903097 1.543125 0.306667 0.666667 0.715491
270 0.899241 1.513516 0.290000 0.633333 0.711991
271 0.897982 1.549585 0.306667 0.650000 0.749492
272 0.895406 1.496974 0.300000 0.650000 0.705490
273 0.898603 1.538972 0.300000 0.633333 0.722991
274 0.900803 1.456101 0.313333 0.633333 0.715556
275 0.900999 1.539447 0.296667 0.666667 0.725991
276 0.899453 1.403440 0.306667 0.666667 0.696491
277 0.899719 1.500291 0.310000 0.666667 0.709991
278 0.899923 1.546075 0.333333 0.683333 0.712991
279 0.896875 1.527657 0.276667 0.666667 0.766991
280 0.894252 1.559931 0.290000 0.650000 0.704659
281 0.899306 1.424589 0.290000 0.633333 0.689491
282 0.900614 1.573746 0.306667 0.666667 0.707990
283 0.896283 1.544236 0.300000 0.650000 0.716491
284 0.898755 1.449153 0.300000 0.633333 0.713491
285 0.895454 1.560373 0.266667 0.650000 0.713493
286 0.900747 1.487210 0.353333 0.650000 0.714489
287 0.896175 1.584944 0.280000 0.683333 0.715991
288 0.897531 1.529690 0.320000 0.633333 0.706991
289 0.892979 1.517627 0.283333 0.650000 0.715991
290 0.894066 1.537414 0.306667 0.666667 0.739491
291 0.893829 1.521752 0.293333 0.650000 0.715376
292 0.900309 1.516614 0.326667 0.633333 0.705492
293 0.895294 1.488177 0.313333 0.666667 0.692046
294 0.898179 1.553349 0.296667 0.650000 0.725492
295 0.896506 1.465509 0.316667 0.650000 0.734989
296 0.894637 1.453335 0.286667 0.666667 0.730991
297 0.893534 1.477440 0.276667 0.616667 0.715992
298 0.892791 1.490009 0.286667 0.683333 0.728991
299 0.891321 1.514491 0.286667 0.616667 0.710935
300 0.895751 1.479844 0.310000 0.616667 0.718490
301 0.900866 1.509531 0.316667 0.650000 0.718597
302 0.890021 1.567908 0.280000 0.650000 0.740989
303 0.893569 1.520638 0.293333 0.650000 0.719490
304 0.890693 1.537821 0.286667 0.650000 0.722685
305 0.892409 1.502648 0.313333 0.650000 0.716726
306 0.890303 1.482915 0.306667 0.666667 0.702922
307 0.890694 1.574998 0.276667 0.650000 0.712492
308 0.892573 1.469983 0.323333 0.650000 0.716992
309 0.892593 1.516662 0.290000 0.666667 0.711989
310 0.893104 1.526790 0.330000 0.616667 0.710477
311 0.888380 1.574157 0.273333 0.633333 0.709992
312 0.892115 1.500358 0.306667 0.650000 0.718990
313 0.891347 1.493217 0.283333 0.666667 0.712493
314 0.887677 1.541948 0.280000 0.650000 0.716989
315 0.889589 1.518282 0.276667 0.650000 0.713995
316 0.890586 1.588591 0.296667 0.666667 0.713487
317 0.889970 1.581745 0.316667 0.650000 0.725996
318 0.890502 1.546701 0.283333 0.650000 0.726631
319 0.889998 1.530903 0.306667 0.650000 0.712388
320 0.886643 1.507182 0.283333 0.683333 0.707990
321 0.892742 1.546773 0.316667 0.650000 0.725502
322 0.890561 1.506240 0.316667 0.650000 0.712491
323 0.885743 1.547237 0.283333 0.666667 0.744991
324 0.886482 1.529686 0.283333 0.650000 0.730492
325 0.886298 1.491445 0.283333 0.650000 0.714992
326 0.891277 1.523510 0.306667 0.633333 0.695592
327 0.886757 1.615587 0.296667 0.666667 0.710119
328 0.886848 1.498849 0.286667 0.650000 0.719911
329 0.890289 1.545196 0.306667 0.650000 0.699991
330 0.886650 1.563992 0.286667 0.650000 0.712105
331 0.888414 1.573726 0.310000 0.700000 0.718991
332 0.884281 1.536549 0.273333 0.650000 0.734186
333 0.887546 1.557036 0.290000 0.650000 0.722990
334 0.888363 1.581320 0.313333 0.666667 0.709491
335 0.885590 1.521253 0.293333 0.650000 0.717491
336 0.883711 1.501403 0.270000 0.666667 0.700993
337 0.886440 1.545682 0.300000 0.633333 0.701400
338 0.882684 1.562092 0.270000 0.650000 0.699754
339 0.885404 1.545974 0.300000 0.650000 0.714990
340 0.883232 1.540597 0.283333 0.650000 0.751492
341 0.883469 1.508840 0.293333 0.650000 0.715775
342 0.883371 1.561110 0.266667 0.650000 0.715491
343 0.882206 1.514950 0.276667 0.650000 0.718990
344 0.883018 1.586547 0.270000 0.650000 0.709211
345 0.883405 1.518424 0.293333 0.650000 0.715991
346 0.883568 1.547356 0.300000 0.650000 0.723991
347 0.880822 1.510368 0.266667 0.650000 0.765524
348 0.883780 1.619718 0.286667 0.683333 0.744490
349 0.882103 1.554986 0.296667 0.650000 0.711991
350 0.881888 1.497146 0.290000 0.616667 0.759991
351 0.880178 1.523301 0.263333 0.666667 0.747489
352 0.881630 1.571813 0.280000 0.633333 0.751491
353 0.880963 1.487277 0.296667 0.650000 0.873909
354 0.881888 1.507177 0.300000 0.633333 0.741839
355 0.882805 1.526965 0.306667 0.666667 0.713494
356 0.882908 1.548831 0.300000 0.633333 0.711989
357 0.880555 1.531176 0.286667 0.650000 0.714491
358 0.881328 1.604432 0.296667 0.666667 0.720991
359 0.883484 1.554035 0.290000 0.650000 0.707992
360 0.880029 1.620798 0.290000 0.650000 0.705624
361 0.879364 1.523789 0.283333 0.650000 0.725991
362 0.878250 1.573790 0.276667 0.650000 0.722491
363 0.879939 1.592868 0.290000 0.650000 0.707491
364 0.878755 1.555785 0.280000 0.650000 0.818490
365 0.881321 1.574510 0.300000 0.633333 0.720442
366 0.878244 1.556496 0.273333 0.650000 0.726036
367 0.877547 1.574286 0.283333 0.650000 0.766378
368 0.879870 1.546865 0.306667 0.666667 0.725995
369 0.878494 1.561144 0.286667 0.666667 0.766499
