## Cluster analysis and scaling examples |

**carpref.sas**- Principal components analysis of car preferences. An application of the biplot to preference scaling.
**country.sas**- Average judged similarity of countries analyzed by MDS.
**gmds1.sas**- 2D and 3D MDS Solution for Hawaiian InterMarriage Rates. Rates of intermarriage among 8 ethnic groups in Hawaii were used by Parkman & Sawyer (1967) to define distances among these groups which were studied by MDS techniques.
**gmds2.sas**- 3D MDS Solution for Hawaiian InterMarriage Rates
**gtree1.sas**- An illustration of the use of the GTREE macro to draw a graphic tree showing the result of a cluster analysis. This example uses intercity flying distances among cities in the US.
**helm.sas**- Individual Differences in Color Perception explored with MDS, using and INDSCAL solution. Data: Helm's data on color perception in normal and color-blind subjects.
**linpro2.sas**- The LINPRO algorithm, which finds scale positions for a set of variables so that observations have nearly linear profiles, is illustrated with the data on crime rates in US cities.
**mdsauto1.sas**- Similarities of Automobiles. Uses the MDS procedure to find a multidimensional space to prepresent the similarities among 11 models of automobiles. This example illustrates the use of a subsequent cluster analysis as an aid to interpreting the dimensions.
**mdsauto4.sas**- Similarities of Automobiles. An illustration of the MDS procedure to find a multidimensional space to prepresent the similarities among 11 models of automobiles. Ratings of those automobiles on 12 unidimensional properties are then fit to the multidimensional space to aid in interpretation of the dimensions, using the vector model and ideal point model.
**mdsiml.sas**- Classical (metric) MDS with PROC IML
**mutation.sas**- Mutation distances among 20 species analyzed with MDS and CLUSTER The mutation distances were used by Fitch & Margoliash to derive a phylogenetic tree.
**things.sas**- Dissimilarity data for 25 Natural Concepts. Distance estimates of 25 natural concepts were obtained from ratings by psychology undergraduates and biology graduate students (Schvanaveldt data): The items were chosen to have both hierarchical and non-hierarchical relationships, rather than underlying continuous dimensions.
**tregex6.sas**- Preference ratings for automobiles manufactured in 1980 analyzed with PRINQUAL (an MDPREF analysis) and PREFMAP.