copying files to /scratch...
starting benchmark...
/scratch/knn/venv/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
running only kgraph
order: [Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 100, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 80, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 20, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 2, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 70, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 50, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 10, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 3, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 1, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 30, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 4, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 90, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 60, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 5, {'reverse': -1}, False]), Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 40, {'reverse': -1}, False])]
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 100, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 100, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0064 accuracy: 1.59623 cost: 0.00633344 M: 10 delta: 1 time: 0.602133 one-recall: 0.01 one-ratio: 1.95911
iteration: 2 recall: 0.0544 accuracy: 0.62013 cost: 0.0100109 M: 10 delta: 0.854945 time: 0.819914 one-recall: 0.05 one-ratio: 1.47978
iteration: 3 recall: 0.358 accuracy: 0.180699 cost: 0.0153112 M: 11.543 delta: 0.827699 time: 1.09838 one-recall: 0.43 one-ratio: 1.1397
iteration: 4 recall: 0.8112 accuracy: 0.0263972 cost: 0.0212912 M: 11.9693 delta: 0.60285 time: 1.39627 one-recall: 0.83 one-ratio: 1.03613
iteration: 5 recall: 0.9496 accuracy: 0.00527856 cost: 0.0304056 M: 16.866 delta: 0.265472 time: 1.81531 one-recall: 0.97 one-ratio: 1.01155
iteration: 6 recall: 0.9828 accuracy: 0.00117401 cost: 0.0394944 M: 22.5709 delta: 0.105577 time: 2.22287 one-recall: 0.99 one-ratio: 1.00163
iteration: 7 recall: 0.9928 accuracy: 0.000380949 cost: 0.0431718 M: 24.4175 delta: 0.0840912 time: 2.4382 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 45.5
Index size:  100344.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0006066667
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0300000000, query time of that 0.0147349320, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2500000000, query time of that 0.1321213420, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Accept!
  -> Decision True in time 2.5300000000, query time of that 1.3131303310, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1500000000, query time of that 0.0153514990, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.1501256800, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
2106.83 < 2109.68
  -> Decision False in time 0.4300000000, query time of that 0.0483237970, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3800000000, query time of that 0.0168360560, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Accept!
  -> Decision True in time 13.4700000000, query time of that 0.1635552770, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1651.26 < 1660.91
  -> Decision False in time 11.0700000000, query time of that 0.1343792810, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 80, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 80, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0044 accuracy: 1.81839 cost: 0.00633344 M: 10 delta: 1 time: 6.90066 one-recall: 0 one-ratio: 2.07916
iteration: 2 recall: 0.052 accuracy: 0.699626 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1934 one-recall: 0.09 one-ratio: 1.52392
iteration: 3 recall: 0.3332 accuracy: 0.210691 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4956 one-recall: 0.33 one-ratio: 1.20405
iteration: 4 recall: 0.8208 accuracy: 0.0243909 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1882 one-recall: 0.9 one-ratio: 1.02615
iteration: 5 recall: 0.966 accuracy: 0.00248577 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8795 one-recall: 0.99 one-ratio: 1.0012
iteration: 6 recall: 0.9872 accuracy: 0.000561977 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0777 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.99 accuracy: 0.000358431 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6638 one-recall: 1 one-ratio: 1
iteration: 8 recall: 0.9912 accuracy: 0.00031344 cost: 0.0443167 M: 24.8843 delta: 0.0806719 time: 35.5699 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.91999999999999
Index size:  92232.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0012200000
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0108588320, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2300000000, query time of that 0.1070911310, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Accept!
  -> Decision True in time 2.2500000000, query time of that 1.0440084470, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1400000000, query time of that 0.0120086230, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1899.09 < 2004.21
  -> Decision False in time 0.4300000000, query time of that 0.0365779400, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Accept!
  -> Decision True in time 13.4700000000, query time of that 1.1835176310, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0139835620, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Accept!