370 0.878595 1.622733 0.283333 0.650000 0.701090
371 0.876514 1.565781 0.273333 0.650000 0.698194
372 0.879579 1.565139 0.306667 0.633333 0.691491
373 0.880292 1.578502 0.283333 0.633333 0.726118
374 0.881314 1.522828 0.276667 0.650000 0.713991
375 0.884296 1.581449 0.320000 0.650000 0.709991
376 0.879847 1.544302 0.296667 0.666667 0.695102
377 0.878963 1.581624 0.286667 0.650000 0.705476
378 0.880456 1.573666 0.286667 0.666667 0.700491
379 0.881732 1.555567 0.283333 0.633333 0.720489
380 0.877074 1.574166 0.273333 0.650000 0.707756
381 0.879782 1.633106 0.283333 0.666667 0.711991
382 0.889287 1.585200 0.353333 0.633333 0.715491
383 0.880346 1.574907 0.296667 0.666667 0.726490
384 0.881504 1.537956 0.290000 0.650000 0.702052
385 0.875310 1.584399 0.270000 0.650000 0.717036
386 0.877183 1.580556 0.276667 0.650000 0.715989
387 0.875188 1.692006 0.276667 0.666667 0.722992
388 0.875956 1.618534 0.280000 0.650000 0.708992
389 0.875367 1.640223 0.260000 0.650000 0.734489
390 0.875160 1.598551 0.290000 0.650000 0.718492
391 0.874072 1.608121 0.273333 0.650000 0.713492
392 0.876029 1.636369 0.280000 0.650000 0.713989
393 0.877707 1.614132 0.300000 0.650000 0.714992
394 0.874190 1.655139 0.286667 0.650000 0.785990
395 0.877916 1.619916 0.290000 0.650000 0.766990
396 0.875877 1.550215 0.286667 0.650000 0.750992
397 0.878525 1.616913 0.283333 0.650000 0.775491
398 0.875100 1.572469 0.280000 0.633333 0.700219
399 0.876116 1.618767 0.293333 0.633333 0.711489
400 0.875702 1.622121 0.286667 0.650000 0.708491
401 0.874363 1.621276 0.290000 0.650000 0.735993
402 0.874559 1.562164 0.290000 0.650000 0.735547
403 0.874244 1.624241 0.280000 0.633333 0.707790
404 0.873605 1.642987 0.276667 0.633333 0.714989
405 0.873240 1.608235 0.266667 0.650000 0.731992
406 0.873850 1.668524 0.283333 0.650000 0.718489
407 0.873502 1.686036 0.270000 0.616667 0.754991
408 0.873101 1.597906 0.283333 0.650000 0.734491
409 0.872933 1.666261 0.276667 0.650000 0.716612
410 0.872963 1.614813 0.266667 0.650000 0.728352
411 0.872608 1.628926 0.290000 0.650000 0.722492
412 0.874058 1.680394 0.280000 0.650000 0.725491
413 0.872153 1.604510 0.293333 0.633333 0.730163
414 0.871958 1.617887 0.280000 0.650000 0.738991
415 0.875027 1.679277 0.276667 0.633333 0.731490
416 0.871264 1.671314 0.283333 0.650000 0.733700
417 0.872043 1.582371 0.270000 0.650000 0.740536
418 0.877161 1.671151 0.286667 0.633333 0.743989
419 0.874298 1.659939 0.296667 0.666667 0.751490
420 0.874269 1.627162 0.263333 0.633333 0.743990
421 0.871233 1.595870 0.290000 0.650000 0.735991
422 0.870608 1.598035 0.290000 0.650000 0.722991
423 0.873166 1.590912 0.300000 0.650000 0.713161
424 0.870908 1.614878 0.263333 0.650000 0.759990
425 0.873370 1.662284 0.290000 0.650000 0.721492
426 0.872355 1.615805 0.273333 0.650000 0.765207
427 0.872800 1.622500 0.286667 0.650000 0.709991
428 0.869820 1.661372 0.260000 0.616667 0.724990
429 0.874033 1.642106 0.273333 0.650000 0.715992
430 0.869969 1.670286 0.276667 0.666667 0.711990
431 0.870706 1.651205 0.293333 0.650000 0.720991
432 0.870515 1.611015 0.276667 0.650000 0.712492
433 0.870322 1.672141 0.280000 0.650000 0.712534
434 0.869745 1.628996 0.293333 0.633333 0.731596
435 0.869306 1.678158 0.263333 0.666667 0.718991
436 0.869416 1.636314 0.273333 0.650000 0.717991
437 0.868472 1.688545 0.270000 0.633333 0.707991
438 0.870761 1.612736 0.293333 0.650000 0.712991
439 0.871125 1.607177 0.283333 0.683333 0.710990
440 0.870926 1.624488 0.280000 0.633333 0.714992
441 0.870700 1.673018 0.280000 0.650000 0.719990
442 0.868553 1.625210 0.290000 0.650000 0.721491
443 0.869257 1.632908 0.263333 0.633333 0.717991
444 0.871212 1.657312 0.286667 0.650000 0.712991
445 0.870331 1.643186 0.290000 0.666667 0.721492
446 0.869313 1.647304 0.270000 0.666667 0.713992
447 0.869091 1.594635 0.273333 0.650000 0.704845
448 0.873389 1.581443 0.320000 0.650000 0.723491
449 0.868678 1.659853 0.276667 0.650000 0.717490
450 0.867949 1.626451 0.290000 0.666667 0.717492
451 0.870831 1.603199 0.286667 0.650000 0.711756
452 0.869287 1.679238 0.276667 0.650000 0.710040
453 0.866493 1.615914 0.273333 0.666667 0.714345
454 0.865810 1.701975 0.250000 0.650000 0.741123
455 0.866254 1.614077 0.266667 0.666667 0.727997
456 0.869822 1.675660 0.286667 0.666667 0.725987
457 0.867208 1.592642 0.283333 0.650000 0.723207
458 0.868752 1.642348 0.270000 0.650000 0.745118
459 0.867862 1.597228 0.276667 0.650000 0.733174
460 0.867082 1.686385 0.256667 0.650000 0.707489
461 0.866599 1.617280 0.280000 0.666667 0.718215
462 0.866375 1.702017 0.283333 0.650000 0.746991
463 0.866850 1.572758 0.293333 0.650000 0.716991
464 0.866315 1.700451 0.270000 0.633333 0.759552
465 0.865192 1.667085 0.280000 0.650000 0.708122
466 0.865778 1.672499 0.276667 0.616667 0.728320
467 0.864890 1.634231 0.260000 0.650000 0.785493
468 0.866757 1.661181 0.266667 0.633333 0.721488
469 0.865168 1.624720 0.263333 0.683333 0.709491
470 0.865030 1.692556 0.270000 0.633333 0.722990
