cells: 393
control (uncorrupted, my target): rejected 0/393
corruption P x 0.9: rejected 393/393
corruption c0 - 1e-3: rejected 393/393
corruption c0 - 1e-5: rejected 393/393
corruption c6 = 0 (cells with c5 != 0): rejected 375/375
corruption u2 + 1/20: rejected 393/393
corruption a1 - 1/20 (cells with a1 >= 1/20): rejected 363/363
corruption a2 + 1/100: rejected 277/393
lowered target r2 = 1.4141: rejected 0/393
lowered target r2 = 1.414: rejected 0/393
lowered target r2 = 1.41: rejected 114/393
lowered target r2 = 1.4: rejected 274/393
smallest float slacks min_r h(r)/T (cell index, u, a):
   3.393e-05  #10  u[1.55370,1.58715] a[0.03750,0.04500]
   1.014e-04  #324  u[1.65405,1.67077] a[0.65250,0.66000]
   1.154e-04  #88  u[1.57879,1.58715] a[0.67875,0.68250]
   2.058e-04  #199  u[1.60387,1.61224] a[0.68812,0.69000]
   2.068e-04  #173  u[1.60387,1.62060] a[0.66750,0.67125]
tiling of cellsA with the first cell removed: gap detected at (1.555790625, 0.00375)
quartic corruption c4 x 100: rejected 128/128
quartic corruption c4 / 100: rejected 128/128
quartic corruption c4 x 3: rejected 128/128
quartic corruption c4 / 3: rejected 99/128
quartic boxes meeting t = sqrt2-1 moved to u in [1.74, 1.74+w] (q < 1/4): rejected 8/8
same boxes moved to u in [1.733, 1.733+w] (just above sqrt3 = 1.73205): rejected 8/8
near-dictator constant at a = 9/10: 1.2899 (>= 1, so the analytic bound cannot start at 0.9)