  -> Decision True in time 13.4000000000, query time of that 0.1338248910, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1741.12 < 1773.8
  -> Decision False in time 10.6100000000, query time of that 0.1032954050, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 20, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 20, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.004 accuracy: 1.75774 cost: 0.00633344 M: 10 delta: 1 time: 6.88388 one-recall: 0 one-ratio: 2.03464
iteration: 2 recall: 0.0608 accuracy: 0.646443 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1769 one-recall: 0.09 one-ratio: 1.47702
iteration: 3 recall: 0.3568 accuracy: 0.180946 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.479 one-recall: 0.38 one-ratio: 1.17399
iteration: 4 recall: 0.8284 accuracy: 0.0237156 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1706 one-recall: 0.86 one-ratio: 1.02268
iteration: 5 recall: 0.9608 accuracy: 0.00313867 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8631 one-recall: 0.97 one-ratio: 1.00167
iteration: 6 recall: 0.9844 accuracy: 0.000793035 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0632 one-recall: 0.99 one-ratio: 1.00002
iteration: 7 recall: 0.9908 accuracy: 0.000356182 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6493 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.01000000000002
Index size:  90960.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0078483333
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0062258840, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
2103.6 < 2202.51
  -> Decision False in time 0.0600000000, query time of that 0.0179339160, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1941.8 < 2079.44
  -> Decision False in time 0.1900000000, query time of that 0.0584469860, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0065340000, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1883.27 < 1884.83
  -> Decision False in time 0.4300000000, query time of that 0.0212109220, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
2044.9 < 2114.8
  -> Decision False in time 0.5400000000, query time of that 0.0255709410, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3600000000, query time of that 0.0074150790, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
2019.54 < 2036.62
  -> Decision False in time 3.4900000000, query time of that 0.0199617690, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1592.97 < 1787.58
  -> Decision False in time 3.1100000000, query time of that 0.0165907120, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 2, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 2, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.006 accuracy: 1.65471 cost: 0.00633344 M: 10 delta: 1 time: 6.87629 one-recall: 0.01 one-ratio: 2.0093
iteration: 2 recall: 0.0564 accuracy: 0.608626 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1695 one-recall: 0.12 one-ratio: 1.44684
iteration: 3 recall: 0.3732 accuracy: 0.17047 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4734 one-recall: 0.44 one-ratio: 1.13883
iteration: 4 recall: 0.8276 accuracy: 0.0241551 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1691 one-recall: 0.9 one-ratio: 1.02079
iteration: 5 recall: 0.9644 accuracy: 0.00292808 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.867 one-recall: 0.98 one-ratio: 1.00081
iteration: 6 recall: 0.986 accuracy: 0.000854818 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.074 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.992 accuracy: 0.000590742 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6627 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.01999999999998
Index size:  90968.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0388550000
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0100000000, query time of that 0.0040348080, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1970.56 < 2203.42
  -> Decision False in time 0.0100000000, query time of that 0.0024783790, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
2243.07 < 2279.38
  -> Decision False in time 0.0600000000, query time of that 0.0149164290, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Reject!
1988.42 < 2114.79
  -> Decision False in time 0.0100000000, query time of that 0.0001611800, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
2223.33 < 2272.2
  -> Decision False in time 0.0000000000, query time of that 0.0003109310, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1988.29 < 2311.92
  -> Decision False in time 0.0800000000, query time of that 0.0029522630, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Reject!
2146.35 < 2209.71
  -> Decision False in time 0.8200000000, query time of that 0.0036538510, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1840.59 < 2310.86
  -> Decision False in time 2.5200000000, query time of that 0.0110250580, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
2565.23 < 2604.64
  -> Decision False in time 0.6900000000, query time of that 0.0029483450, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 70, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 70, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0036 accuracy: 1.7649 cost: 0.00633344 M: 10 delta: 1 time: 6.87294 one-recall: 0 one-ratio: 2.10464
iteration: 2 recall: 0.0604 accuracy: 0.635748 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.164 one-recall: 0.06 one-ratio: 1.52118
iteration: 3 recall: 0.3776 accuracy: 0.169024 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4661 one-recall: 0.46 one-ratio: 1.20421
iteration: 4 recall: 0.8328 accuracy: 0.0231786 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1599 one-recall: 0.89 one-ratio: 1.04324
iteration: 5 recall: 0.9596 accuracy: 0.00447756 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8503 one-recall: 0.97 one-ratio: 1.00944
iteration: 6 recall: 0.9812 accuracy: 0.00168385 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0493 one-recall: 0.99 one-ratio: 1.00557
iteration: 7 recall: 0.9892 accuracy: 0.000901436 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6341 one-recall: 0.99 one-ratio: 1.00557
iteration: 8 recall: 0.992 accuracy: 0.000417126 cost: 0.0443167 M: 24.8843 delta: 0.0806719 time: 35.5392 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.900000000000034
Index size:  92232.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0012433333
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0106928610, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2200000000, query time of that 0.0991697350, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1908.17 < 1936.86
  -> Decision False in time 0.5200000000, query time of that 0.2296217860, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0114017220, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.3300000000, query time of that 0.1107393990, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1714.77 < 1935.29
  -> Decision False in time 12.0400000000, query time of that 0.9697488040, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0121411550, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Accept!
