The software has been tested on Windows NT, Windows 2000, Windows XP and Windows Vista. Only 32-bit operating systems are supported.


These programs will be made available online and offered at no charge to those who have registered their purchase(s) of this book (and provided an email address).


Bruker1r
Automatically access a directory of Bruker 1D files (1r) copied to a PC and convert the 32-bit binary data into a series of simple ASCII (text) files that can then be assembled into a data matrix suitable for direct input into most multivariate analysis software. Thus, one can acquire any number of data sets (especially relevant to data acquired with autosamplers) and automatically convert them with no user intervention other than selecting the directory in which the data exist.

Peak Align
A program to take individual spectra or even a matrix of data that have been assembled and perform a peak alignment. Based on PAGA but using beam search.


Denoise
For somewhat noisier data it is entirely possible to improve existing quality by application of a denoising algorithm. This wavelet-based method preserves peak positions and peak shapes, a critical requirement for chemometric / metabonomic analyses.


Univariate Var Sel
A selection of univariate preprocessing options are offered, each of which automatically creates new data matrices of reduced (variable) dimensions for subsequent calculations.


Overlay
A plotting utility that allows users to investigate data quality for large numbers of objects/samples.


Offset
A utility to view and correct baseline offsets.


dPCA
A simple but efficient PCA program, including a high quality plotting / graphics utility.


kernelPLS
A PLS algorithm that employs the kernel-method to allow for very fast calculations of large data sets.


Lanczos PLS
Another PLS algorithm that employs a different algorithm for very fast calculations of large data sets.


Concat
A simple utility to create large matrices from individual files. Can be used to combine files one-by-one or to automatically concatenate an entire directory of individual files.


MVA Train/Test
Takes a data matrix that has been assembled and creates separate train/test (and validation) matrices using a variety of options (e.g., random selection, every other sample, every n'th sample, etc). Very effective option to cut-and-paste options that might only exist in most spreadsheets.


dPreprocess
A more complete program containing the ability to load large data sets and then process them in a variety of ways according to many of the methods discussed in the book.