================================================================================================
READOUT-LOCALITY SCISSOR STUDY  (depth-1 IQP; shared splits; symmetric nested-CV)
================================================================================================

Phase A -- per-dataset scissor (advantage at each locality k; expect adv <= 0 everywhere)
------------------------------------------------------------------------------------------------
dataset       q   readout   dim    acc  advAcc      p  ktaGap  concentr
  running haberman ...
haberman      3       k=1     9  0.729  +0.003   0.31  -0.001  1.23e-01
haberman      3       k=2    36  0.732  +0.007   0.09  -0.000  7.67e-02
haberman      3       k=3    63  0.732  +0.007   0.34  -0.000  7.47e-02
haberman      3  fidelity     8  0.745  +0.020   0.10  -0.020  5.72e-02
  running blood ...
blood         4       k=1    12  0.775  +0.010   0.05  -0.026  3.40e-02
blood         4       k=2    66  0.775  +0.010   0.09  -0.018  5.72e-02
blood         4       k=3   174  0.783  +0.018   0.03  -0.017  7.10e-02
blood         4       k=4   255  0.780  +0.015   0.07  -0.017  7.53e-02
blood         4  fidelity    16  0.767  +0.002   0.41  -0.019  6.62e-02
  running diabetes ...
diabetes      6       k=1    18  0.642  -0.122   0.99  -0.101  8.82e-02
diabetes      6       k=2   153  0.645  -0.120   0.99  -0.075  2.95e-02
diabetes      6       k=3   693  0.655  -0.110   0.99  -0.068  1.83e-02
diabetes      6       k=4  1908  0.648  -0.117   0.99  -0.066  1.32e-02
diabetes      6  fidelity    64  0.668  -0.097   0.99  -0.066  9.68e-03
  running ilpd ...
ilpd          6       k=1    18  0.718  +0.010   0.15  -0.025  8.49e-02
ilpd          6       k=2   153  0.715  +0.008   0.21  -0.027  3.83e-02
ilpd          6       k=3   693  0.718  +0.010   0.12  -0.028  2.72e-02
ilpd          6       k=4  1908  0.712  +0.005   0.31  -0.028  2.19e-02
ilpd          6  fidelity    64  0.708  +0.000   0.50  -0.038  1.81e-02
  running breast_cancer ...
breast_cancer 6       k=1    18  0.630  -0.327   1.00  -0.565  1.14e-01
breast_cancer 6       k=2   153  0.665  -0.292   1.00  -0.541  1.43e-02
breast_cancer 6       k=3   693  0.672  -0.285   1.00  -0.531  3.76e-03
breast_cancer 6       k=4  1908  0.635  -0.322   1.00  -0.535  1.64e-03
breast_cancer 6  fidelity    64  0.688  -0.270   1.00  -0.541  8.87e-04
  running ionosphere ...
ionosphere    6       k=1    18  0.710  -0.210   1.00  -0.088  1.23e-01
ionosphere    6       k=2   153  0.758  -0.162   1.00  -0.077  2.88e-02
ionosphere    6       k=3   693  0.789  -0.131   1.00  -0.101  1.34e-02
ionosphere    6       k=4  1908  0.749  -0.171   1.00  -0.117  9.00e-03
ionosphere    6  fidelity    64  0.792  -0.128   0.99  -0.117  7.00e-03
  running qsar_biodeg ...
qsar_biodeg   6       k=1    18  0.675  -0.202   1.00  -0.166  1.06e-01
qsar_biodeg   6       k=2   153  0.717  -0.160   1.00  -0.152  1.76e-02
qsar_biodeg   6       k=3   693  0.722  -0.155   1.00  -0.149  6.91e-03
qsar_biodeg   6       k=4  1908  0.708  -0.170   1.00  -0.152  3.87e-03
qsar_biodeg   6  fidelity    64  0.772  -0.105   1.00  -0.151  2.48e-03
  running sonar ...
sonar         6       k=1    18  0.577  -0.264   1.00  -0.072  9.41e-02
sonar         6       k=2   153  0.500  -0.342   1.00  -0.064  1.03e-02
sonar         6       k=3   693  0.591  -0.250   1.00  -0.062  2.23e-03
sonar         6       k=4  1908  0.543  -0.298   1.00  -0.063  8.58e-04
sonar         6  fidelity    64  0.587  -0.255   1.00  -0.062  4.32e-04

  significant advantage (adv>0 & p<0.05) at any locality on any dataset: 2 of 39 (dataset x readout) cells
Saved -> results/scissor_phaseA.csv