  -> Decision True in time 13.4900000000, query time of that 0.1248718030, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1523.44 < 1935.29
  -> Decision False in time 7.4800000000, query time of that 0.0708808670, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 50, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 50, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0044 accuracy: 1.68691 cost: 0.00633344 M: 10 delta: 1 time: 6.8952 one-recall: 0.02 one-ratio: 1.95821
iteration: 2 recall: 0.0608 accuracy: 0.628781 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1871 one-recall: 0.12 one-ratio: 1.41714
iteration: 3 recall: 0.398 accuracy: 0.170086 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4899 one-recall: 0.48 one-ratio: 1.09856
iteration: 4 recall: 0.8544 accuracy: 0.0205763 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1829 one-recall: 0.9 one-ratio: 1.01567
iteration: 5 recall: 0.9676 accuracy: 0.00268387 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8755 one-recall: 0.99 one-ratio: 1.00065
iteration: 6 recall: 0.988 accuracy: 0.000656383 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0783 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.9912 accuracy: 0.000477346 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6625 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.01999999999998
Index size:  90964.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0030666667
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0081132740, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2000000000, query time of that 0.0801107930, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
2029.65 < 2234.01
  -> Decision False in time 0.1100000000, query time of that 0.0405371960, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0087637910, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.3100000000, query time of that 0.0919730420, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1965.16 < 1978.79
  -> Decision False in time 0.3700000000, query time of that 0.0260574460, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0108714790, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Accept!
  -> Decision True in time 13.4000000000, query time of that 0.1042542850, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1932.66 < 1978.79
  -> Decision False in time 25.6600000000, query time of that 0.1975908430, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 10, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 10, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.008 accuracy: 1.74437 cost: 0.00633344 M: 10 delta: 1 time: 6.87778 one-recall: 0 one-ratio: 2.03444
iteration: 2 recall: 0.0652 accuracy: 0.637073 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1717 one-recall: 0.06 one-ratio: 1.4837
iteration: 3 recall: 0.3984 accuracy: 0.166915 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4746 one-recall: 0.49 one-ratio: 1.13879
iteration: 4 recall: 0.862 accuracy: 0.0177386 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1706 one-recall: 0.92 one-ratio: 1.01668
iteration: 5 recall: 0.976 accuracy: 0.00199604 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8686 one-recall: 0.99 one-ratio: 1.00077
iteration: 6 recall: 0.99 accuracy: 0.000758827 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0772 one-recall: 0.99 one-ratio: 1.00077
iteration: 7 recall: 0.9932 accuracy: 0.000473901 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6646 one-recall: 0.99 one-ratio: 1.00077
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.01999999999998
Index size:  90980.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0154166667
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0051978360, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1725.19 < 1753.33
  -> Decision False in time 0.0100000000, query time of that 0.0056118130, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1867.25 < 1908.69
  -> Decision False in time 0.0300000000, query time of that 0.0059516560, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Reject!