471 0.865794 1.695530 0.276667 0.650000 0.730991
472 0.864937 1.679997 0.273333 0.633333 0.724492
473 0.865411 1.638261 0.283333 0.666667 0.716492
474 0.865549 1.674066 0.273333 0.650000 0.718126
475 0.865614 1.709901 0.270000 0.650000 0.730991
476 0.865140 1.714752 0.283333 0.633333 0.711118
477 0.867914 1.706460 0.273333 0.633333 0.734718
478 0.863578 1.720007 0.286667 0.650000 0.725490
479 0.865149 1.675914 0.270000 0.650000 0.709491
480 0.863452 1.685003 0.276667 0.633333 0.719674
481 0.863893 1.759197 0.250000 0.666667 0.744990
482 0.864161 1.626745 0.266667 0.666667 0.717274
483 0.863320 1.663999 0.270000 0.633333 0.719991
484 0.863838 1.663616 0.280000 0.650000 0.718992
485 0.864359 1.696927 0.273333 0.666667 0.717991
486 0.863390 1.671009 0.280000 0.650000 0.718105
487 0.863345 1.652037 0.286667 0.633333 0.711491
488 0.864030 1.644885 0.266667 0.633333 0.722516
489 0.863824 1.688134 0.266667 0.650000 0.714991
490 0.865264 1.680263 0.283333 0.650000 0.716492
491 0.863334 1.715220 0.273333 0.650000 0.711493
492 0.862462 1.706616 0.273333 0.633333 0.717989
493 0.862048 1.752756 0.243333 0.650000 0.707491
494 0.861917 1.677643 0.266667 0.650000 0.714491
495 0.862748 1.668679 0.283333 0.650000 0.720491
496 0.862555 1.677117 0.270000 0.666667 0.752848
497 0.861982 1.704741 0.260000 0.633333 0.736990
498 0.861399 1.713654 0.270000 0.633333 0.740989
499 0.861950 1.676173 0.263333 0.666667 0.748000
500 0.861709 1.647092 0.276667 0.650000 0.775539
501 0.861497 1.737871 0.253333 0.633333 0.742489
502 0.861538 1.719586 0.263333 0.616667 0.864490
503 0.860704 1.737500 0.266667 0.650000 0.721505
504 0.861511 1.694644 0.286667 0.650000 0.718596
505 0.863016 1.729979 0.273333 0.650000 0.720099
506 0.862472 1.715306 0.276667 0.616667 0.712991
507 0.860228 1.690946 0.250000 0.666667 0.708991
508 0.861324 1.667499 0.266667 0.650000 0.701492
509 0.863097 1.684604 0.296667 0.666667 0.707430
510 0.861013 1.740663 0.280000 0.633333 0.714990
511 0.861116 1.678211 0.290000 0.650000 0.717491
512 0.860875 1.721847 0.270000 0.633333 0.715991
513 0.862146 1.671926 0.283333 0.616667 0.711328
514 0.861097 1.726235 0.273333 0.633333 0.786992
515 0.861542 1.655710 0.296667 0.666667 0.707936
516 0.860743 1.668881 0.266667 0.633333 0.714037
517 0.859445 1.730753 0.263333 0.616667 0.700709
518 0.860072 1.691003 0.260000 0.666667 0.723989
519 0.860013 1.727261 0.273333 0.633333 0.735991
520 0.859820 1.669123 0.260000 0.633333 0.714100
521 0.859811 1.620919 0.283333 0.633333 0.720988
522 0.861332 1.744039 0.276667 0.633333 0.702992
523 0.858968 1.693151 0.266667 0.666667 0.709990
524 0.861228 1.680861 0.280000 0.650000 0.748991
525 0.858719 1.690025 0.270000 0.633333 0.727494
526 0.857677 1.698327 0.266667 0.633333 0.707995
527 0.858649 1.659882 0.280000 0.666667 0.710992
528 0.857969 1.718820 0.253333 0.633333 0.718901
529 0.860285 1.648697 0.270000 0.650000 0.711427
530 0.859732 1.657783 0.263333 0.633333 0.712490
531 0.858260 1.667328 0.270000 0.650000 0.710991
532 0.858824 1.632672 0.266667 0.650000 0.719991
533 0.858520 1.727402 0.263333 0.633333 0.711598
534 0.858155 1.640477 0.283333 0.650000 0.709991
535 0.858314 1.676578 0.270000 0.633333 0.716991
536 0.859556 1.677101 0.266667 0.633333 0.701383
537 0.857556 1.691113 0.266667 0.666667 0.715491
538 0.858208 1.685157 0.263333 0.633333 0.720991
539 0.857413 1.695742 0.263333 0.633333 0.757489
540 0.857820 1.685056 0.273333 0.666667 0.726498
541 0.857594 1.634304 0.283333 0.666667 0.718492
542 0.856919 1.721528 0.250000 0.633333 0.718991
543 0.857451 1.634909 0.280000 0.633333 0.717232
544 0.856601 1.715201 0.250000 0.633333 0.717990
545 0.856303 1.690851 0.253333 0.666667 0.716991
546 0.858722 1.732867 0.276667 0.633333 0.710492
547 0.856560 1.681793 0.270000 0.666667 0.717991
548 0.857401 1.677749 0.266667 0.650000 0.765490
549 0.856809 1.699012 0.256667 0.650000 0.699757
550 0.856606 1.712394 0.270000 0.633333 0.711489
551 0.856607 1.696491 0.260000 0.650000 0.733322
552 0.856176 1.657492 0.266667 0.633333 0.695953
553 0.856602 1.709597 0.273333 0.650000 0.720993
554 0.855857 1.717186 0.263333 0.633333 0.715989
555 0.855999 1.692184 0.266667 0.633333 0.711993
556 0.855016 1.686159 0.253333 0.633333 0.738490
557 0.856133 1.721931 0.270000 0.650000 0.723990
558 0.855630 1.625017 0.280000 0.633333 0.711901
559 0.855812 1.696420 0.253333 0.650000 0.705491
560 0.855316 1.716409 0.246667 0.650000 0.723990
561 0.855645 1.710380 0.260000 0.633333 0.713491
562 0.855019 1.729007 0.273333 0.633333 0.715992
563 0.854737 1.700652 0.256667 0.650000 0.720991
564 0.854438 1.688203 0.266667 0.650000 0.717991
565 0.854634 1.695100 0.270000 0.633333 0.714813
566 0.854661 1.695061 0.266667 0.650000 0.716491
567 0.855628 1.743360 0.266667 0.650000 0.722990
568 0.854491 1.700141 0.263333 0.650000 0.716338
569 0.854584 1.717868 0.256667 0.650000 0.756989
570 0.854286 1.697904 0.263333 0.666667 0.716991
571 0.855479 1.691437 0.266667 0.633333 0.756606
