Last modified: 2004 Oct 20
Multivariate Statisitics
We will include under this heading factor analysis, cluster
analysis, and multi-dimensional scaling.
-
LINCS provides programs for solving the Covariance Structure problem,
which includes the LISREL model and the Confirmatory Factor Analysis problem as special cases.
(Schoenberg 1994)
The author has provided explicit permission for the code and documentation to reside on and be redistributed from this GAUSS archive.
-
GAUSS-library for multivariate statistical analysis
(Schlittgen May99).
A zip-file with many multivariate statistical procedures.
Includes:
a postscript file with a description of the procedures in the typical shape of the GAUSS manual.
- Cluster analysis.
Peforms agglomerative clustering, plots a dendrogram,
give the clusters up to a specified distance,
k-means clustering.
- PLOTS: starplot, biplot, scatterplotmatrix,...
-
MOMENTS: polychoric corelation, polyserial correlation, one step
minimum volume estimator, ...
-
DISCRIMINANT ANALYSIS: linear discriminant analysis,
logistic regression
- FACTOR ANALYSIS: ML-factor a., PC-factor analysis,
varimax rotation
- MULTIDIMENSIONAL SCALING: metric and nonmetric mds,
monotone regression
- GLIM: generalized linear models with quite a lot
of link functions
- REGRESSION TREES: recursive partition with binary
splits; this proc provides several split criteria and allows for
right censoring (survival data)