1564.35 < 1941.07
  -> Decision False in time 0.0900000000, query time of that 0.0043637680, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1962.08 < 2047.77
  -> Decision False in time 0.2900000000, query time of that 0.0123589720, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1929.97 < 2043.78
  -> Decision False in time 0.5900000000, query time of that 0.0260577190, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3600000000, query time of that 0.0065729000, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1970.59 < 2036.26
  -> Decision False in time 0.0100000000, query time of that 0.0003228030, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1951.16 < 2079.7
  -> Decision False in time 2.0000000000, query time of that 0.0101523010, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 3, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 3, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.004 accuracy: 1.49738 cost: 0.00633344 M: 10 delta: 1 time: 6.8811 one-recall: 0 one-ratio: 1.99571
iteration: 2 recall: 0.0468 accuracy: 0.631223 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1719 one-recall: 0.02 one-ratio: 1.52908
iteration: 3 recall: 0.3512 accuracy: 0.191174 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4741 one-recall: 0.32 one-ratio: 1.17563
iteration: 4 recall: 0.8116 accuracy: 0.0284491 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1665 one-recall: 0.86 one-ratio: 1.02276
iteration: 5 recall: 0.9532 accuracy: 0.00452028 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8596 one-recall: 0.99 one-ratio: 1.00003
iteration: 6 recall: 0.9824 accuracy: 0.000906052 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0619 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.9884 accuracy: 0.000479649 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6451 one-recall: 1 one-ratio: 1
iteration: 8 recall: 0.9908 accuracy: 0.000330768 cost: 0.0443167 M: 24.8843 delta: 0.0806719 time: 35.5503 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.91000000000008
Index size:  92232.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0251566667
  Testing...
|S| = 20
|T| = 283
Reject!
1709.13 < 1980.18
  -> Decision False in time 0.0000000000, query time of that 0.0003121120, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1416.27 < 1650.93
  -> Decision False in time 0.0800000000, query time of that 0.0195364940, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
2069.62 < 2200.97
  -> Decision False in time 0.0000000000, query time of that 0.0012429480, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Reject!
2016.88 < 2025.41
  -> Decision False in time 0.1100000000, query time of that 0.0043429360, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1790.11 < 1904.89
  -> Decision False in time 0.2500000000, query time of that 0.0100641360, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1515.78 < 1543.44
  -> Decision False in time 0.4200000000, query time of that 0.0159818490, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Reject!
1669.95 < 1735.64
  -> Decision False in time 0.7800000000, query time of that 0.0039432210, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1338.07 < 1440.36
  -> Decision False in time 2.1200000000, query time of that 0.0098488880, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
2042.86 < 2092.12
  -> Decision False in time 2.6200000000, query time of that 0.0116851580, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 1, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 1, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0072 accuracy: 1.78614 cost: 0.00633344 M: 10 delta: 1 time: 6.87946 one-recall: 0.01 one-ratio: 2.04356
iteration: 2 recall: 0.0672 accuracy: 0.657772 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1726 one-recall: 0.08 one-ratio: 1.47674
iteration: 3 recall: 0.3812 accuracy: 0.178449 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4744 one-recall: 0.38 one-ratio: 1.16785
iteration: 4 recall: 0.878799 accuracy: 0.0154212 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.167 one-recall: 0.93 one-ratio: 1.01494
iteration: 5 recall: 0.9824 accuracy: 0.00111758 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8607 one-recall: 0.99 one-ratio: 1.00036
iteration: 6 recall: 0.9956 accuracy: 0.000240658 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0638 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 32.41999999999996
Index size:  86668.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0517233333
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0100000000, query time of that 0.0047100740, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1692.18 < 1703.53
  -> Decision False in time 0.0100000000, query time of that 0.0014134030, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
2206.75 < 2274.61
  -> Decision False in time 0.0400000000, query time of that 0.0104034710, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Reject!
1656.24 < 1977.9
  -> Decision False in time 0.0300000000, query time of that 0.0013228500, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1656.22 < 1872.92
  -> Decision False in time 0.0200000000, query time of that 0.0010093530, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
2201.38 < 2226.83
  -> Decision False in time 0.4700000000, query time of that 0.0182891740, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Reject!