572 0.854763 1.689504 0.276667 0.666667 0.715492
573 0.854335 1.745204 0.260000 0.616667 0.726991
574 0.853667 1.667847 0.253333 0.666667 0.729491
575 0.854496 1.727062 0.260000 0.633333 0.717990
576 0.854651 1.720617 0.263333 0.650000 0.724491
577 0.854550 1.683457 0.253333 0.650000 0.761992
578 0.853690 1.691279 0.263333 0.650000 0.714990
579 0.853173 1.726047 0.240000 0.633333 0.746491
580 0.852959 1.734140 0.253333 0.633333 0.722140
581 0.852986 1.705237 0.266667 0.650000 0.714988
582 0.853251 1.734701 0.250000 0.633333 0.717490
583 0.853274 1.704391 0.266667 0.650000 0.719491
584 0.853152 1.711632 0.273333 0.650000 0.718491
585 0.852396 1.750426 0.263333 0.633333 0.717992
586 0.854194 1.680958 0.266667 0.650000 0.738957
587 0.853710 1.754557 0.243333 0.633333 0.699160
588 0.853268 1.696252 0.270000 0.633333 0.703798
589 0.853080 1.682671 0.270000 0.666667 0.696847
590 0.851980 1.734145 0.256667 0.650000 0.711989
591 0.852049 1.728825 0.263333 0.666667 0.724991
592 0.852374 1.758674 0.263333 0.633333 0.723491
593 0.852771 1.680090 0.266667 0.666667 0.721490
594 0.852099 1.757278 0.250000 0.633333 0.718992
595 0.852232 1.717485 0.260000 0.650000 0.708490
596 0.852278 1.736163 0.263333 0.633333 0.729576
597 0.851725 1.692470 0.263333 0.666667 0.716746
598 0.852021 1.718486 0.263333 0.666667 0.713491
599 0.852779 1.785339 0.256667 0.633333 0.714490
600 0.853676 1.719848 0.270000 0.633333 0.730491
601 0.851678 1.715313 0.276667 0.633333 0.723403
602 0.851956 1.785474 0.246667 0.616667 0.716491
603 0.851725 1.713537 0.263333 0.650000 0.756490
604 0.851317 1.706689 0.260000 0.650000 0.743992
605 0.851024 1.736208 0.253333 0.650000 0.723798
606 0.850734 1.751404 0.250000 0.650000 0.702298
607 0.851422 1.702428 0.270000 0.650000 0.717785
608 0.850872 1.733211 0.266667 0.633333 0.724993
609 0.850778 1.723672 0.253333 0.650000 0.720989
610 0.851891 1.688285 0.256667 0.650000 0.708991
611 0.850888 1.736414 0.243333 0.633333 0.717493
612 0.850808 1.716385 0.260000 0.650000 0.724167
613 0.851165 1.715252 0.256667 0.666667 0.714491
614 0.851023 1.756586 0.270000 0.633333 0.717492
615 0.851068 1.719593 0.280000 0.650000 0.726990
616 0.850446 1.718237 0.273333 0.633333 0.720991
617 0.851107 1.686212 0.266667 0.650000 0.716990
618 0.850683 1.731017 0.276667 0.633333 0.731492
619 0.850695 1.718095 0.260000 0.650000 0.718490
620 0.850843 1.706388 0.273333 0.650000 0.714492
621 0.849974 1.743849 0.250000 0.633333 0.719990
622 0.850226 1.693373 0.266667 0.650000 0.725607
623 0.850281 1.747150 0.240000 0.633333 0.762676
624 0.849601 1.752017 0.253333 0.650000 0.727491
625 0.849564 1.732304 0.256667 0.633333 0.741755
626 0.849494 1.740631 0.253333 0.650000 0.699849
627 0.850355 1.731675 0.250000 0.633333 0.715991
628 0.849349 1.764241 0.253333 0.633333 0.713991
629 0.849874 1.744944 0.260000 0.666667 0.717638
630 0.849519 1.770177 0.253333 0.633333 0.713491
631 0.849194 1.743980 0.256667 0.666667 0.711491
632 0.848974 1.777501 0.250000 0.633333 0.718290
633 0.849006 1.733679 0.253333 0.666667 0.728993
634 0.849138 1.774917 0.243333 0.633333 0.717607
635 0.849903 1.731170 0.260000 0.650000 0.717490
636 0.848997 1.763890 0.256667 0.633333 0.752991
637 0.848924 1.750159 0.256667 0.633333 0.723994
638 0.849477 1.709134 0.270000 0.650000 0.718491
639 0.848913 1.792704 0.256667 0.633333 0.732491
640 0.849260 1.740912 0.263333 0.633333 0.713227
641 0.848851 1.747591 0.260000 0.650000 0.701355
642 0.848106 1.747434 0.253333 0.633333 0.720490
643 0.848415 1.747531 0.260000 0.650000 0.734992
644 0.848636 1.757827 0.253333 0.666667 0.737989
645 0.848461 1.734583 0.270000 0.650000 0.727745
646 0.848158 1.754395 0.253333 0.616667 0.720990
647 0.848207 1.754383 0.260000 0.633333 0.723490
648 0.847819 1.731915 0.270000 0.650000 0.714991
649 0.847820 1.721542 0.266667 0.633333 0.709991
650 0.847849 1.753271 0.240000 0.616667 0.733491
651 0.847374 1.742448 0.256667 0.650000 0.707501
652 0.848256 1.700672 0.260000 0.666667 0.717490
653 0.847582 1.766712 0.246667 0.650000 0.720992
654 0.847727 1.766739 0.243333 0.633333 0.724491
655 0.847800 1.758333 0.260000 0.616667 0.715991
656 0.847435 1.767767 0.266667 0.666667 0.723303
657 0.848462 1.721827 0.266667 0.650000 0.736492
658 0.848557 1.758670 0.256667 0.633333 0.716490
659 0.847579 1.766289 0.253333 0.633333 0.751491
660 0.847179 1.721189 0.260000 0.650000 0.733189
661 0.846855 1.757603 0.256667 0.633333 0.724876
662 0.847119 1.731915 0.256667 0.666667 0.718717
663 0.846889 1.759142 0.250000 0.650000 0.720604
664 0.846822 1.726385 0.260000 0.650000 0.721490
665 0.846787 1.727688 0.260000 0.650000 0.712995
666 0.846748 1.730275 0.256667 0.650000 0.710489
667 0.846794 1.762465 0.236667 0.650000 0.714489
668 0.846575 1.717050 0.260000 0.666667 0.718491
669 0.846589 1.784856 0.236667 0.616667 0.724490
670 0.846524 1.761271 0.253333 0.633333 0.718448
671 0.846267 1.747753 0.246667 0.666667 0.717511
672 0.846128 1.772712 0.256667 0.650000 0.715993
