Forward coding

Forward modelling is the first step for inverting data for the reconstruction of the defect. For this purpose, the specialised code is better adapted because it is faster than the general code. 

Sponsor/Developing Organization LLNL. Developer. Laurence E. Fried LLNL, P.O. Box 808 Livermore, CA 94551, E-mail cheetah llnl.gov. Hardware-. IBM-PC or clone, Windows 3.1, Windows 95, Mac OS 7.x or later, SUN and SGI workstations, 4.3 MB of hard disk. Software ANSI C. Run execution time for typical problem (CPU or real time). Standard run About 30 seconds on a Power Macintosh 6100/80. Cost None from LLNL. Source code is available, with the stipulations that all modifications be preapproved and forwarded to the sponsor for tracking... 

The events and causal factors chart for this incident is shown in Figure 7.9. The primary sequence of events is shown horizontally in bold boxes. Secondary events are shown in the other boxes, and conditions are in ovals. From the diagram three causal factors were identified and carried forward to the Root Cause Coding to establish the root causes of the causal factors. 

The number and quality of the safety shut-off valves required is specified in the relevant Standards and Codes of Practice. These are usually electrically operated valves conforming to class 1 or class 2 as defined in BS 5963, Specification for electrically-operated automatic gas shutoff valves (1981). This specifies forward pressures and reverse pressures that the valve must withstand in the closed position together with closing times. 

The species shown in red, HB and HA, act as Bronsted-Lowry acids in the forward and reverse reactions, respectively A and B (blue) act as Bronsted-Lowry bases. (We will use this color coding consistently throughout the chapter in writing Bronsted-Lowry acid-base equations.)... 

McGregor DB, Brown A, Cattanach P, et al. 1988. Responses of the L5178Y tk+/tk- mouse lymphoma cell forward mutation assay III. 72 coded chemicals. Environ Mol Mutagen 12 85-154. 

Migrating only the raw data - the characters, numbers, bits, and bytes - forward is not enough to ensure usability. The meta data and the context for the application or database must also be migrated forward. Meta data are the code to the machine-stored bits and bytes. Meta data are the data about the data. They describe the data in the database. The meta data documentation describes the method of data capture, the application used to access the data, security rules for the tables and columns, and other descriptive and procedural information. For derived or calculated data, the algorithm or protocol that was used must be known. The documentation then becomes something else that must be preserved. Without the meta data, the reader will only see a series of alphabetic characters. Without the entire described context associated with the data, the data have no meaning. 

The underlined values in the sample data fit the definition of baseline. You can see how the last non-missing value should be carried forward. The following SAS code selects those proper baseline values. 

McGregor, D.B., A.G.Brown, S.Howgate, D.McBride, C.Riach, and W.J.Caspary. 1991. Responses of the L5178Y mouse lymphoma cell forward mutation assay. V. 27 coded chemicals. Environ. Mol. Mutagen. 17 196-219. 

When the user, whether working on stand-alone software or through a spreadsheet, supplies only the values of the problem functions at a proposed point, the NLP code computes the first partial derivatives by finite differences. Each function is evaluated at a base point and then at a perturbed point. The difference between the function values is then divided by the perturbation distance to obtain an approximation of the first derivative at the base point. If the perturbation is in the positive direction from the base point, we call the resulting approximation a forward difference approximation. For highly nonlinear functions, accuracy in the values of derivatives may be improved by using central differences here, the base point is perturbed both forward and backward, and the derivative approximation is formed from the difference of the function values at those points. The price for this increased accuracy is that central differences require twice as many function evaluations of forward differences. If the functions are inexpensive to evaluate, the additional effort may be modest, but for large problems with complex functions, the use of central differences may dramatically increase solution times. Most NLP codes possess options that enable the user to specify the use of central differences. Some codes attempt to assess derivative accuracy as the solution progresses and switch to central differences automatically if the switch seems warranted. 

To solve the problem a sequential quadratic programming code was used in the outer loop of calculations. Inner loops were used to evaluate the physical properties. Forward-finite differences with a step size of h = 10 7 were used as substitute for the derivatives. Equilibrium data were taken from Holland (1963). The results shown in Table E12.1B were essentially the same as those obtained by Sargent and Gaminibandara. 

The basic idea of plugging in to a code framework shows up in different ways in different languages, including C++ templates, component/connector technologies, and class hierarchies. The latter tend to be overused it is often better to replace the inheritance with composition and explicit forwarding and to use types to document the subtle call patterns between superclasses and subclasses. 

---