1898.95 < 1922.81
  -> Decision False in time 0.4800000000, query time of that 0.0022453600, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1799.77 < 2189.12
  -> Decision False in time 0.0000000000, query time of that 0.0002479650, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1518.91 < 1860.49
  -> Decision False in time 0.4900000000, query time of that 0.0024552650, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 30, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 30, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0048 accuracy: 1.58384 cost: 0.00633344 M: 10 delta: 1 time: 6.87725 one-recall: 0.01 one-ratio: 2.0718
iteration: 2 recall: 0.0596 accuracy: 0.628493 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1713 one-recall: 0.07 one-ratio: 1.47418
iteration: 3 recall: 0.3788 accuracy: 0.170552 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4724 one-recall: 0.5 one-ratio: 1.15294
iteration: 4 recall: 0.8484 accuracy: 0.0207751 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1652 one-recall: 0.92 one-ratio: 1.01538
iteration: 5 recall: 0.9736 accuracy: 0.00176498 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.857 one-recall: 1 one-ratio: 1
iteration: 6 recall: 0.9896 accuracy: 0.000550482 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0608 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.9944 accuracy: 0.000259975 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6475 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.0
Index size:  90956.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0065250000
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0068109780, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1670.75 < 1712.27
  -> Decision False in time 0.0600000000, query time of that 0.0193601000, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1922.47 < 2229.55
  -> Decision False in time 0.0700000000, query time of that 0.0254843790, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0073141940, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1808.3 < 2210.02
  -> Decision False in time 0.7400000000, query time of that 0.0417362180, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1921.41 < 2003.77
  -> Decision False in time 0.1800000000, query time of that 0.0099997770, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0086131290, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1751.9 < 1826.1
  -> Decision False in time 12.0200000000, query time of that 0.0773336640, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1889.06 < 1991.16
  -> Decision False in time 1.3200000000, query time of that 0.0086150670, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 4, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 4, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.008 accuracy: 1.72963 cost: 0.00633344 M: 10 delta: 1 time: 6.88649 one-recall: 0 one-ratio: 2.01043
iteration: 2 recall: 0.066 accuracy: 0.635277 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1779 one-recall: 0.1 one-ratio: 1.42655
iteration: 3 recall: 0.4096 accuracy: 0.162884 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4802 one-recall: 0.55 one-ratio: 1.08928
iteration: 4 recall: 0.8664 accuracy: 0.0168036 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1722 one-recall: 0.94 one-ratio: 1.01843
iteration: 5 recall: 0.9752 accuracy: 0.00155631 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8649 one-recall: 1 one-ratio: 1
iteration: 6 recall: 0.992 accuracy: 0.000346113 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0656 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 32.39999999999998
Index size:  86660.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0310166667
  Testing...
|S| = 20
|T| = 283
Reject!
2120.87 < 2411.61
  -> Decision False in time 0.0100000000, query time of that 0.0035686610, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1464.33 < 2199.94
  -> Decision False in time 0.1100000000, query time of that 0.0276854460, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1836.26 < 1879.03
  -> Decision False in time 0.0400000000, query time of that 0.0101249600, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Reject!
2107.84 < 2114.67
  -> Decision False in time 0.1300000000, query time of that 0.0044291340, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
1915.58 < 2043.99
  -> Decision False in time 0.2400000000, query time of that 0.0092478450, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1925.46 < 2028.2
  -> Decision False in time 0.2100000000, query time of that 0.0082221900, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Reject!
1701.2 < 2061.69
  -> Decision False in time 0.6200000000, query time of that 0.0028884020, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1820.68 < 1912.66
  -> Decision False in time 0.3400000000, query time of that 0.0018336500, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1995.32 < 2047.3
  -> Decision False in time 3.3300000000, query time of that 0.0148547200, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 90, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 90, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.004 accuracy: 1.5407 cost: 0.00633344 M: 10 delta: 1 time: 6.89169 one-recall: 0 one-ratio: 2.00156
iteration: 2 recall: 0.0588 accuracy: 0.610834 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.184 one-recall: 0.06 one-ratio: 1.45495
iteration: 3 recall: 0.358 accuracy: 0.174056 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4858 one-recall: 0.39 one-ratio: 1.14383
iteration: 4 recall: 0.8292 accuracy: 0.0240465 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1796 one-recall: 0.87 one-ratio: 1.02334
iteration: 5 recall: 0.964 accuracy: 0.00325969 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8714 one-recall: 0.99 one-ratio: 1.00319
iteration: 6 recall: 0.9868 accuracy: 0.000732637 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0724 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.9916 accuracy: 0.0004699 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6571 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.01999999999998
Index size:  90968.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0010516667
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0300000000, query time of that 0.0126501470, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2300000000, query time of that 0.1095713050, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
1909.93 < 1957.19
  -> Decision False in time 1.3800000000, query time of that 0.6599747610, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1400000000, query time of that 0.0130926590, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.3400000000, query time of that 0.1223643460, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1632.1 < 1697.82
  -> Decision False in time 0.9700000000, query time of that 0.0913909170, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0137583520, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Accept!