673 0.846932 1.756635 0.273333 0.633333 0.722989
674 0.846649 1.758952 0.260000 0.633333 0.721491
675 0.846125 1.745503 0.256667 0.650000 0.710492
676 0.845948 1.730958 0.266667 0.650000 0.732490
677 0.846137 1.764670 0.236667 0.633333 0.749991
678 0.846157 1.717314 0.233333 0.650000 0.720492
679 0.846127 1.796233 0.243333 0.633333 0.724490
680 0.846281 1.759177 0.260000 0.616667 0.709955
681 0.846053 1.759336 0.243333 0.650000 0.731492
682 0.845468 1.767471 0.266667 0.650000 0.720990
683 0.846031 1.793072 0.253333 0.633333 0.742491
684 0.845506 1.761027 0.253333 0.650000 0.723991
685 0.845445 1.776348 0.240000 0.650000 0.721991
686 0.845360 1.783542 0.253333 0.650000 0.722492
687 0.845422 1.747765 0.260000 0.650000 0.720490
688 0.845596 1.795611 0.243333 0.633333 0.724490
689 0.845150 1.741239 0.263333 0.666667 0.723491
690 0.845520 1.790352 0.246667 0.633333 0.731844
691 0.845257 1.769590 0.266667 0.633333 0.713113
692 0.845405 1.756942 0.250000 0.650000 0.709653
693 0.845237 1.742316 0.260000 0.650000 0.721490
694 0.844729 1.742972 0.256667 0.633333 0.744992
695 0.844939 1.750323 0.250000 0.666667 0.751489
696 0.844973 1.764516 0.256667 0.650000 0.756491
697 0.845070 1.745681 0.266667 0.650000 0.711491
698 0.844846 1.764241 0.256667 0.633333 0.727492
699 0.844280 1.741678 0.256667 0.650000 0.712990
700 0.844537 1.757084 0.246667 0.633333 0.716491
701 0.844666 1.786718 0.243333 0.633333 0.726769
702 0.844680 1.773166 0.260000 0.633333 0.721991
703 0.844661 1.798020 0.233333 0.633333 0.721557
704 0.844776 1.751512 0.263333 0.666667 0.726490
705 0.845030 1.785487 0.260000 0.633333 0.722076
706 0.844533 1.789471 0.253333 0.633333 0.724991
707 0.844287 1.788081 0.243333 0.633333 0.719491
708 0.843988 1.780204 0.246667 0.650000 0.730491
709 0.844075 1.772917 0.260000 0.633333 0.707135
710 0.843974 1.765075 0.263333 0.633333 0.715490
711 0.843847 1.781767 0.243333 0.650000 0.720991
712 0.843905 1.776582 0.246667 0.633333 0.729992
713 0.844214 1.764894 0.263333 0.650000 0.731387
714 0.844035 1.768875 0.246667 0.650000 0.720992
715 0.844019 1.765326 0.250000 0.633333 0.734618
716 0.843749 1.802159 0.256667 0.616667 0.728195
717 0.843525 1.790133 0.246667 0.650000 0.718991
718 0.843758 1.760171 0.263333 0.650000 0.741995
719 0.843614 1.746911 0.253333 0.650000 0.719486
720 0.843683 1.794315 0.233333 0.633333 0.725991
721 0.843514 1.764827 0.243333 0.650000 0.730117
722 0.843546 1.770654 0.253333 0.650000 0.701440
723 0.843610 1.754816 0.263333 0.650000 0.720989
724 0.843685 1.784720 0.243333 0.633333 0.728578
725 0.843428 1.794124 0.236667 0.633333 0.759449
726 0.843529 1.793035 0.226667 0.633333 0.729991
727 0.843557 1.785281 0.246667 0.633333 0.728063
728 0.843297 1.757240 0.246667 0.650000 0.725491
729 0.843598 1.793587 0.246667 0.633333 0.716491
730 0.843182 1.796083 0.243333 0.633333 0.718491
731 0.844021 1.760881 0.263333 0.650000 0.726280
732 0.843353 1.747677 0.263333 0.650000 0.716991
733 0.843748 1.773978 0.233333 0.650000 0.717495
734 0.843012 1.812366 0.246667 0.633333 0.715489
735 0.842824 1.796337 0.243333 0.633333 0.718990
736 0.842907 1.786316 0.243333 0.633333 0.716491
737 0.842955 1.787559 0.250000 0.650000 0.738992
738 0.843086 1.788122 0.243333 0.633333 0.747128
739 0.842928 1.757024 0.246667 0.650000 0.716991
740 0.843066 1.771170 0.240000 0.650000 0.718492
741 0.842646 1.751854 0.253333 0.650000 0.715489
742 0.842873 1.776929 0.246667 0.633333 0.726492
743 0.842725 1.815724 0.236667 0.633333 0.744675
744 0.842630 1.783221 0.243333 0.650000 0.719989
745 0.842697 1.790586 0.243333 0.633333 0.736601
746 0.842547 1.787754 0.253333 0.633333 0.720992
747 0.842565 1.768919 0.250000 0.633333 0.717989
748 0.842485 1.803114 0.243333 0.633333 0.721491
749 0.842540 1.780595 0.250000 0.616667 0.727491
750 0.843213 1.762808 0.260000 0.650000 0.720171
751 0.842549 1.762322 0.263333 0.650000 0.719693
752 0.842532 1.775852 0.253333 0.666667 0.721491
753 0.842389 1.784484 0.256667 0.633333 0.711493
754 0.842596 1.798270 0.250000 0.633333 0.719489
755 0.842171 1.767370 0.250000 0.650000 0.722266
756 0.842167 1.784293 0.250000 0.650000 0.751989
757 0.842196 1.800234 0.246667 0.633333 0.752359
758 0.842181 1.781207 0.243333 0.650000 0.729492
759 0.842197 1.755111 0.250000 0.650000 0.713085
760 0.842072 1.783528 0.250000 0.633333 0.724490
761 0.842114 1.787048 0.253333 0.633333 0.714491
762 0.842127 1.770124 0.250000 0.650000 0.717562
763 0.842126 1.786544 0.246667 0.616667 0.718174
764 0.842061 1.782562 0.233333 0.633333 0.731490
765 0.841841 1.759974 0.250000 0.633333 0.714991
766 0.841846 1.785870 0.243333 0.633333 0.709505
767 0.841859 1.788493 0.240000 0.633333 0.702644
768 0.842077 1.798046 0.256667 0.633333 0.709490
769 0.842073 1.762545 0.256667 0.633333 0.716492
770 0.842065 1.755212 0.253333 0.650000 0.729990
771 0.841829 1.759444 0.246667 0.650000 0.737636
772 0.841909 1.795214 0.246667 0.633333 0.723991
773 0.841807 1.791062 0.240000 0.633333 0.725493
