Based on a concrete set of working MATLAB programs, the book begins with a short program of snippets describing basic principles and proceeds to a complete application program filling a central analytical need: obtaining pure spectra from the observed overlapping spectra, with standard deviations for the
solutions obtained.
The strength of the book is the open nature of the programs. All programs can be read, tested and modified by the user. The mathematical principles needed for the proper treatment of experimental data are thoroughly described. The first part of the book describes how data is
collected, converted and prepared for the actual deconvolution. Practical working algorithms such as the Savitzky-Golay smoothing method are emphasized. The reader can see how searches in spectral libraries can be greatly speeded up by proper formulation of the calculation. Basic signal processing is described by illustrative examples.
The main part of the book describes the deconvolution of overlapping chromatographic peaks. Principle component analysis is described and used as a useful tool. The main emphasis is a discussion of a deconvolution method, OSCAR (Optimization by Stepwise Constraining of Alternating Regression), developed by the authors. The results can be validated as OSCAR calculates confidence intervals for the spectra and elution curves.
For users who do not want to enter the programs by hand a separate CD-ROM is available. It contains the programs and extensive sample data files. The CD-ROM has instructional multimedia showing step by step how the programs are used for the problems in the book. The MATLAB programs and data
files can be directly run on Windows® and Macintosh® computers having a MATLAB interpreter.
Sprache
Verlagsort
Verlagsgruppe
Elsevier Science & Techn.
ISBN-13
978-0-08-054475-5 (9780080544755)
Schweitzer Klassifikation
Part 1 - Analysis of Hyphenated Data 1. Analysis of overlapping spectra. 2. Data preprocessing. 3. Compression with principal components. 4. Techniques for library searches. 5. Neighborhood operations on hyphenated data. 6. Alternating regression. 7. Applying the OSCAR algorithm - A practical example. 8. Applications in other spectroscopies. 9. Validation. 10. AR and factor analysis. 11. Looking ahead.Part 2 - Computer Programs for Hyphenated Data 12. The starting point. 13. Selecting and preprocessing the raw data. 14. Gathering the statistics from the AR experiment. 15. AR statistics in two dimensions. 16. AR statistics in three dimensions. 17. Selecting an optimal paramater set. 18. Displaying the spectra and elution curves. 19. Calculating the confidence ranges of spectra and elution curves. 20. Looking at the confidence rangs of the spectra and elution curves. 21. Looking at the measurements. 22. Finding help. Appendix. Index.