  -> Decision True in time 13.4100000000, query time of that 0.1360291490, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1802.91 < 1842.08
  -> Decision False in time 9.3100000000, query time of that 0.0937683690, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 60, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 60, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.004 accuracy: 1.6918 cost: 0.00633344 M: 10 delta: 1 time: 6.88308 one-recall: 0 one-ratio: 2.12757
iteration: 2 recall: 0.0552 accuracy: 0.670215 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1748 one-recall: 0.09 one-ratio: 1.58026
iteration: 3 recall: 0.38 accuracy: 0.177198 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4767 one-recall: 0.48 one-ratio: 1.17305
iteration: 4 recall: 0.846 accuracy: 0.0222695 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1684 one-recall: 0.92 one-ratio: 1.02262
iteration: 5 recall: 0.9636 accuracy: 0.00320592 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8594 one-recall: 0.98 one-ratio: 1.00458
iteration: 6 recall: 0.9852 accuracy: 0.00131027 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.058 one-recall: 0.98 one-ratio: 1.00458
iteration: 7 recall: 0.9912 accuracy: 0.000846318 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6414 one-recall: 0.99 one-ratio: 1.00454
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.0
Index size:  90976.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0020716667
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0300000000, query time of that 0.0095860830, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.2000000000, query time of that 0.0856953720, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Accept!
  -> Decision True in time 2.0900000000, query time of that 0.8734833610, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1400000000, query time of that 0.0103577940, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.3100000000, query time of that 0.0974100630, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
2198.89 < 2327.3
  -> Decision False in time 1.6000000000, query time of that 0.1195625840, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0109712200, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1935.11 < 1986.4
  -> Decision False in time 6.6000000000, query time of that 0.0574034950, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1921.92 < 1925.36
  -> Decision False in time 32.2600000000, query time of that 0.2634686440, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 5, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 5, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0056 accuracy: 1.41157 cost: 0.00633344 M: 10 delta: 1 time: 6.88879 one-recall: 0.01 one-ratio: 1.99155
iteration: 2 recall: 0.0544 accuracy: 0.60534 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1803 one-recall: 0.04 one-ratio: 1.45143
iteration: 3 recall: 0.3392 accuracy: 0.190145 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4815 one-recall: 0.37 one-ratio: 1.15719
iteration: 4 recall: 0.7912 accuracy: 0.03203 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1741 one-recall: 0.87 one-ratio: 1.01805
iteration: 5 recall: 0.9604 accuracy: 0.00284024 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8656 one-recall: 1 one-ratio: 1
iteration: 6 recall: 0.9828 accuracy: 0.00106873 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0669 one-recall: 1 one-ratio: 1
iteration: 7 recall: 0.99 accuracy: 0.000613841 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6509 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.00999999999999
Index size:  90968.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0258416667
  Testing...
|S| = 20
|T| = 283
Reject!
2341.72 < 2354.25
  -> Decision False in time 0.0000000000, query time of that 0.0025684920, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Reject!
1862.36 < 1948.85
  -> Decision False in time 0.0100000000, query time of that 0.0014143140, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Reject!
2469.96 < 2498.82
  -> Decision False in time 0.0500000000, query time of that 0.0132756800, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0043211610, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Reject!
2141.56 < 2451.65
  -> Decision False in time 0.8000000000, query time of that 0.0298399580, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
1838.71 < 1922.07
  -> Decision False in time 0.1900000000, query time of that 0.0077728330, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Reject!
