### Table 12. Recursive declarations: rec D

909

"... In PAGE 12: ... The constructs that we have specified so far allow the declaration of (call by value) functions, using a combination of value-declarations and value-abstractions, but they do not (directly) support recursive function declarations. The construct recursive(D) specified in Table12 has the effect of making ordinary function declarations D recursive (it could easily be extended to mutually-recursive dec- larations). The auxiliary operation unfold(D,B) is defined by the (equational) rule (31).... ..."

### Table 1-1. Stale data errors.

1994

"... In PAGE 22: ... More information about theses traces can be found in [Baker91b]. Table1 -1 presents the results of these simulations. A 60-second refresh interval would have resulted in many uses of stale data each hour, and one-half of all users would have accessed stale... In PAGE 24: ... The same trace data from Sprite helps answer the second question: whether consistency actions are invoked often enough to degrade overall performance. Table1 -2 presents this data. About one in every 300 file opens causes a file to be opened for reading and writing on more than one machine.... In PAGE 25: ... This thesis evaluates three such methods for recovering distributed state, all implemented in Sprite. The first method, described in chapter 3, is client-driven recovery, in which clients initiate Table1 -2. Consistency action frequency.... ..."

### Table 9: Recursive Computation of Tn(x)

1997

"... In PAGE 5: ...aple V.3/V.4 uses the three-term recurrence equation to obtain the collection of polyno- mials Tk(x) (k = 0; : : : ; n). Table9 of x 7 gives a fair comparison for this approach between the six systems, which shows that for large n 2 N, Mathematica is faster in this case and can compute a larger list than Maple. However, since the memory and storage requirements are so immense, we think that an e cient computation of a single Tn(x) is the most important task.... In PAGE 10: ...Table 9: Recursive Computation of Tn(x) Table9 shows the timings for this approach. REDUCE generates variable stack over ow since it does not have a remember feature.... ..."

Cited by 2

### Table 6. DTP efficiency by varying proportional distribution of patients at different compliance levels amp; case-finding efficiency

"... In PAGE 15: ... For example, the DTP efficiency of SR would then increase from 16.4% at the current rates (shown in column 2 of Table6 ) to 24.2%, when all the cases comply at level 4 (column 8, Table 6) with CF efficiency remaining at 33%.... In PAGE 15: ...4% at the current rates (shown in column 2 of Table 6) to 24.2%, when all the cases comply at level 4 (column 8, Table6 ) with CF efficiency remaining at 33%. The DTP efficiency would be three times the above figure (49.... In PAGE 15: ...7% for the current compliance structure and 73.3% when all the cases comply at level 4) if the CF efficiency is 100% (last row of Table6 , Part a). ... In PAGE 16: ...respective sputum conversion rates for compliance structure given in columns 2 to 8 given in Appendix Table.; ** Structure of compliance:-please see Appendix Table with SCC is calculated, except that improvement with SCC over SR, when rates of compliance are increased to 100% at level 4, increases only from 24% to 29% at CF efficiency of 33% (row 1 of Table6 Part b). The DTP efficiency at 100% CF efficiency improves to nearly three times (78.... In PAGE 16: ... The DTP efficiency at 100% CF efficiency improves to nearly three times (78.9%) the value attainable for CF efficiency at 33%, at the current rates of compliance (column 2, last rowof Table6 b). Compliance, regimen mixes, CF efficiency and DTP efficiency Another factor that could be studied through the model is the effect of varying the mixes of relative proportions of cases on SR and SCC.... ..."

