next up previous contents
Next: Example Up: Truss Stiffness Previous: Data sample for the   Contents

Checking Results of Analysis

It is tempting to "check" the results by comparison with another program. This is probably the simplest check. Whilst such check is undoubtedly useful, it is not necessarily easy nor is it necessarily adequate.

First, it implies that one has at hand another and independent computer program. But wait, how do we know that it is "independent". I am sure that this text has typographical errors. What if both programs followed the methods elucidated in this text, but without their authors bothering to understand, to check the formulae of this text and to correct the errors? If both programmers were competent in programming, they would produce programs that would give the same result, but the wrong result!

Secondly, in many situation one would afford only one program - they are costly. Of course, the Python program listed in this text is useful for the checking of other programs and is not costly, but the first limitation certainly still applies.

For these two reasons, particularly the first one, we consider truly independent checks.



Subsections
next up previous contents
Next: Example Up: Truss Stiffness Previous: Data sample for the   Contents
Algis Kabaila 2006-12-03