774 0.841794 1.777393 0.253333 0.633333 0.718988
775 0.841778 1.766383 0.263333 0.633333 0.729992
776 0.842226 1.762950 0.260000 0.650000 0.730494
777 0.841600 1.754846 0.250000 0.650000 0.770990
778 0.841628 1.768701 0.240000 0.650000 0.730111
779 0.841769 1.775931 0.256667 0.633333 0.724991
780 0.841590 1.777325 0.246667 0.633333 0.735492
781 0.841482 1.752720 0.253333 0.650000 0.737990
782 0.841388 1.767062 0.246667 0.650000 0.731408
783 0.841410 1.781687 0.243333 0.633333 0.741491
784 0.841366 1.755660 0.256667 0.633333 0.735991
785 0.841432 1.760047 0.243333 0.650000 0.753491
786 0.841506 1.800321 0.236667 0.633333 0.751991
787 0.841171 1.796874 0.243333 0.633333 0.759490
788 0.841423 1.765795 0.256667 0.650000 0.768897
789 0.841349 1.739605 0.246667 0.650000 0.742378
790 0.841238 1.776689 0.243333 0.633333 0.728063
791 0.841485 1.753674 0.240000 0.650000 0.726495
792 0.841390 1.765785 0.246667 0.633333 0.700906
793 0.841398 1.774083 0.236667 0.633333 0.784989
794 0.841126 1.766968 0.250000 0.650000 0.745359
795 0.841019 1.762359 0.243333 0.650000 0.712113
796 0.840989 1.761560 0.250000 0.650000 0.728991
797 0.841036 1.767084 0.260000 0.633333 0.712491
798 0.841026 1.759657 0.250000 0.650000 0.714993
799 0.840950 1.765417 0.250000 0.650000 0.746990
800 0.841316 1.800128 0.246667 0.633333 0.767490
801 0.840979 1.757689 0.236667 0.650000 0.743759
802 0.840974 1.777697 0.243333 0.633333 0.738989
803 0.840881 1.772521 0.243333 0.633333 0.904990
804 0.841112 1.778117 0.243333 0.633333 0.747491
805 0.840840 1.776776 0.246667 0.633333 0.736490
806 0.841011 1.790095 0.240000 0.633333 0.721490
807 0.840958 1.788473 0.240000 0.633333 0.716991
808 0.840854 1.767145 0.256667 0.650000 0.723896
809 0.840594 1.779591 0.246667 0.633333 0.757492
810 0.840721 1.795999 0.246667 0.633333 0.735991
811 0.840733 1.784631 0.250000 0.633333 0.695655
812 0.840520 1.796097 0.253333 0.633333 0.711491
813 0.840704 1.785285 0.240000 0.650000 0.726990
814 0.840523 1.790345 0.256667 0.633333 0.719491
815 0.840479 1.783261 0.240000 0.633333 0.716991
816 0.840703 1.760045 0.243333 0.650000 0.710450
817 0.840606 1.776908 0.240000 0.650000 0.721991
818 0.840508 1.796477 0.240000 0.633333 0.718159
819 0.840403 1.797226 0.243333 0.633333 0.763692
820 0.840497 1.786389 0.240000 0.650000 0.708203
821 0.840451 1.784953 0.240000 0.633333 0.717990
822 0.840466 1.792018 0.240000 0.633333 0.718992
823 0.840443 1.778264 0.253333 0.650000 0.720675
824 0.840423 1.781797 0.250000 0.650000 0.725491
825 0.840396 1.786188 0.250000 0.633333 0.722491
826 0.840323 1.790597 0.250000 0.633333 0.711490
827 0.840365 1.784677 0.246667 0.633333 0.763490
828 0.840314 1.783196 0.243333 0.633333 0.719491
829 0.840348 1.776775 0.250000 0.633333 0.711491
830 0.840237 1.767693 0.250000 0.650000 0.715991
831 0.840328 1.775675 0.256667 0.650000 0.721991
832 0.840243 1.792496 0.243333 0.633333 0.721570
833 0.840267 1.785063 0.253333 0.633333 0.720990
834 0.840241 1.793217 0.246667 0.633333 0.721991
835 0.840291 1.812642 0.243333 0.633333 0.715993
836 0.840202 1.789156 0.240000 0.633333 0.703991
837 0.840187 1.788845 0.250000 0.633333 0.717997
838 0.840106 1.780335 0.250000 0.633333 0.722991
839 0.840145 1.777659 0.250000 0.633333 0.720982
840 0.840202 1.771223 0.250000 0.633333 0.714991
841 0.840175 1.783134 0.250000 0.633333 0.722991
842 0.840110 1.770663 0.253333 0.650000 0.720062
843 0.840065 1.786528 0.243333 0.633333 0.718489
844 0.840120 1.774858 0.246667 0.650000 0.720991
845 0.840119 1.779332 0.243333 0.633333 0.719992
846 0.840077 1.790671 0.250000 0.633333 0.722718
847 0.839943 1.781551 0.246667 0.633333 0.712473
848 0.839982 1.787872 0.243333 0.633333 0.722990
849 0.839908 1.787999 0.243333 0.633333 0.726991
850 0.839943 1.790039 0.246667 0.633333 0.720991
851 0.839907 1.783117 0.250000 0.633333 0.714991
852 0.839880 1.790007 0.243333 0.633333 0.717492
853 0.839914 1.781068 0.253333 0.650000 0.712492
854 0.839857 1.786884 0.253333 0.633333 0.716490
855 0.839829 1.788914 0.243333 0.633333 0.734491
856 0.839859 1.780697 0.250000 0.633333 0.716491
857 0.839828 1.778002 0.243333 0.650000 0.722991
858 0.839733 1.794352 0.246667 0.633333 0.724991
859 0.839802 1.804818 0.240000 0.633333 0.718992
860 0.839782 1.788687 0.246667 0.633333 0.715492
861 0.839772 1.788898 0.250000 0.633333 0.699144
862 0.839736 1.789753 0.253333 0.633333 0.701037
863 0.839827 1.781146 0.250000 0.650000 0.726624
864 0.839715 1.787821 0.246667 0.633333 0.732990
865 0.839693 1.795283 0.240000 0.633333 0.718491
866 0.839646 1.792202 0.236667 0.633333 0.719991
867 0.839695 1.797694 0.243333 0.633333 0.717490
868 0.839814 1.783678 0.250000 0.650000 0.717992
869 0.839661 1.783072 0.243333 0.633333 0.721491
870 0.839639 1.790421 0.236667 0.633333 0.723990
871 0.839594 1.800014 0.236667 0.633333 0.720991
872 0.839658 1.795600 0.240000 0.633333 0.718991
873 0.839580 1.790724 0.236667 0.633333 0.728490
874 0.839616 1.788395 0.246667 0.650000 0.721491
