MVARTOOLS is a Matlab [1] toolbox for doing multivariate analysis and calibration. It is distributed under the GNU General Public License.

MVARTOOLS was made during my M.Sc. thesis work, but was not a part of the thesis. I made it just to familiarize myself with Matlab [2] (and Octave [3]), and to get a better understanding of the PCA, PCR and PLS algorithms. Since 2001, I have not done any development/bug fixing on MVARTOOLS. You have been warned!

MVARTOOLS is not a state-of-the-art industrial strength multivariate package. If that's what you are looking for, I strongly recommend MUST Analyze! [4] (or PLS_Toolbox [5] if you want to work in Matlab).

CONTENTS

• Help and information:
  • contents - package contents
  • mvreadme - release notes
  • mvchanges - changes since previous release
  • mvwarranty - warranty information
  • mvcopying - the GNU general public license
• Data Scaling and Preprocessing:
  • mvcenter - mean centers the data in a row-wise matrix
  • mvobjcenter - object centers a row-wise matrix
  • mvrecenter - re-centers data in a row-wise matrix
  • mvvarscale - scales the data in a matrix to unit variance
  • mvrevarscale - rescales the data in a variance scaled matrix
  • mvdelobjects - delete objects (rows) in a matrix
  • mvmedsmooth - moving median smoothing
  • mvmeansmooth - moving mean smoothing
  • mvsavgol - Savitsky-Golay polynomial smoothing or derivation
  • mvsnv - Standard Normal Variate normalisation
  • mvdigfilt - 1st order digital filter
• NaN values:
  • mvfindnan - find objects (rows) with non-numbers
  • mvreplacenan - find and replace elements with non-numbers
  • mvremnan - removes any rows in a matrix containing NaN
• Plotting:
  • mvscoreplot - plot scores from PCA or PLS
  • mvnumplot - plot with numbers as plot symbols
  • mvstrplot - plot with text strings as plot symbols
  • mvstatplot - plot mean spectrum, std and relative std
  • mvpmplot - predicted vs. measured plot
  • mvregplot - add legend with filename and date
  • mvdensityplot - data density image plot
• Statistics:
  • mvvariance - finds the variance of each column in a matrix
  • mvpmstats - predicted/measured statistics
• Principal Components:
  • mvpcanipals - NIPALS algorithm for Principal Component Analysis
  • mvpcasvd - PCA using singular value decomposition
• Linear Regression:
  • mvmlr - multiple linear regression
  • mvpcrsvd - Principal Component Regression based on SVD
  • mvplsnipals - Partial Least Squares regression using NIPALS
  • mvplsnipcore - PLS NIPALS core
  • mvplsregcoeff - Calculate PLSR regression coefficients
  • mvrpls - Recursive Partial Least Squares regression
  • mvpredict - predict y-values from linear multivariate model
  • mvfullpredict - predict y-values from PLS loading weights
  • mvoptlv - find optimal number of latent variables
• Validation:
  • mvcrossval - cross validation (full, random, block)
  • mvcvbyclass - cross validation by class belongings
  • mvtestval - test-set validation of multivariate model
• Experimental design:
  • mvffdesign - create two-level full/fractional factorial design matrix
• Spectroscopy:
  • mvabstrans - Convert spectral absorbance data to transmittance
  • mvtransabs - Convert spectral transmittance data to absorbance
  • mvintf - calculates interference filter wavelength (and wavenumber)
  • mvaddsine - adds a sine to the input data
  • mvaddrandom - adds random noise to spectra
  • mvaddnrandom - adds normally distributed random noise to spectra
  • mvaddslope - add slope to spectrum
  • mvaddoffset - add offset to spectrum
  • mvaddmult - add multiplicative effect to spectrum
  • mvwlshift - wavelength shift input spectrum
• Miscellaneous:
  • mvmdist - mahalanobis distance
  • mvedist - euclidian distance
  • mvextractclass- extract X and y data for one specific class
  • mvwritecals - write matrix to a CALS table
• Matlab/Octave compability
  • mvomtest - test whether the application is Octave or Matlab
• MVARTOOLS Test data-sets and demo-scripts:
  • acoustics1 - acoustics data: spectra and concentration
  • acoustics2 - acoustics data: spectra, concentration and temperature
  • synthetic - synthetic data: three constituents in a closed data-set
  • iris - the famous iris data-set
  • distdata - data for testing distance measures (mahalanobis/euclidian)
  • mvtest - script for testing all MVARTOOLS functions
  • mvdemo1 - demo-script for doing PCA on the iris data-set
  • mvdemo2 - demo script using PCR
  • mvdemo3 - demo script using PLS
  • mvdemo4 - demo script using factorial design
  • mvdemo5 - demo script using cross- and testset validation
  • mvdemo6 - demo script
  • mvdemo7 - demo script
  • mvdemo8 - demo script
  • mvdemo9 - demo script

NO WARRANTY

Because the program is licensed free of charge, there is no warranty for the program, to the extent permitted by applicable law. Except when otherwise stated in writing the copyright holders and/or other parties provide the program "as is" without warranty of any kind, either expressed or implied, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. The entire risk as to the quality and performance of the program is with you. Should the program prove defective, you assume the cost of all necessary servicing, repair or correction.

In no event unless required by applicable law or agreed to in writing will any copyright holder, or any other party who may modify and/or redistribute the program as permitted above, be liable to you for damages, including any general, special, incidental or consequential damages arising out of the use or inability to use the program (including but not limited to loss of data or data being rendered inaccurate or losses sustained by you or third parties or a failure of the program to operate with any other programs), even if such holder or other party has been advised of the possibility of such damages.