2380.24 < 2444.46
  -> Decision False in time 0.2700000000, query time of that 0.0015829370, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
1709.83 < 1765.68
  -> Decision False in time 2.1800000000, query time of that 0.0098247970, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
2106.9 < 2394.96
  -> Decision False in time 2.3600000000, query time of that 0.0110375720, with c1=5.0000000000, c2=0.1000000000
Definition(algorithm='kgraph', constructor='KGraph', module='ann_benchmarks.algorithms.kgraph', docker_tag='ann-benchmarks-kgraph', arguments=['euclidean', 40, {'reverse': -1}, False]) ...
Trying to instantiate ann_benchmarks.algorithms.kgraph.KGraph(['euclidean', 40, {'reverse': -1}, False])
Got a train set of size (60000 * 784)
Generating control...
Initializing...
iteration: 1 recall: 0.0068 accuracy: 1.63493 cost: 0.00633344 M: 10 delta: 1 time: 6.88801 one-recall: 0.01 one-ratio: 2.05973
iteration: 2 recall: 0.0572 accuracy: 0.643555 cost: 0.0100076 M: 10 delta: 0.854952 time: 10.1802 one-recall: 0.09 one-ratio: 1.46203
iteration: 3 recall: 0.3896 accuracy: 0.166479 cost: 0.0152986 M: 11.5308 delta: 0.828164 time: 14.4833 one-recall: 0.48 one-ratio: 1.15249
iteration: 4 recall: 0.850399 accuracy: 0.0212351 cost: 0.0212851 M: 11.9674 delta: 0.604399 time: 19.1747 one-recall: 0.88 one-ratio: 1.02536
iteration: 5 recall: 0.9692 accuracy: 0.00334765 cost: 0.0303842 M: 16.8423 delta: 0.26712 time: 25.8682 one-recall: 0.97 one-ratio: 1.00691
iteration: 6 recall: 0.9844 accuracy: 0.00194847 cost: 0.0394649 M: 22.5291 delta: 0.106382 time: 32.0681 one-recall: 0.99 one-ratio: 1.00353
iteration: 7 recall: 0.9884 accuracy: 0.00121479 cost: 0.0431501 M: 24.3677 delta: 0.0850189 time: 34.6521 one-recall: 1 one-ratio: 1
iteration: 8 recall: 0.9932 accuracy: 0.000294001 cost: 0.0443167 M: 24.8843 delta: 0.0806719 time: 35.5576 one-recall: 1 one-ratio: 1
Graph completion with reverse edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Reranking edges...

0%   10   20   30   40   50   60   70   80   90   100%
|----|----|----|----|----|----|----|----|----|----|
***************************************************
Built index in 35.919999999999845
Index size:  92244.0
Run 1/1...
  Calculating distance...
  -> Distance: 0.0021300000
  Testing...
|S| = 20
|T| = 283
Accept!
  -> Decision True in time 0.0200000000, query time of that 0.0083861650, with c1=0.0500000000, c2=0.0010000000
|S| = 196
|T| = 283
Accept!
  -> Decision True in time 0.1900000000, query time of that 0.0726643740, with c1=0.0500000000, c2=0.0100000000
|S| = 1960
|T| = 283
Accept!
  -> Decision True in time 1.9400000000, query time of that 0.7258897110, with c1=0.0500000000, c2=0.1000000000
|S| = 20
|T| = 2821
Accept!
  -> Decision True in time 0.1300000000, query time of that 0.0076574930, with c1=0.5000000000, c2=0.0010000000
|S| = 196
|T| = 2821
Accept!
  -> Decision True in time 1.2800000000, query time of that 0.0819842140, with c1=0.5000000000, c2=0.0100000000
|S| = 1960
|T| = 2821
Reject!
2227.63 < 2262.16
  -> Decision False in time 1.4900000000, query time of that 0.0937385340, with c1=0.5000000000, c2=0.1000000000
|S| = 20
|T| = 28201
Accept!
  -> Decision True in time 1.3700000000, query time of that 0.0091203130, with c1=5.0000000000, c2=0.0010000000
|S| = 196
|T| = 28201
Reject!
572.798 < 585.885
  -> Decision False in time 8.4800000000, query time of that 0.0608905330, with c1=5.0000000000, c2=0.0100000000
|S| = 1960
|T| = 28201
Reject!
1276.86 < 1371.6
  -> Decision False in time 7.5600000000, query time of that 0.0552507350, with c1=5.0000000000, c2=0.1000000000