875 0.839575 1.785680 0.243333 0.650000 0.720440
876 0.839528 1.790324 0.243333 0.633333 0.710990
877 0.839539 1.789025 0.243333 0.650000 0.713990
878 0.839573 1.790711 0.240000 0.633333 0.719491
879 0.839575 1.789318 0.243333 0.633333 0.715991
880 0.839575 1.780312 0.256667 0.650000 0.725991
881 0.839528 1.785591 0.236667 0.650000 0.724991
882 0.839511 1.783402 0.253333 0.650000 0.715169
883 0.839465 1.789444 0.236667 0.650000 0.723286
884 0.839506 1.799843 0.243333 0.633333 0.721185
885 0.839468 1.790043 0.250000 0.633333 0.755090
886 0.839525 1.792450 0.240000 0.633333 0.723490
887 0.839432 1.783869 0.236667 0.650000 0.716991
888 0.839457 1.783690 0.253333 0.650000 0.718357
889 0.839459 1.790293 0.246667 0.633333 0.717625
890 0.839429 1.800294 0.240000 0.633333 0.728494
891 0.839473 1.797583 0.233333 0.633333 0.732989
892 0.839447 1.804665 0.236667 0.633333 0.714693
893 0.839468 1.793253 0.246667 0.633333 0.716490
894 0.839409 1.791585 0.243333 0.633333 0.726991
895 0.839376 1.794638 0.236667 0.633333 0.730651
896 0.839370 1.799651 0.236667 0.633333 0.722992
897 0.839376 1.799263 0.240000 0.633333 0.717990
898 0.839349 1.792257 0.243333 0.633333 0.712492
899 0.839318 1.802353 0.243333 0.633333 0.718992
900 0.839390 1.798039 0.250000 0.633333 0.709804
901 0.839446 1.784986 0.250000 0.633333 0.725991
902 0.839374 1.798835 0.240000 0.633333 0.729491
903 0.839447 1.791793 0.243333 0.633333 0.719385
904 0.839337 1.786384 0.246667 0.633333 0.711804
905 0.839395 1.775760 0.250000 0.650000 0.717991
906 0.839379 1.782198 0.253333 0.633333 0.724491
907 0.839256 1.796142 0.240000 0.633333 0.716991
908 0.839292 1.788515 0.246667 0.633333 0.719492
909 0.839313 1.787415 0.243333 0.633333 0.719490
910 0.839250 1.786654 0.236667 0.633333 0.718491
911 0.839245 1.794315 0.230000 0.633333 0.725992
912 0.839276 1.799465 0.236667 0.633333 0.710573
913 0.839236 1.797653 0.233333 0.633333 0.719491
914 0.839222 1.795163 0.243333 0.633333 0.720491
915 0.839366 1.786866 0.253333 0.633333 0.711998
916 0.839288 1.785493 0.250000 0.633333 0.726984
917 0.839203 1.789837 0.243333 0.633333 0.725491
918 0.839243 1.793860 0.240000 0.633333 0.721991
919 0.839263 1.791858 0.233333 0.633333 0.718490
920 0.839257 1.787891 0.246667 0.633333 0.708992
921 0.839244 1.793670 0.236667 0.633333 0.723989
922 0.839214 1.790013 0.240000 0.633333 0.715991
923 0.839260 1.788540 0.246667 0.633333 0.714992
924 0.839326 1.782442 0.250000 0.633333 0.724489
925 0.839326 1.783044 0.243333 0.633333 0.717364
926 0.839265 1.788289 0.243333 0.633333 0.726367
927 0.839301 1.798126 0.243333 0.633333 0.735492
928 0.839239 1.783418 0.246667 0.633333 0.732490
929 0.839244 1.794045 0.240000 0.633333 0.805490
930 0.839173 1.787916 0.243333 0.633333 0.714431
931 0.839192 1.792935 0.240000 0.633333 0.779989
932 0.839185 1.792569 0.240000 0.633333 0.756202
933 0.839222 1.781454 0.250000 0.650000 0.790489
934 0.839202 1.784504 0.246667 0.650000 0.800990
935 0.839190 1.784375 0.250000 0.650000 0.801991
936 0.839161 1.782186 0.246667 0.650000 0.742254
937 0.839170 1.790564 0.240000 0.633333 0.795489
938 0.839228 1.781475 0.240000 0.650000 0.718623
939 0.839174 1.790921 0.236667 0.633333 0.721991
940 0.839177 1.787340 0.250000 0.633333 0.709492
941 0.839231 1.783915 0.250000 0.633333 0.707491
942 0.839261 1.780819 0.253333 0.650000 0.711990
943 0.839158 1.782591 0.250000 0.650000 0.720991
944 0.839111 1.787482 0.236667 0.633333 0.716493
945 0.839160 1.793535 0.236667 0.633333 0.712990
946 0.839113 1.790392 0.243333 0.633333 0.698491
947 0.839115 1.787908 0.243333 0.633333 0.700246
948 0.839192 1.791934 0.233333 0.633333 0.718489
949 0.839153 1.788899 0.243333 0.633333 0.721492
950 0.839145 1.789635 0.243333 0.633333 0.715123
951 0.839193 1.787677 0.246667 0.633333 0.718453
952 0.839146 1.789029 0.243333 0.633333 0.715491
953 0.839103 1.787875 0.243333 0.633333 0.717119
954 0.839176 1.801280 0.236667 0.633333 0.711989
955 0.839082 1.792311 0.240000 0.633333 0.714312
956 0.839137 1.792779 0.243333 0.633333 0.719491
957 0.839225 1.785078 0.243333 0.633333 0.726490
958 0.839110 1.787800 0.243333 0.633333 0.709992
959 0.839130 1.787311 0.250000 0.633333 0.719315
960 0.839170 1.781849 0.253333 0.633333 0.727561
961 0.839162 1.783794 0.243333 0.633333 0.718492
962 0.839193 1.796991 0.240000 0.633333 0.715114
963 0.839089 1.793486 0.243333 0.633333 0.724491
964 0.839150 1.786736 0.250000 0.633333 0.710883
965 0.839151 1.783355 0.243333 0.633333 0.718383
966 0.839131 1.788603 0.243333 0.633333 0.715602
967 0.839091 1.795747 0.240000 0.633333 0.718571
968 0.839158 1.790060 0.243333 0.633333 0.732489
969 0.839080 1.791901 0.243333 0.633333 0.717991
970 0.839107 1.793483 0.240000 0.633333 0.713991
971 0.839169 1.794731 0.236667 0.633333 0.720490
972 0.839119 1.792233 0.240000 0.633333 0.721991
973 0.839204 1.797086 0.236667 0.633333 0.718717
974 0.839159 1.791988 0.240000 0.633333 0.733491
975 0.839064 1.787736 0.243333 0.633333 0.731795
