$ python3 n_point.py 6 points/n6_continued.json
n=6: fan 14 cones, dim 3; point n6_continued.json
  planar invariants: P(1, 3)=1/2, P(1, 4)=-2/3, P(1, 5)=-3/4, P(2, 4)=-2/5, P(2, 5)=-5/7, P(2, 6)=-3/5, P(3, 5)=-4/7, P(3, 6)=-5/8, P(4, 6)=-5/9
  axis census: continued m=1: 5, convergent: 37
  all continued (dim, kappa) classes have been validated against an independent computation (README.md "Validated continuation classes"; python3 n_point.py --list-validated)
  NOTE: n=6 P13-line point — the kinsys operator-transport fast path is available (--engine kinsys; seconds; validated, not certified).
  route: tm
  TM route: 14 cones, NGRID=8, ORD=[8], floor rad<=1e-6|mid| per cone, one NGRID+1 retry if a cone misses the floor
  per-cone: 0:+/1/+; 1:+/+/1; 2:+/1/+; 3:+/+/1; 4:+/+/+; 5:+/+/1; 6:+/+/+; 7:+/+/+; 8:+/+/+; 9:+/+/+; 10:+/+/+; 11:+/+/+; 12:+/+/+; 13:+/+/+
  PRE-RUN ESTIMATE (measured basis: the n=7 convergent-region and continued points and the n=6 and n=5 runs under validation/; NGRID scaling (n/ref)^dim and m>=2 overheads are EXTRAPOLATIONS): median-class 0.35 core-h, worst-class 0.91 core-h; + retries/tightening after a missed floor (measured at the n=7 continued point: 10.02 core-h first pass -> 19.92 core-h total, validation/n7/); at 1 core(s): 0.35-0.91 h wall; per-cone CPU limit 3600 s
dry run only; add --run --cores c1,c2,...
# exit code 0
