### Table 4. Hamiltonian cycle

"... In PAGE 7: ... As for computing Hamiltonian cycles on complete graphs, lookahead cannot reduce the depth of the search tree while bringing on extra constraint propagations. The experimental results ( Table4 ) show smodels can be several hundred times slower than smodels without lookahead. For the above two problems, A-smodels works largely as smodels without lookahead.... In PAGE 7: ... A-smodels turned off lookahead perma- nently after some literals are assigned during the process of solving this problem. For the Hamiltonian cycle problem, the performance of A-smodels is in between smodels and smodels without lookahead - it is about 20 to 30 times faster than smodels and 2 to 10 times slower than smodels without lookahead ( Table4 ). We notice that if the parameter look score is set to be 2 and ratio 0.... ..."

### Table 3: Number of Non-Hamiltonian Graphs from Gn

1998

Cited by 16

### Table 3: Number of Non-Hamiltonian Graphs from G

1998

Cited by 16

### TABLE 1. Counts of cubic graphs with additional properties. The percentages show Kotzig graphs among the hamiltonian graphs.

### Table 1: Hamiltonian path CPU times

"... In PAGE 29: ... The addConstraint(CL, CLimit) is a predicate that makes an external function call to add the list CL of constraints into the GENET solver with the resource CLimit, which may return either true, false, or unknown. Tables 1 and 2 show the performance of ECLiPSe and the different models on Hamiltonian path problems using fixed-limit strategies with limits set at 1,000; Table1 shows CPU seconds for each benchmark under the different models, and Table 2 shows the average number of backtracks in the search. The first two problems are coded from some interesting real-life examples in graph theory [Pra76].... ..."

### Table 2. Performances on the Hamiltonian cycle problem

"... In PAGE 11: ... In addition, we consider graphs far from that point: solvable instances with e = 3p=2, and unsolvable ones with e = p=2. Table2 shows our average results of 5 instances for n = 15, 17, and 20 (p = 60, 70, and 86, respectively). We note that compilation is quite fast, while the SAT solver is very slow.... ..."

### Table 1: Hamiltonian path CPU times

1998

"... In PAGE 29: ... The addConstraint(CL, CLimit) is a predicate that makes an external function call to add the list CL of constraints into the GENET solver with the resource CLimit, which may return either true; false, or unknown. Tables 1 and 2 show the performance of ECLiPSe and the di erent models on Hamiltonian path problems using xed-limit strategies with limits set at 1,000; Table1 shows CPU seconds for each benchmark under the di erent models, and Table 2 shows the average number of backtracks in the search. The rst two problems are coded from some interesting real-life examples in graph theory [Pra76].... ..."

### Table 6: Experimental and approximate theoretical values for the location of the 50% Hamiltonian point for Degreebound graphs of various sizes.

1998

"... In PAGE 20: ... A larger variance in hardness was observed with the Hamiltonian graphs. Table6 shows the distribution with respect to the number of search nodes required. Unlike Gn;m and Degreebound graphs, these graphs could not be solved in only n search nodes.... ..."

Cited by 16

### TABLE I CHARACTERISTICS OF BENCHMARK DATA FLOW GRAPHS.

2004

Cited by 9

### Table 1: Static control flow graph statistics for

"... In PAGE 3: ... Table1 shows the static number of basic blocks, block-to-block transitions, and additional transi- tion blocks required for arc-based profiling for the SPECint92 benchmarks. Surprisingly, the average number of transitions is only approximately 1.... ..."