These programs perform various types of multivariate statistical analyses. The input variables can be based on conventional linear measurements or shape variables derived from landmark or outline data.
|RVComparison||Given two a priori defined groups of observations (i.e. two species, two populations) and a set of variables divided in two blocks (modules), this software computes the Escoufier RV coefficient (Escoufier, 1973) for each of the groups, the difference in the coefficient and then performs a permutation test of the null hypothesis of no difference in the RV coefficient between the two groups of observations (Fruciano et al., 2013). Developed by Adriano Franchini. Version 1.0.||7/16/13|
|lanmark reliability||MATLAB software from the 2008 study "Adaptation of Generalizability Theory for Inter-Rater Reliability for Landmark Localization" by ERCAN I, Ocakoglu G, Guney I, Yazici B.
Journal of Tomography & Statistics. Vol. 9, No. S08:51-58.
Example data and software: Link. Developed by Ercan et al.
Coriandis provides a set of graphical and analytical tools to study associations among multivariate datasets (e.g. shapes), using distances among measured individual or species. Coriandis accepts 2D landmark, as well as non-landmark data, including distance matrices (irrespective of how they are computed). For Windows. Link. Developed by Eladio Marquez. Version 1.0.
A program to do fluctuating asymmetry/symmetry analyses with 2D landmark data. For Windows. Link. Developed by Eladio Marquez. Version 1.03.
|Mace and Mace3D||
Programs to do matrix correlations with landmark data in 2D and 3D. For Windows. Link. Developed by Eladio Marquez. Mace version 1.01 and Mace3D version 1.0.
PAD performs a discriminant analysis. It allows one to perform permutations among groups as a non-parametric way to examine the statistical significance of each Mahalanobis distance. For Windows and Linux. Link. Developed by Jean-Pierre Dujardin.
BAC performs a principal component analyses on the total variance matrix or on the consensus matrix when there are various groups . Both kinds of PCA allow bootstraps of the input data to be performed in order to estimate the range of variation of the eigenvector coefficients and of the corresponding allometric coefficients (if data are log-transformed measurements). When performing two or more successive PCA (ACP total), it computes the angle(s) between the corresponding first principal components. There is also the possibility to construct shape components from log-shape ratios. Windows and Linux. Link. Developed by Jean-Pierre Dujardin.
Common Principal Components for dependent random vectors. SAS. The dCPC model considers a situation where the same p variables are measured in k groups (e.g., multiple growth stages) for each observation. The model assumes that there is a common set of principal components that are uncorrelated not only within, but also across groups. A SAS program by Christian Peter Klingenberg.
|LINDA||Linear Discriminant Analysis and Hotteling's T-Squared. Windows 95+. This program computes Hotelling's T-Squared and the linear discriminant function for two groups, in which each individual has been measured in respect of several variables. Supports NTSYSpc format. and the simple "matrix" used by tpsrw. Version 5.23. By Mauro J. Cavalcanti, Centro de Ciências Biológicas, Universidade Santa Úrsula. The older 16 bit Windows version is still available.||6/20/08|
|JACKIE||Principal Components Analysis with jackknife. Windows 95+. R-mode principal component analysis, with jackknife estimates of eigenvectors and eigenvalues. Supports NTSYSpc format. Version 1.32. By Mauro J. Cavalcanti, Centro de Ciências Biológicas, Universidade Santa Úrsula. .||6/20/08|
|MANTEL||Mantel test. Windows 95+. Program for Mantel test. version 1.19. Supports NTSYSpc format. By Mauro J. Cavalcanti, Centro de Ciências Biológicas, Universidade Santa Úrsula.||6/20/08|
|PAST||PAST is a free, easy-to-use data analysis
package aimed at paleontology. Windows 95+.
Multivariate statistical methods include: Principal Components (with Minimal
Spanning Tree), Principal Coordinates (ten distance measures), Non-metric
Multidimensional Scaling (ten distance measures), Detrended Correspondence
Analysis, Cluster analysis (three algorithms, 11 distance measures), k-means
clustering, seriation, discriminant analysis, one-way MANOVA, Hotelling's
t-squared, Box's M, Canonical Variates Analysis. Also includes Fourier shape
analysis, thin-plate splines, parsimony, and time series analysis among
. University of Oslo.
|sizepca||Size-constrained PCA. DOS. An archive of PC software to compute size-constrained principal components analysis. By Keith Somers. Version 6.||4/21/93|
|ICOLATER||Primer by Marcus & Corti. SAS. A primer for a morphometrics workshop containing a variety of SAS IML procedures. Corti and Marcus.||8/9/93|
|ITC5||Data analysis in systematics SAS. Files from "Data analysis in systematics" by Marcus and Corti.||8/9/93|
|Multidimensional paleobiology SAS. Files from Reyment's "Multidimensional Paleobiology". The REYBLCKO file contains the SAS routines, their output, and a readme file. By Les Marcus.||8/9/93|
|Applied factor analysis in the natural sciences MATLAB. Programs and data files from Reyment and Joreskog's "Applied Factor Analysis in the Natural Sciences" By Les Marcus. . The FACDOC file is a Word for Windows file giving additional documentation.||8/9/93|
|CPC||Software for common principal components analysis. DOS, Windows, Mac, Linux, and Unix. Useful for comparing patterns of covariation between populations as developed by Flury. Provides significance testing of the Flury hierarchy via a randomization test, and therefore does not depend on the assumption of normally distributed data and chi-square significance test. Can be used on both phenotypic and quantitative genetic data. Currently in beta test. Used in Phillips and Arnold (1999, Evolution 53:1506).Available with documentation and sample files from the www site by Patrick Phillips.||12/18/00|
|rprimer||A primer on multivariate analysis by Richard Reyment.||12/23/94|
Last modified on
May 4, 2010
Back to SB Morphometrics home page.