### Table 3: Input Parameters to our RTDBS Model This metric is an indicator of the QT processing e ciency of the system. Fraction of Stale Reads (FSR): A read is stale when it is given access to a non-recent value. This metric measures what fraction of all reads issued in the system accessed stale values, and is computed as:

in Databases for Active, Rapidly, Changing data Systems (ARCS) and Handling Sensor Data in such Systems

1996

Cited by 3

### Table 2. Performance values of the synchronization protocols

1998

"... In PAGE 5: ... Continuous values are also monitored by means of trace file outputs that provide detailed in- formation about the internal node voltages in the analog islands. The timings presented in Table2 are the average of three different runs. Averaging was done to reduce the effect of system load fluctuations.... In PAGE 5: ... Figure 9 illustrates the process distribution for the two clock generators example. Table2 presents a comparison of the different synchronization protocols. The effect of increased computation power and communication cost on the synchronization protocols can be Clock Generator on two processors Execution Time (sec) ext.... In PAGE 5: ... The preliminary performance evaluation illustrates some interesting facts about the individual protocols. The performance values illustrated in Table2 , favor the SES protocol when it is a multiprocessor simulation. In most cases, the SES protocol performs as good as or better than the FES protocol.... ..."

Cited by 6

### Table 2. Performance values of the synchronization protocols

1998

"... In PAGE 5: ... Continuous values are also monitored by means of trace file outputs that provide detailed in- formation about the internal node voltages in the analog islands. The timings presented in Table2 are the average of three different runs. Averaging was done to reduce the effect of system load fluctuations.... In PAGE 5: ... Figure 9 illustrates the process distribution for the two clock generators example. Table2 presents a comparison of the different synchronization protocols. The effect of increased computation power and communication cost on the synchronization protocols can be Clock Generator on two processors Execution Time (sec) ext.... In PAGE 5: ... The preliminary performance evaluation illustrates some interesting facts about the individual protocols. The performance values illustrated in Table2 , favor the SES protocol when it is a multiprocessor simulation. In most cases, the SES protocol performs as good as or better than the FES protocol.... ..."

Cited by 6

### Table 1 Index Vector by Recursion

1996

"... In PAGE 3: ... For example, weknow that the last n ; 2 bits of ~ k j are zero for k j lt; 4. This approach leads to the recursive algorithm illustrated in Table1 [42]. At eachstepwemultiply the currentlistby 2 and concatenate 1 plus the list.... In PAGE 26: ... Elster instead uses the \add half quot; procedure which involves halving an integer instead of doubling an array.Table 14 can be compared with Table1 to see the di erence in the way the index vector is built. Since Elster apos;s approach never changes an entry once it has been computed, it saves arithmetic and time, as shown in Table 15.... In PAGE 26: ... Since Elster apos;s approach never changes an entry once it has been computed, it saves arithmetic and time, as shown in Table 15. Elster apos;s algorithm and that in Table1 are both special cases of a more general formulation that comes from the idea of a tensor sum[33]. The tensor sum of two vectors w = u v is de ned as w =[u + v 0 ;;u+ v 1 ;; .... ..."

Cited by 8

### Table 1: Order of computations required. Brute force Recursive

in The inner product matrix of discrete autocorrelation wavelets: efficient computation and application

2001

"... In PAGE 2: ... A recursive scheme for the calculation of those (initialising) elements which lie upon the top row of the inner product matrix is proposed in section 4. The computational efficiency of the various schemes considered in sections 2 and 3 are compared to the brute-force alternatives in Table1 . Finally, motivated by recent work on the modelling of two-dimensional random fields (Eckley, Nason and Treloar [7]), we propose a recursive construction for the inner product matrix of (separable) two-dimensional discrete autocorrelation wavelets.... ..."

### Table 1: Order of computations required. Brute force Recursive

in The Inner Product Matrix of Discrete Autocorrelation Wavelets: Efficient Computation and Application

"... In PAGE 2: ...ying on a given diagonal. Figure 1 illustrates our recursive scheme. A recursive scheme for the calculation of those (initialising) elements which lie upon the top row of the inner product matrix is proposed in section 4. The computational efficiency of the various schemes considered in sections 2 and 3 are compared to the brute-force alternatives in Table1 . Finally, motivated by recent work on the modelling of two-dimensional random fields (Eckley, Nason and Treloar [7]), we propose a recursive construction for the inner product matrix of (separable) two-dimensional discrete autocorrelation wavelets.... ..."