No need to feel guilty about posting here. It's a learning experience. And do remember that not everything that is posted here is correct -- collaboratory members each speak for themselves, no one else. The general rule is caveat emptor. If you don't understand something, ask. You might get an explanation, you might discover that there is nothing to understand.
There is nothing wrong in using Mathcad to analyze results calculated elsewhere. If you could do the calculations in Mathcad, you'd have some additional avenues of approach and you could directly play with the z parameter (BTW, Mathcad supports Greek letters, if the literature uses ε it may be easier for you (and possibly those familiar with the literature) if you use that also). If, for whatever reason, it is impractical to do the calculations in Mathcad you have to rely on some sort of fitting and interpolation.
Whether the unnormalized data would be better is unclear. Sometimes the original data has simple functional forms that make fitting work better. Sometimes not. But it's still better to start with the original data (if available), as it's easy enough to normalize relative to the smallest value. It's not so easy (if at all possible) to recover the normalization factor, once it's been applied.
Another consideration for calculated data. Retyping data tends to be time consuming and error prone. Not to mention boring. Mathcad has reasonable flexibility in reading files. If you can post the data as data files, as produced by the calculating system, you can usually get Mathcad to read the data, and then manipulate it as needed. Worst case is writing scripted components to read and parse data, as VBScript has goo facilities for dealing with arrays and text strings.
__________________
� � � � Tom Gutman
