### Table 1: Values of the two roots of Rn(x) for = 1 10 and = q419 Rn polynomials tend to values very near from the roots of the Melnikov function, as pointed out in [1]. But if one takes = 8 and = q419 , Melnikov theory still predicts two (circle like) limit cycles of the same radii (Melnikov function does not depend on ), while the Rn polynomials have no real root of odd multiplicity after n = 12 (see table (2)). The fact that the two real roots disappear indicates that there is no more limit cycle for = 8. 2

### Table 1. Computing speed of di erent algorithms on di erent computer architectures. \s(ingle) bit quot;, \particle quot;, \intermed(iate) quot;, and \netw(ork) quot; mean the corresponding algorithms described in the text. For each machine and algorithm, the rst table entry gives the time each computer needed to simulate a system of size 10;000 km. From this gure, we derive the other two entries: the real time limits in km and in vehicle sec/sec.

"... In PAGE 6: ... With respect to the vehicle oriented approach, the problem is the same as in parallel Molecular Dynamics approaches: How to nd the neighbors for inter- action, for example for intersection dynamics, lane changing, or car following. Solutions to this are possible but elaborate, and the expected computational speed gain (not more than a factor of about four, see Table1 ) did not warrant the additional programming complexity at the current state of the project. These observations led to a third, intermediate approach.... In PAGE 6: ... More details can be found in [17]. 6 Computational speeds on di erent supercomputers Table1 gives an overview of the computational speeds on selected computers. When comparing performance data, it is necessary to give the size of the sim- ulated system.... ..."

### Table 1. Tests in real scenario

"... In PAGE 8: ... No contacts between the robot and people or objects in the environment have been registered during the experiments. In Figure 4, we show a snapshot of the environment and of the robot, the entire map with the path covered by the robot during one of the experiments (Experiment 2 in Table1 ), and some details of the person following behavior (little squares are the robot... In PAGE 9: ...Table 1. Tests in real scenario In Table1 we show the results of the experiments in the real scenario, with total covered distance and total execution time. In the experiments the maximum robot speed was limited to 0.... ..."

### Table 1: The functions on the right are the fractional Fourier transforms of the functions on the left. j is an arbitrary integer, and and are real constants. For certain isolated values of a, the above expressions should be interpreted in the limiting sense (equation 17). In the last pair, gt; 0 is required for convergence.

1999

Cited by 1

### Table II applies this expression to existing machines and to T30, assuming that n n3d10in all cases. For the SSC [3], the real half cell length appears to be conservatively small, while RHIC is close to the limit predicted by Equation 1.

### Table 6: Lagrangean algorithm and CPLEX results - Instance set III

"... In PAGE 17: ... Actually, the largest instances of this set overestimate the size of real cellular networks. Table6 reports the results of the Lagrangean algorithm and CPLEX. The meaning of each column is the same previously defined for table 2.... ..."

### Table 3: Performance equations of a 1m m standard cell library. We applied these results to the different inverter cells of an industrial 1m m library, which performance equations are given in Table 3 together with the values of Tlimit, the maximum load allowed and the input capacitance of each template. These values constitute the first determination, at the cell level, of real buffer insertion limits allowing, by a direct inspection of loading factors on the output nodes, the definition of speed up or reconfiguration strategies as discussed later. In Table 4 we

### Table 3: Results of the generalized versions of DCFA* and BDP-PB for 4 and 5 dimensional real protein sequence sets, with minimum and maximum sequence lengths for each set indicated. Times are in seconds. Problems unsolvable by BDP-PB, due to memory limitations, are denoted by a apos;* apos;.

"... In PAGE 6: ... While DCFA* uses memory more efficiently than BDP-PB, both algorithms are limited by space, with the pro grams filling the available memory in less than an hour. Performance of the algorithms on real sequence sets is pre sented in Table3 . As for randomly-generated sequences, DCFA* is faster than BDP-PB for four and five dimensional problems.... ..."

### Table 12 Real Effective Foreign Exchange Rates

"... In PAGE 15: ... Several observers contend that the Asian current account deficits were problematic in that they were caused by a loss of competitiveness. This view is consistent with the behavior of the real exchange rate: as Table12 shows, most of the Asean-5 economies experienced appreciation relative to 1990. This tendency sharpened in late 1995, as the US dollar (to which these countries apos; currencies where de facto or de jure pegged) gained on the japanese yen.... In PAGE 16: ... 12 The third caveat comes from comparisons with other countries as well as other times, which suggest that real appreciation in Asia was not a such a large problem. Table12 shows that, in the last decade, emerging economies such as Argentina, Brazil and Chile experienced much greater appreciations, yet no crisis struck. Also, Table 12 shows that the change in the real exchange rate in the Asean-5 countries is much smaller if the reference point is taken to be 1988 or 1989 instead of 1990.... In PAGE 16: ... To that subject we now turn. 12 In particular, Chinn (1998) estimates a structural model of real exchange rate determination and finds that, once one corrects for underlying structural change, the extent of misalignment is quite limited, and smaller than the real appreciation numbers in Table12... ..."

### Table 5: Summary of predictions, data and upper limits. Column `Pred. apos; gives predictions from models [1] [2] and [3]. `Real prompt pairs apos; gives the numbers of candidate events in mass windows of J= , (2S) and . Numbers of prompt + ? and e+e? pairs are shown, with the additional conditions on b veto or tag. Columns apos;BRrate apos; and apos;BRlimit apos; give branching ratio estimates (see text) and 90% c.l. upper limits.

1995