WHAT CHECK algorithms

There are various approaches to the data-reduction task. An often used one consists of computing the modulus of the complex phase-detector signal. This removes all offset imperfections as well as any receiver phase misad-justment, bringing us theoretically to what we would have by summing the outputs of two independent, ideal diode detectors. In this case, however, the original signals are still available and can be used to check various aspects of data quality, carry out additional corrections (such as removal of noise-rectification artifacts), or submitted to alternative evaluation algorithms. 

Note diat in die implementation used in this text the PLS loadings are neither normalised nor orthogonal. There are several different PLS 1 algorithms, so it is useful to check exactly what method a particular package uses, although the resultant concen-tration estimates should be identical for each method (unless there is a problem with convergence in iterative approaches). 

It is pointless to check the algorithm s stability when the differential equation is ill-conditioned. What is important is to check the stability of the algorithm for well-conditioned problems. 

We wrote a program to remove in-house idiosyncrasies from the WLN file and to reformat it for CROSSBOW. Programs from this suite were then used to remove outdated conventions, check for what errors could be detected algorithmically and generate molecular formulae. This was achieved and the file, then of some 13,000 FPL Compounds, was subsequently maintained by CROSSBOW. 

Independent of how the internal failure modes are merged and what other abstraction algorithms are applied, the mapping between the realization and the specification must be specified as part of the realization. Additionally, the kind of relation must be specified as part of the specification. Only in this way is it possible to guarantee traceability and to check if the assumed relation between failure realization and specification is true. For example, if modularization is used, the internal modules of the failure realization and the corresponding internal failure modes of the specification must be known in order to check equivalence. This is also true if the internal failure modes are only renamed from the realization to the specification. To efficiently check the equivalence or other qualitative relations between failure specification and realization. Binary Decision Diagrams (BDDs) are used. 

The entire spreadsheet took only a few hours for preparation and checking. This is considerably less than what would have been required for writing a normal algorithmic program to perform the same tasks, especially for a non-skilled programmer. 

---