976 0.839075 1.794297 0.240000 0.633333 0.731986
977 0.839149 1.788210 0.243333 0.633333 0.715991
978 0.839156 1.783921 0.243333 0.633333 0.715918
979 0.839188 1.785025 0.246667 0.633333 0.723993
980 0.839231 1.789402 0.236667 0.633333 0.744683
981 0.839159 1.798841 0.236667 0.633333 0.714491
982 0.839187 1.792764 0.240000 0.633333 0.725991
983 0.839166 1.795872 0.240000 0.633333 0.734991
984 0.839131 1.793638 0.240000 0.633333 0.707992
985 0.839164 1.787463 0.253333 0.633333 0.724489
986 0.839116 1.791776 0.243333 0.633333 0.711480
987 0.839175 1.784701 0.243333 0.633333 0.704154
988 0.839092 1.792227 0.236667 0.633333 0.711776
989 0.839133 1.787433 0.253333 0.633333 0.709342
990 0.839130 1.789557 0.240000 0.633333 0.720991
991 0.839104 1.786225 0.243333 0.633333 0.721491
992 0.839099 1.790042 0.243333 0.633333 0.715491
993 0.839192 1.793197 0.240000 0.633333 0.728990
994 0.839103 1.792420 0.240000 0.633333 0.716991
995 0.839128 1.792937 0.240000 0.633333 0.703079
996 0.839062 1.791280 0.243333 0.633333 0.714989
997 0.839070 1.787670 0.240000 0.633333 0.727422
998 0.839152 1.782353 0.250000 0.633333 0.719991
999 0.839113 1.785067 0.240000 0.633333 0.717491
1000 0.839105 1.788121 0.246667 0.633333 0.723993
robintibor
commented
5 years ago so you use the code in https://robintibor.github.io/braindecode/notebooks/Cropped_Decoding.html including scheduler='cosine',? or what do you use? How many timesteps are in your trials? So how many seconds are they?
robintibor
commented
5 years ago Ah and do you have reference values from other methods? to see if it could stimply be that you cannot get much better than that?
Mahelita
commented
5 years ago @FinnGu This is unrelated to your problem but:
My suggestion: cut the training to 100 epochs max until the problem is solved for this interval. This way you will have a much more efficient testing cycle. You can still increase the number of epochs when everything is converging properly (but I don't think you will need 1000).
PS: the misclass is in %/100? Meaning you have ~66% validation misclass nach 200 epochs? Which would fit with the model predicting only one class. PS': What loss are you using? PS'': I just saw that you are aware of over-fitting and training extra long to see improvements. Sorry for doubting. I still leave the comments as a documentation for other users.
FinnGu
commented
5 years ago @robintibor
so you use the code in https://robintibor.github.io/braindecode/notebooks/Cropped_Decoding.html including scheduler='cosine',?
Yes, I use that code sample, including the scheduler 'cosine'.
How many timesteps are in your trials?
Each trial consists of 512 timesteps (2 seconds). I had 1024 (4 seconds) before, but changed it to achieve better results.
do you have reference values from other methods?
Sadly, no. But by now, I have reached sufficiently good results in my valid_misclass by recording more trials and changing the trial duration from 2 to 4 seconds. Thank you for all your help!
One last thing that I cannot get to work is saving my trained model and reusing it later. I followed pytorch's tutorial for saving and loading, but when calling model.predict(trial) I get the error AttributeError: 'ShallowFBCSPNet' object has no attribute 'iterator'.
robintibor
commented
5 years ago Amazing @FinnGu ! yes changing trial duration can be also important, especially for cropped decoding as it otherwise might only train on a few samples. Loading and Saving I can imagine it doesn't work... I didn't consider about this use case the time. As a workaround, calling fit again with 0 epochs might work, but I can also have a look at a later point and change the code to make it work somehow hmhm
robintibor
commented
5 years ago Made a new issue here: https://github.com/robintibor/braindecode/issues/35
First of all, thank you for publishing this library alongside your article "Deep learning with convolutional neural networks for EEG decoding and visualization". I am trying to use your library to classify imagined movements of a subject's left hand, right hand or both hands in my bachelors thesis.
I closely followed the instructions given for Cropped Decoding, only replaced the eeg data loaded with my own recordings (I use a g.USBamp at 256Hz and record data from Cz, C3 and C4 in mV). During recording I apply a bandpass filter (0.1-60Hz) and a notch filter at 50Hz, after that no further preprocessing happens.
When training my model, the
train_lossandvalid_lossquickly decrease to around 1.1, but then do not improve any more.train_misclassandvalid_misclassare always at 0.66, because the model always predicts one of the three possible classes for all trials. The predicted class varies between different trained models, I assume this is due to the random initialization of weights.Am I missing an important step? I thought about normalizing my data, but read that your library accepts raw EEG-data. Any help would be greatly appreciated!