There is a new standard Fortran. The official title is “American National Standard Programming Language Fortran, X3.9-1978,” but it is more commonly referred to as “Fortran 77,” since its development was completed in 1977. It replaces the Fortran standard designated X3.9-1966. This paper describes many of the features of Fortran 77 and also provides some information about how and why the standard was developed.
October 1978 - Vol. 21 No. 10
Features
Models for parallel processing within programs: application to CPU: I/O and I/O: I/O overlap
Approximate queueing models for internal parallel processing by individual programs in a multiprogrammed system are developed in this paper. The solution technique is developed by network decomposition. The models are formulated in terms of CPU:I/O and I/O:I/O overlap and applied to the analysis of these problems. The percentage performance improvement from CPU:I/O overlap is found to be greatest for systems which are in approximate CPU:I/O utilization balance and for low degrees of multiprogramming. The percentage improvement from I/O:I/O overlap is found to be greatest for systems in which the I/O system is more utilized than the CPU.
Jump searching: a fast sequential search technique
When sequential file structures must be used and binary searching is not feasible, jump searching becomes an appealing alternative. This paper explores variants of the classic jump searching scheme where the optimum jump size is the square root of the number of records. Multiple level and variable size jump strategies are explored, appropriate applications are discussed and performance is evaluated.
Optimal histogram matching by monotone gray level transformation
This paper investigates the problem of optimal histogram matching using monotone gray level transformation, which always assigns all picture points of a given gray level i to another gray level T(i) such that if i ≥ j, then T(i) ≥ T(j). The objective is to find a transformed digital picture of a given picture such that the sum of absolute errors between the gray level histogram of the transformed picture and that of a reference picture is minimized. This is equivalent to placing k1 linearly ordered objects of different sizes one by one into k2 linearly ordered boxes of assorted sizes, such that the accumulated error of space underpacked or overpacked in the boxes is minimized; the placement function is monotonic, which ensures a polynomial time solution to this problem. A tree search algorithm for optimal histogram matching is presented which has time complexity O(k1 × k2). If the monotone property is dropped, then the problem becomes NP-complete, even if it is restricted to k2 = 2.
Counting large numbers of events in small registers
It is possible to use a small counter to keep approximate counts of large numbers. The resulting expected error can be rather precisely controlled. An example is given in which 8-bit counters (bytes) are used to keep track of as many as 130,000 events with a relative error which is substantially independent of the number n of events. This relative error can be expected to be 24 percent or less 95 percent of the time (i.e. &sgr; = n/8). The techniques could be used to advantage in multichannel counting hardware or software used for the monitoring of experiments or processes.
An analysis of algorithms for the Dutch National Flag Problem
Solutions to the Dutch National Flag Problem have been given by Dijkstra [1] and Meyer [3]. Dijkstra starts with a simple program and arrives at an improved program by refinement. Both of the algorithms given by Dijkstra are shown to have an expected number of swaps which is 2/3N + &thgr;(1) and that these values differ at most by 1/3 of a swap and asymptotically by 1/4 of a swap. The algorithm of Meyer is shown to have expected swap complexity 5/9 N.
Implementing Quicksort programs
This paper is a practical study of how to implement the Quicksort sorting algorithm and its best variants on real computers, including how to apply various code optimization techniques. A detailed implementation combining the most effective improvements to Quicksort is given, along with a discussion of how to implement it in assembly language. Analytic results describing the performance of the programs are summarized. A variety of special situations are considered from a practical standpoint to illustrate Quicksort's wide applicability as an internal sorting method which requires negligible extra storage.
Scatter tables for open addressing benefit from recursive entry displacements, cutoffs for unsuccessful searches, and auxiliary cost functions. Compared with conventional methods, the new techniques provide substantially improved tables that resemble exact-solution optimal packings. The displacements are depth-limited approximations to an enumerative (exhaustive) optimization, although packing costs remain linear—O(n)—with table size n. The techniques are primarily suited for important fixed (but possibly quite large) tables for which reference frequencies may be known: op-code tables, spelling dictionaries, access arrays. Introduction of frequency weights further improves retrievals, but the enhancement may degrade cutoffs.
Cold-start vs. warm-start miss ratios
In a two-level computer storage hierarchy, miss ratio measurements are often made from a “cold start”, that is, made with the first-level store initially empty. For large capacities the effect on the measured miss ratio of the misses incurred while filling the first-level store can be significant, even for long reference strings. Use of “warm-start” rather than “cold-start” miss ratios cast doubt on the widespread belief that the observed “S-shape” of lifetime (reciprocal of miss ratio) versus capacity curve indicates a property of behavior of programs that maintain a constant number of pages in main storage. On the other hand, if cold-start miss ratios are measured as a function of capacity and measurement length, then they are useful in studying systems in which operation of a program is periodically interrupted by task switches. It is shown how to obtain, under simple assumptions, the cache miss ratio for multiprogramming from cold-start miss ratio values and how to obtain approximate cold-start miss ratios from warm-start miss ratios.
A comparison of heaps and the TL structure for the simulation event set
Following publication of our paper [2], questions arose with respect to the superiority of the TL structure over heaps,1 particularly in the face of the remarks of Gonnet [3], concerning the use of heaps for the physical realization of the simulation event set. Gonnet's communication was in response to the Vaucher and Duval paper [5], and suggested the heap to be a more efficient structure than any proposed in [5]. As regards a comparison of heaps and the TL structure we can make the following remarks:
